thesis for the degree doctor of philosophy ... · thesis for the degree doctor of philosophy by...
TRANSCRIPT
Thesis for the degree Doctor of Philosophy
By Chunxiu Yu
Advisor Prof. Ehud Ahissar
February 2008
Submitted to the Scientific Council of the Weizmann Institute of Science
Rehovot, Israel
י שפם חולדה" קורטיקלי של מגע פעיל ע-עיבוד תלמו Thalamocortical Processing of Active Vibrissal Touch
חבור לשם קבלת התואר דוקטור לפילוסופיה
מאת צנשיו יו
ח"תשס, 'אדר א
מוגש למועצה המדעית של מכון ויצמן למדע
ישראל, רחובות
Regular Format
מנחהחישראהוד א' פרופ
1
Acknowledgements
First and foremost, I am grateful to my mentor, Prof. Ehud Ahissar, for his excellent
guidance, advice, help and encouragement, on both a professional and personal level,
throughout my work. His remarkable generosity and warm caring during the years I
was in his laboratory, made it possible for me to complete my doctoral research.
I would like to thank my colleagues Sebastian Haidarliu, Dori Derdikman, Knarik
Bagdasarian, Naama Rubin and Daniel Goldman. Without their excellent technical
instruction during the initial time and invaluable assistance throughout the course of
my work, I would not be possible to completing my experiments described.
I would like to thank Per Magne Knutsen, Marcin Szwed and all present members of
the laboratory for their friendship, support, and technical assistance, and for creating a
scientifically stimulating and very pleasant atmosphere.
I would also like to express my appreciation and gratitude to Prof. Shabtai Barash and
Dr. Ilan Lamp for their excellent and thoughtful advices and great help during my
doctoral research.
I am forever indebted to Xiaohu Tang, my husband, for his daily support, and for his
love and encouragement.
2
Table of Contents SUMMARY.............................................................................................................................................3 5..........................................................................................................................................................סיכוםINTRODUCTION ..................................................................................................................................7
ACTIVE VIBRISSAL TOUCH....................................................................................................................7 AFFERENT PATHWAYS VIA THE THALAMUS ..........................................................................................9 SECONDARY SOMATOSENSORY CORTEX .............................................................................................11
CHAPTER 1: PARALLEL THALAMIC PATHWAYS FOR WHISKING AND TOUCH SIGNALS IN THE RAT ......................................................................................................................14
SUMMARY ..........................................................................................................................................14 BACKGROUND ....................................................................................................................................15 MATERIALS AND METHODS................................................................................................................16 RESULTS.............................................................................................................................................18
Anatomical Borders ......................................................................................................................18 Specificity of Thalamic Responses ................................................................................................20 Comparison of Latency and Duration of Responses .....................................................................23
DISCUSSION........................................................................................................................................25 SUPPORTING INFORMATION................................................................................................................28
CHAPTER 2: THALAMIC CODING OF ACTIVE VIBRISSAL TOUCH...................................32 SUMMARY ..........................................................................................................................................32 BACKGOUND ......................................................................................................................................33 METHODS ...........................................................................................................................................34 RESULTS.............................................................................................................................................34
Artificial whisking paradigm ........................................................................................................34 Representations of object location during artificial whisking ......................................................36 Representations of whisking frequency during artificial whisking ...............................................42
DISCUSSION........................................................................................................................................46 CHAPTER 3: CODING OF OBJECT LOCATION IN THE SECONDARY SOMATOSENSORY CORTEX DURING ACTIVE VIBRISSAL TOUCH..................................52
SUMMARY ..........................................................................................................................................52 BACKGROUND ....................................................................................................................................53 MATERIALS AND METHODS................................................................................................................55 RESULTS.............................................................................................................................................57
Coding of object locations ............................................................................................................58 Comparison of response dynamics................................................................................................62 Comparison of adaptation ............................................................................................................67 Comparison of response latencies ................................................................................................68
DISCUSSION........................................................................................................................................70 DISCUSSION........................................................................................................................................77
PARALLEL THALAMIC AND CORTICAL PROCESSING OF ACTIVE TOUCH ...............................................77 SENSORY MOTOR LOOPS .....................................................................................................................79 CODING OF HORIZONTAL OBJECT LOCATION.......................................................................................82 ARTIFICIAL WHISKING ........................................................................................................................84 EFFECTS OF ANESTHESIA ....................................................................................................................86
REFERENCES .....................................................................................................................................87 LIST OF PUBLICATIONS .................................................................................................................97 DECLARATION OF INDEPENDENCE OF PH.D WORK............................................................98
3
SUMMARY In active sensation, sensory information is acquired via movements of sensory
organs; rats move their whiskers repetitively to scan the environment, thus detecting,
localizing, and identifying objects. Recently, general principles of encoding of active
touch by the rat trigeminal ganglion neurons were uncovered. However, the processes
employed by the rat brain to decode sensory information obtained during active
sensing are not yet understood. The general objective of my Ph.D study has been to
understand the processing of sensory information performed in the thalamus and
cortex of the vibrissal system during active touch. For this purpose, in the first part of
my Ph.D study, I recorded the activity of individual neurons from three thalamic
nuclei of the whisker system, each belonging to a different major afferent pathway:
paralemniscal, extralemniscal (a recently discovered pathway), or lemniscal – while
inducing artificial whisking in anesthetized rats. I found that different sensory signals
related to active touch are conveyed separately via the thalamus by these three parallel
afferent pathways. The paralemniscal pathway conveys sensor motion (whisking)
signals, the extralemniscal conveys contact (touch) signals, and the lemniscal pathway
conveys combined whisking-touch signals (Yu et al., 2006). Furthermore, I
investigated how thalamic neurons encode the horizontal object location and whisking
frequency (Chapter 2). I found that Touch neurons encoded the horizontal coordinate
of object location by first-spike timing. In contrast, Whisking-Touch neurons did not
convey any information about object location by their first-spike timing. Some
Whisking-Touch neurons encoded horizontal object location by inter-spike intervals
and by spike counts. POm neurons did not encode object touch, but exhibited both
spike counts and first-spike timing modulations as whisking frequency was modulated
during both free air whisking and touch conditions. My findings suggest that
thalamocortical processing occurs in at least three streams: The paralemniscal
pathway processing self-generated motion of the whiskers, primarily in the temporal
domain. The lemniscal pathway processing object identity, primarily in the spatial
domain, though tightly coordinated with whisker motion. The extralemniscal pathway
processing contact information related to object location.
Anatomical studies indicate that the major projections of VPMvl are to deep layers
of the secondary somatosensory cortex (S2), which is also a target of POm. Thus, we
conjectured that Whisking and Contact signals could be integrated in S2 in order to
4
decode horizontal object location. So far, the function of S2 has been poorly
understood. Thus, in the last part of my Ph.D study (Chapter 3), I decided to explore
the processing of tactile information performed in S2 during active whisking.
Specifically, I investigated what sensory information is conveyed by S2 neurons and
how S2 neurons represent the horizontal coordinate of object's location during active
vibrissal touch. I found that the deeper layers of S2 (S2-L46) contain a relatively large
proportion of neurons (41%) which are selective to specific horizontal locations of
contacted objects. In comparison, the superficial layers of S2 (S2-L23, 19%), the three
tactile thalamic nuclei (<23%), and three layers in S1 (<13%) exhibited significantly
smaller proportions. Response dynamic analysis suggests that there are at least two
cortical streams that process object information in parallel.
Overall, the results of my Ph.D study demonstrate that sensory processing in the rat
vibrissal system utilizes parallel processing streams that are organized in functionally
specialized thalamic and cortical microcircuitry, and reveal those circuits that are
actually involved in object localization via whisking.
5
סיכום
מערכת את מניעות חולדות: החוש אברי תנועת באמצעות נרכש חושי מידע, פעילה חישה במסגרת
. המרחבי ומיקומם אופיים, עצמים מזהות, זה ובכלל, סביבתן את לסרוק כדי מחזורי באופן השפם שיערות
, זאת עם. מינאלי'הטריג בגנגליון אשר העצב בתאי הפעילה החישה של הקידוד עקרונות התגלו, לאחרונה
המטרה. עדיין מובנים אינם, החולדה במוח, פעילה בחישה שנרכש, החושי המידע מפוענח בהם התהליכים
השפם למערכת המיוחדים באזורים החושי דעהמי מעובד בו האופן את להבין הייתה שלי בתיזה המנחה
פעילותם את רשמתי עבודתי של הראשון בחלקה, כך לשם. פעילה חישה בזמן, ובקורטקס בתלמוס
מהם אחד כל, השפם שערות ממערכת מידע המעבדים תלמיים גרעינים בשלושה עצב תאי של החשמלית
למניסקלי-האקסטרה, )POm גרעין (למניסקלי-הפרה: העיקריים האפרנטיים המסלולים לאחד שייך
שפם תזוזת עירור כדי תוך ערכתי הרישומים את). VPMdml גרעין (הלמניסקלי או, )VPMvl גרעין(
בנפרד מעובדים פעילה לחישה השייכים שונים תחושתיים אותות כי מצאתי. מורדמות בחיות מלאכותית
תנועת את מעבד למניסקלי-הפרה המסלול. מקבילים במסלולים, שלעיל התלמיים הגרעינים שלושת י"ע
משלב -והלמניסקלי, (touch) עצמים עם השפם מגע את -למניקלי-האקסטרה, (whisking) החוש איבר
את מקודדים תלמיים עצב תאי כיצד חקרתי בנוסף. Yu et al., 2006), ראשון פרק (ומגע תנועה אותות
השפם תנועות תדר ואת עצמים של) קאודאלי -רורוסט: החולדה של האורך ציר על (האופקי המיקום
הפעולה פוטנציאל תזמון י"ע עצם של האופקי מיקומו את מקודדים מגע תאי כי מצאתי). שני פרק(
פוטנציאל תזמון י"ע העצם מיקום לגבי כלשהו מידע ייצגו לא מגע -התנועה תאי, לכך בניגוד. הראשון
בין המפריד הזמן באמצעות האופקי מיקום קודדו מגע -התנועה מתאי אחדים. שלהם הראשון הפעולה
מגע קודדו לא POm -ב אשר העצב תאי. הפעולה פוטנציאלי מספר באמצעות וכן הפעולה פוטנציאלי
בתדר שינויים עם הראשון פ"הפ תזמון של והן הפעולה פוטנציאלי מספר של הן אפנון הראו אולם, כלל
כי מרמזים ממצאי). באוויר חופשית תנועה (לאו אם ובין התנועה לבמסלו עצם היה אם בין, השפם תנועת
תנועות מעבד למניסקלי-הפרה המסלול: מסלולים שלושה בלפחות נעשה קורטקס -התלמוס ברמת עיבוד
-המרחב במימד בעיקר, עצמים זהות מעבד, הלמניסקלי המסלול. הזמן במימד בעיקר, השפם של עצמיות
הקשור מגע לגבי מידע מעבד, למניסקלי-האקסרטה המסלול. השפם נועתת עם המתואם באופן אולם
.במרחב עצם למיקום
הקורטקס של העמוקות השכבות את בעיקר מעצבב VPMvl שה כך על מצביעים אנטומיים מחקרים
עשויים ותנועה מגע אותות כי שיערנו, לכן. POm -מה גם עצבוב שמקבל, (S2) המשני הסומאטוסנסורי
S2 של התפקיד את הבנתנו הייתה, עתה עד. עצם של האופקי מיקומו את לייצג בכדי S2 -ב להתמזג
מידע מעובד שבו האופן את לחקור החלטתי, )3 פרק (עבודתי של האחרון בחלקה, לכן. בחסר לוקה
-ה בתאי המעובד החושי המידע של אופיו את לתאר ביקשתי, בפרט. פעילה חישה בזמן S2 -ב המישוש
S2 של יותר העמוקות בשכבות כי מצאתי. אקטיבית חישה בזמן עצם של האופקי מיקומו מיוצג צדוכי S2
בעצם למגע מגיבים דהיינו, סלקטיבית שמגיבים תאים של) %41 (יחסית ניכר אחוז קיים) 4-6(
, יותר נמוכים באחוזים מופיעים סלקטיביים תאים, לעומתן. בהם ורק מסוימים אופקיים במיקומים
6
S1 -ב השכבות ובשלוש, (23%>) התלמיים הגרעינים בשלושת, (S2-L23, 19%) העליונות בותבשכ
לעיבוד מסלולים שני לפחות של האפשרי קיומם על הצביע התגובות דינאמיקת של ניתוח. (13%>)
.הסומאטוסנסורי בקורטקס מקביל באופן עצבים
על מתבסס חולדות של השפם יערותש במערכת חושי עיבוד כי מראות שלי המחקר תוצאות, ככלל
, )וקורטיקליים תלמיים (מותאמים פונקציונאליים מעגלים בתתי המאורגנים, מקבילים עיבוד מסלולי
. פעילה חישה בזמן, במרחב עצמים של המיקום בתפישת המעורבים המעגלים על ומצביעות
7
INTRODUCTION
Tactile sensation is an active process that requires a precise and efficient interface
between receptor arrays and sensed object. To perceive stationary objects, the sensory
organ has to move. Thus primates move their fingers across surfaces they try to
identify, and rodents, such as rats, move their whiskers in order to localize and
identify objects (Ahissar and Arieli, 2001).
Rodent whisking system is a good model for studying sensory systems, both at the
anatomical, physiological and behavioral levels (Kleinfeld et al., 2006; Staiger, 2006;
Brecht, 2007; Gopal and Hartmann, 2007; Petersen, 2007; Schubert et al., 2007).
Studies conducted so far revealed important aspects regarding the principles of
operation of the vibrissal system, such as the principles of peripheral encoding of
object location and texture, mechanisms of sensory adaptation, responses and
effective power of individual neurons at all processing levels, response selectivity,
plasticity of cortical representations and circuits, temporal and spatial characteristics
of sensory-motor feedback processes, patterns of whisking behavior, and involvement
of whisking patterns in solving behavioral tasks, etc. Specifically, the vibrissal
sensory and motor pathways and representations are relatively large and well-defined
in the rodent brain. Many parts of the vibrissal sensorimotor system have already been
characterized, at least partially, and the connectivity scheme is largely known.
Active vibrissal touch During exploration, rats use a combination of body, head and whisker movements
to effectively comb the environment for interesting features. The immediate space
which surrounds the snout of the rat can be geometrically described by three
orthogonal axes, which correspond to the geometry of the whisker pad to describe this
space in head-centered coordinates: a vertical axis runs along the whisker arcs, a
horizontal axis along whisker rows and a radial axis extends outwards from the skin.
In this scheme, the movements of the whiskers can be described in terms of the head-
centered whisker angles azimuth and elevation.
Behavioral evidence shows that rats are excellent at discriminating the horizontal
location of objects. Rats outperform the horizontal acuity limit imposed by the
granularity of the whisker array. Interestingly, if all but one whisker are removed on
8
each side of the snout, rats still reach hyperacuity and even improve their performance
compared to rats localizing with all whiskers intact (Knutsen et al., 2006).
Rats are nocturnal animals, and dwell in extensive networks of underground
tunnels. In such an environment, the texture and position of tunnel walls provide rich
tactile experiences as rats navigate around their living quarters. The proximity of a
tunnel wall is an important stimulus, as it regulates running direction and velocity
(Horev et al., 2007). As a consequence, rats are easily trained and quite efficient at
discriminating the radial distance of objects, reaching a radial acuity of about 3 mm
(Krupa et al., 2001; Shuler et al., 2002) through whole-body movements.
Behavioral evidence further reveals that whisker movements along the vertical axis
are small, and typically smaller than the spacing between two neighboring rows.
These vertical movements are caused by the activity of extrinsic muscles that act to
move the whole whisker-pad together. Thus, even if the absolute elevation of the
whiskers may vary, the change in relative elevation of different whiskers is always
much smaller (Bermejo et al., 2002). Combining with the fact that receptive fields of
first-order trigeminal afferents are always single-whisker (Zucker and Welker, 1969;
Gibson and Welker, 1983b; Lichtenstein et al., 1990; Szwed et al., 2003), these
observations suggest that the vertical dimension is most likely spatially encoded,
possibly using the identities of activated whiskers.
Studying sensory processing in behaving animals is a challenge since sensory-
evoked efferent responses may interfere with sensory response on relatively fast time-
scales (Nguyen and Kleinfeld, 2005). To eliminate such effects, Szwed and colleagues
in our lab opened this system of closed-loops by anesthetizing rats and cutting the
facial motor nerve (Szwed et al., 2003). They then attached electrodes to the distal
part of the motor nerve and applied patterns of electrical stimulation that produced
movements mimicking natural free-air whisking (Zucker and Welker, 1969; Brown
and Waite, 1974). With this method, whiskers are moved forward by their muscles,
and thus whisker-object interaction mimics that which occurs naturally, i.e., forces are
applied both to the whisker’s follicle and to the whisker’s shaft. Using this
electrically-driven whisking, Szwed and colleagues successfully investigated the
encoding principle of active touch by rat trigeminal ganglion (TG) neurons (Szwed et
al., 2003; Szwed et al., 2006). Their results indicated that different sensory signals
related to active touch are conveyed separately by different classes of first-neurons in
TG: Whisking, Whisking-touch and Touch. Whisking-responsive neurons fired at
9
specific deflection angles, reporting the actual whiskers' position with high precision.
Whereas Touch-responsive neurons could encode horizontal object locations by spike
timing and radial locations by firing rate. However, it is still largely unknown what
the decoding principle of active touch is at downstream brain stations.
Afferent pathways via the thalamus During tactile exploration, rats actively move their whiskers across objects or
surfaces in repeated rhythmic sweeps (5-25Hz) (Carvell and Simons, 1990; Fanselow
and Nicolelis, 1999; Kleinfeld et al., 1999; Berg and Kleinfeld, 2003b; Knutsen et al.,
2006). Each whisker sits in a follicle that contains various types of mechanosensory
endings, which originate from a branch of the trigeminal maxillary nerve (infraorbital
branch of the 5th cranial nerve) (Rice et al., 1986; Ebara et al., 2002), and provide
input to the first stage of the vibrissal system – trigeminal ganglion neurons
(Lichtenstein et al., 1990; Szwed et al., 2003).
Somatosensory information from the vibrissae of the rat is first encoded by the
trigeminal ganglion neurons, and then carried to the second-order neurons within the
brainstem trigeminal complex; Anatomical studies show that it reaches to the
neocortex via the thalamus in three parallel pathways: the leminiscal pathway,
paralemniscal pathway and extraleminscal pathway (a recently discovered pathway).
The lemniscal pathway ascends via the dorso-medial sector of the ventral
posteromedial nucleus (VPMdm), which mainly receives vibrissal input from the
principal nucleus of the brainstem (PrV) (Bruce et al., 1987; Rhoades et al., 1987;
Chiaia et al., 1991; Williams et al., 1994); the paralemniscal ascends via a rostral
sector of the posterior complex (POm), which receives its main afferent input from
the rostral interpolar nucleus of the spinal trigeminal complex (SpVir) (Erzurumlu and
Killackey, 1980; Chiaia et al., 1991; Friedberg et al., 1999; Veinante et al., 2000), and
extralemniscal pathway ascends via the ventro-lateral sector of the VPM (VPMvl)
(Pierret et al., 2000). We refer to the recently discovered pathway as extralemniscal,
to denote its path, which emerges from the caudal segment of the trigimial interpolar
nucleus (SpVic) in the brainstem and ascends in parallel to the lemniscal and
paralemniscal pathways (Pierret et al., 2000).
These three pathways convey their information to different targets. The leminical
pathway projects principally to “barrel” in layer 4 and to layers 5b and 6a of
10
neocortex (Woolsey and Van der Loos, 1970; Chmielowska et al., 1989; Lu and Lin,
1993). The paralenmiscal pathway terminates more broadly to layers 1 and 5a and to
the septal zones surrounding the layer 4 barrels (Koralek et al., 1988; Chmielowska et
al., 1989; Lu and Lin, 1993). The extralemnical pathway mainly projects to the
second somatosensory area (S2) and the septal regions of the barrel field (Pierret et
al., 2000).
The lemniscal and paralemniscal pathways differ considerably in their responses to
stimuli applied passively to stationary whiskers. In response to passive whisker
stimulation, VPM neurons exhibit short latency responses to a single whisker, but
whisker-sensitive neurons in POm are slowly responsive to multiple whiskers with
longer latency, and have larger receptive field (RF) center under deep anesthesia
(Diamond et al., 1992b). Furthermore, the comparison of response adaptation and
temporal transformations of VPM and POm shows that neurons in VPM respond to
frequency changes with amplitude modulations and constant latencies, whereas
neurons in POm respond with modulations in latency and in amplitude. These
modulation differences suggest that the lemniscal system is involved in processing the
spatially encoded information, whereas paralamniscal pathway processes temporally
encoded information , in which the main processing in paralemniscal system is a
transformation of a temporal code to a rate code (Ghazanfar and Nicolelis, 1997;
Ahissar et al., 2000; Sosnik et al., 2001). Finally, a number of previous reports have
assessed that vibrissae sensitive neurons in VPM exhibit more rapidly and less slowly
adapting responses to sustained vibrissa deflection (Ito, 1988; Simons and Carvell,
1989; Hartings and Simons, 2000). The responses of neurons are directionally well-
tuned (Ito, 1988; Brecht and Sakmann, 2002a; Minnery et al., 2003; Timofeeva et al.,
2003), and highly correlated with the velocity of whisker deflection, but much less
relate to deflection amplitude (Pinto et al., 2000). Little information is available
concerning these response characteristics of neurons in POm. Moreover, no
physiological study has reported distinct response properties for cells in the
extralemniscal pathways. Importantly, information about the signals conveyed by
these pathways, and by the extralemniscal one, during active touch is lacking.
In the first part of this thesis (chapter 1 and 2), using artificial active whisking, I
investigated what the decoding principle of active touch is in the thalamus.
Specifically, I explored the repertoire of response of thalamic neurons and investigate
how different sensory signals related to active touch are conveyed via the thalamus. I
11
found that different sensory signals related to active touch are carried separately by
the thalamus: Whisking signals are conveyed by POm neurons along the
paralemniscal pathway, Touch signals are conveyed by VPMvl neurons along the
extralemniscal pathway and combined Whisking-Touch signals are conveyed by
VPMdm neurons along the lemniscal pathway (Yu et al., 2006) (Chapter 1). In
addition, I investigated that how these thalamic neurons encode the horizontal
location of the objects and what the encoding principle of whisker frequency is during
active whisking (Chapter 2).
Secondary somatosensory cortex Somatosensory cortex in the rat, as in other species, is typically divided into
primary and secondary regions. The primary somatosensory cortex in the rat (S1)
has been well characterized anatomically and functionally in previous studies. In
contrast to S1, the secondary somatosenory cortex (S2) has been more anatomically,
but less functionally explored (Carvell and Simons, 1987; Koralek et al., 1988; Fabri
and Burton, 1991; Spreafico et al., 1987; Remple et al., 2003; Brett-Green et al.,
2004; Kwegyir-Afful and Keller, 2004; Benison et al., 2006; Melzer et al., 2006;
Kamatani et al., 2007)
Anatomical and physiological studies in the rat show that S2 is located posterior
and lateral to S1 in parietal cortex, and shares an extensive common border with the
barrel field of the S1 face area, just rostral to auditory cortex. S2 contains a coarser
representation of the body and vibrissae, and is reciprocally and topographically
interconnected with S1. S2 whisker representations originate from both the septa- and
the barrel-related regions in all extragranular layers of S1 barrel cortex (Fabri and
Burton, 1991; Kim and Ebner, 1999; Alloway et al., 2003; Chakrabarti and Alloway,
2006). S2 also receives substantial inputs from topographically appropriate regions
within the ipsilateral ventrobasal nucleus and from the ipsilateral posterior group
(Carvell and Simons, 1987; Spreafico et al., 1987; Alloway et al., 2000; Pierret et al.,
2000), and has dense reciprocal connections with the ipsilateral motor cortex (Carvell
and Simons, 1987). So far, the exact functional role has remained less understood.
Studies combining somatosensory with either auditory or visual stimulation have
suggested that one of the roles may be in multi-sensory integration (Di et al., 1994;
Barth et al., 1995; Brett-Green et al., 2003; Wallace et al., 2004). However, recent
12
studies based on electrophysiological recording proposed that multisensory cortex
should be considered functionally and anatomically distinct with S2, and that S2 may
perform more unisensory processing tasks (Menzel and Barth, 2005).
Physiological and anatomical investigations in a variety of species have identified
consistent similarities as well as important differences in the organization of S2 and
S1 cortical areas. However, it is still unclear whether information processing in the
cortical somatosensory network is in serial or in parallel. Although a widely held
view of cortical processing is that the sensory information is performed by serial,
hierarchical processing of gradually more complex stimulus features (Pons et al.,
1987; Pons et al., 1992; Burton et al., 1995; Inui et al., 2004), a growing body of
evidences in support of the parallel processing scheme has obtained from direct
inactivation of S1. The studies in a group of species, such as cat (Turman et al., 1992;
Turman et al., 1995), tree shrew (Garraghty et al., 1991), rabbit (Murray et al., 1992),
possum (Coleman et al., 1999) and monkeys (Zhang et al., 1996; Zhang et al., 2001b;
Zhang et al., 2001a), suggest that S1 inactivation has almost no effect on S2 tactile
responsiveness. The evidence from human studies also show that S2 receives direct
thalamic inputs and are simultaneously activated with S1 during the early processing
of sensory input (Karhu and Tesche, 1999; Barba et al., 2002). Correspondingly, it is
still not clear whether whisker-related information is serially or parallely processed in
S1 and S2 of the rat.
Recent electrophysiological study in the rat reported that there are no latency
differences between neurons in all layers of S2 and those in S1 layer 4 barrels
(Kwegyir-Afful and Keller, 2004). Similar to S1 septa neurons, S2 neurons respond to
multiple whiskers with comparable onset latencies (Fabri and Burton, 1991; Brecht
and Sakmann, 2002b; Remple et al., 2003; Kwegyir-Afful and Keller, 2004) and
represent a single whisker deflection with lower response magnitudes. Moreover, S2
neurons present similar angular preferences as those in S1 Layer 4. Collectively, these
similarities in responsiveness suggest that S2 and S1 may play functionally
independent roles in the processing of tactile input. Additionally, in those studies, no
differences were found between layers within S2; using a passive paradigm, neurons
in all layers of S2 exhibit consistent response properties, such as the receptive field,
latencies, angular tuning, response duration and magnitude (Pierret et al., 2000). So
far, the functional organization within S2 is still unknown.
13
In the rat whisker system, the information about active touch flows through three
parallel and functional distinctive channels in the thalamus (Yu et al., 2006).
Moreover, the laminar differences in S1 during active touch match the differences
between their thalamic counterparts: S1L5a neurons are primarily sensitive to
whisking similar to POm; Whereas S1L4 bareel neurons are sensitive to both
whisking and touch similar to those in VPMdm. However, no neurons in S1 are
sensitive to contact only, such as those in the VPMvl (Derdikman et al., 2006a; Yu et
al., 2006). Anatomical studies show that the major projections of VPMvl are to deep
layers of S2, which is also a target of POm (Carvell and Simons, 1987; Spreafico et
al., 1987; Koralek et al., 1988; Alloway et al., 2000; Pierret et al., 2000), Thus, S2
may integrate the Whisking and Contact signals to decode object locations. Therefore,
in the last part of this thesis (chapter 3), I explored the processing of sensory
information performed in S2 during active touch. Specifically, I investigated what
external information related to object localization is conveyed by S2 neurons and how
S2 neurons encode object location compared to S1 and thalamus during active
whisking. My results indicate that there is a functional segregation within S2; the
deeper layers of S2 contain high proportion of location-selective neurons, and these
neurons seem to function in parallel to neurons of S1.
14
CHAPTER 1: Parallel thalamic pathways for whisking and
touch signals in the rat1
Summary
In active sensation, sensory information is acquired via movements of sensory
organs; rats move their whiskers repetitively to scan the environment, thus detecting,
localizing, and identifying objects. Sensory information, in turn, affects future motor
movements. How this motor-sensory-motor functional loop is implemented across
anatomical loops of the whisker system is not yet known. While inducing artificial
whisking in anesthetized rats, we recorded the activity of individual neurons from
three thalamic nuclei of the whisker system, each belonging to a different major
afferent pathway: paralemniscal, a recently discovered pathway (extralemniscal), or
lemniscal. We found that different sensory signals related to active touch are
conveyed separately via the thalamus by these three parallel afferent pathways. The
paralemniscal pathway conveys sensor motion (whisking) signals, the extralemniscal
conveys contact (touch) signals, and the lemniscal pathway conveys combined
whisking-touch signals. This functional segregation of anatomical pathways raises the
possibility that different sensory-motor processes, such as those related to motion
control, object localization, and object identification, are implemented along different
motor-sensory-motor loops.
1 Yu, C., Derdikman, D., Haidarliu, S., and Ahissar, E. (2006) Parallel thalamic pathways for
whisking and touch signals in the rat. PLoS Biol, 4(5): e124.
15
Background
Active touch is a closed-loop process in which sensor motion determines the
sensory input and the sensory input determines future sensor motion (Gibson, 1962;
Katz, 1989, Original work published 1925 Original work published 1925). Both
sensor motion and touch signals are reported to the brain by peripheral neurons. Limb
movements are reported via proprioceptive mechanoreceptors located in joints,
tendons, and muscle spindles. Whisker movements are reported to the brain via
mechanoreceptors located in the whisker follicle. As with limb movements, whisker
movements are reported by a set of receptors that is separated from those sensing
touch (Szwed et al., 2003).
Whisker afferents ascend via the thalamus in three parallel pathways: the lemniscal
pathway ascends via the dorso-medial (dm) sector of the ventral posteromedial
nucleus (VPM) (VPMdm), the paralemniscal ascends via a rostral sector of the
posterior complex (POm), and a recently discovered pathway ascends via the ventro-
lateral sector of the VPM (VPMvl) (Pierret et al., 2000). We refer to the recently
discovered pathway as “extralemniscal” to denote its path, which emerges from
paralemniscal nuclei in the brainstem and ascends in parallel to the lemniscal and
paralemniscal pathways (Pierret et al., 2000). The paralemniscal, extralemniscal, and
lemniscal pathways appear to be trigeminal analogs of the spinal spinothalamic,
neospinothalamic, and dorsal column–lemniscal pathways (Bishop, 1959),
respectively. These pathways convey their information to different targets (Cadusseau
and Roger, 1990; Deschenes et al., 1998; Pierret et al., 2000), which close the
sensory-motor loop at different levels of brain hierarchy (Kleinfeld et al., 1999;
Guillery and Sherman, 2002), with the lemniscal involving the highest, the
extralemniscal a lower, and the paralemniscal a still lower level. The lemniscal and
paralemniscal pathways differ considerably in their responses to stimuli applied
passively to stationary whiskers (Diamond et al., 1992a; Friedberg et al., 1999;
Ahissar et al., 2000; Brecht and Sakmann, 2002b). However, information about the
signals conveyed by these pathways, and by the extralemniscal one, during active
touch is lacking (Pierret et al., 2000).
We combined active whisking with controlled stimulus application and accurate
localization of recording sites by employing artificial whisking in anesthetized rats
(Zucker and Welker, 1969; Brown and Waite, 1974; Szwed et al., 2003; Arabzadeh et
16
al., 2005). With this method, whiskers are moved forward by their muscles, and thus
whisker–object interaction mimics that which occurs naturally, i.e., forces are applied
both to the whisker’s follicle and to the whisker’s shaft. In contrast, when stimulating
passive whiskers, i.e., when the object moves an otherwise stationary whisker, forces
are applied only to the whisker’s shaft. Using this artificial active whisking method,
we previously identified three types of active-touch signals (whisking, touch, and
combined whisking–touch) sent by trigeminal ganglion (TG) neurons to the brain
(Szwed et al., 2003). Here, we used the same method to examine conveyance of
active-touch signals via the thalamus.
Materials and Methods Surgical and recording procedures.
Experiments were performed on 40 male Albino Wistar rats weighing 200–300 g,
using experimental protocols as previously described (Szwed et al., 2003). Briefly,
surgery was performed under general anesthesia (urethane; 1.5 g/kg,
intraperitoneally), with supplemental doses of anesthetic (10%) being administered
when required. Atropine methyl nitrate (0.3 mg/kg, intramuscularly) was administered
to prevent respiratory complications. Anesthetized animals were secured in a
stereotaxic device (SR-6; Narishige, Tokyo, Japan), and their body temperature
maintained at 37 °C. An opening was made in the skull overlying the right thalamus,
and tungsten microelectrodes (0.5–1 MΩ; Alpha Omega Engineering, Nazareth,
Israel) were lowered according to known stereotaxic coordinates of POm and VPM
until units drivable by whisker stimulations were encountered. Up to four electrodes,
spaced 0.33 mm from each other, were lowered in parallel in each recording session.
Standard methods for single-unit recordings were used (Szwed et al., 2003). Single
units were sorted by spike templates. Units were considered single only if they had
homogenous spike shapes that did not overlap with other units or noise and if they
exhibited refractory periods of > 1 ms in their autocorrelation histograms. Artifacts
produced by electrical stimulation were isolated by an online spike-sorter (MSD-3.21;
Alpha-Omega Engineering) and removed from unit recordings. Experimental
procedures were approved by the Institutional Animal Care and Use Committee of
The Weizmann Institute of Science.
17
Experimental paradigms.
We induced trains (5 Hz, 50% duty cycle, 2 s) of artificial whisking followed by
intertrain intervals of 3 s in blocks of 12, 18, or 24 trains (trials) each. Artificial
whisking was induced as described in (Szwed et al., 2003). In brief, the facial nerve
was cut and its distal end mounted on a pair of silver electrodes. Bipolar, rectangular
electrical pulses (0.5–4.0 V, 40 μs duration) were applied through an isolated pulse
stimulator (Model 2100; A-M systems, Sequim, Washington, United States) at 83 Hz,
the lowest frequency that still produces continuous whisker movement. Whisker
movements were recorded at 1,000 frames/sec with a fast digital video camera
(MotionScope PCI 1000; Redlake, San Diego, California, United States). Video
recordings were synchronized with neurophysiological data with 1 ms accuracy
(Szwed et al., 2003; Knutsen et al., 2005). Blocks of free-air artificial whisking were
interleaved with blocks of artificial whisking against an object positioned in front of
the principal whisker, i.e., the whisker that produced the maximal response in the
recorded cell during manual passive stimulations. The object was a vertical pole (2-
mm diameter), positioned at three different horizontal distances from the resting
position of the whisker; the distance of the object from the skin was 70%–90% of the
whisker’s length. Horizontal distances of the object from the resting position of the
whisker ranged from 1 to 9 mm (median = 3 mm). Each of the four whisking
conditions (free-air and three object positions) was repeated in at least two blocks,
interleaved in time. Results of touch trials were averaged over all three object
positions. In order to mimic as close as possible natural conditions, all the whiskers of
the mystacial pad were left intact throughout an experiment. Thus, whiskers other
than the principal whisker also contacted the object during whisker movement.
However, we verified that the principal whisker was always the first whisker to
contact the object during protraction. In the cases (n = 11) in which other whiskers
were between the principal whisker and the object, the other whiskers were moved
rostral to the object prior to each block of trials.
Histology and anatomical analysis.
Procedures were identical to those previously employed (Haidarliu and Ahissar,
2001). Briefly, at the end of each recording session, electrolytic lesions were induced
by passing currents (10 μA, 2 × 4s, unipolar) through the tips of the recording
electrodes. The brains were then removed, fixed, sliced coronally, and stained for CO.
18
Lesions located in the thalamus could be clearly seen (e.g., Figure S1). The
coordinates of lesions in each rat were normalized to the size of the VPM of that rat,
and translated from the coronal plane to an oblique plane (from dorsomedial to
ventrolateral, at 50° to the horizontal plane (Haidarliu and Ahissar, 2001) and placed
on a canonical map of the thalamus on that plane (Figure 1D) (see Figure S1 for a
detailed description of the coordinate transformation process). Sessions in which
recording sites could not be determined were excluded from analysis.
Analysis of whisking and neuronal data.
Of the 97 individual neurons recorded, the recording sites of 14 could not be
accurately localized, five did not respond to our stimulation paradigm, three
responded only during retraction, and eight exhibited nonstationary behavior during
the recording session. These 30 neurons were excluded from analysis, leaving 67
neurons in the dataset to be analyzed. Trajectories of whisker movements were
analyzed offline, using a semi-automatic image processing software (Knutsen et al.,
2005). Whisking onset time was determined from the video records as the time at
which the whiskers started moving. Raster plots and peri-stimulus time histograms
(PSTHs; 1-ms bins, smoothed by convolution with a triangle of area 1 and a base of ±
10 ms) were computed and examined for all trains of each cell. Average response
latencies were computed from PSTHs as the delay from certain events (protraction
onset or contact) to half-peak response. Responses were analyzed during steady-state
periods. We selected cycles 5–10 as those cycles in which virtually all thalamic
neurons exhibited stabilized responses. This selection was based on the following
observations. The intertrial variability (variance/mean) of response spike counts
stabilized, on average, on cycle 1 for POm neurons and cycle 5 for VPM neurons. The
mean cycle-to-cycle difference of four response variables (spike count/cycle, PSTH
amplitude, latency to half-peak, and delay to first spike) stabilized at 0 for all three
nuclei prior to cycle 5, except for spike count/cycle in the VPMdm, which stabilized
on cycle 6.
Results Anatomical Borders
In thalamic slices, the border between VPM and POm is distinct in all planes of
sectioning, due to a high contrast in several anatomical markers, including
19
cytochrome oxidase (CO) (Land et al., 1995). However, the border between VPMdm
and VPMvl is not distinct with standard section planes, i.e., coronal, sagittal, and
horizontal (Land and Simons, 1985b; Pierret et al., 2000). Thus, we explored non-
standard section planes in both young (in which thalamic borders are in general more
clear) (Haidarliu and Ahissar, 2001) and adult rats. We found that a distinct border
between VPMdm and VPMvl is visible in an oblique plane, from dorsomedial to
ventrolateral, at 50° to the horizontal plane (Haidarliu and Ahissar, 2001). In this
plane, the VPMdm–VPMvl border was salient in young rats (Figure 1A and 1B), and
still visible in adult rats (Figure 1C and 1D). We used the anatomical scheme of
Figure 1D as a canonical scheme for our thalamic recordings. Each recording site was
determined by an electrolytic lesion and mapped onto the canonical scheme (Figure
1D, black dots). The method we used for coordinate transformation is described in
Figure S1.
VPL
VPMdm
VPMvl
POmVL
Rt
E DC
B A
1 mm
0.5 mm
R
DMA B
C D
Figure 1. Thalamic Nuclei and Recording Sites
(A) and (C) Oblique sections, dorsomedial to ventrolateral at 50° clockwise to the horizontal plane when the right hemisphere was viewed rostrally (Inset in [B]), through the thalamus of a young (postnatal day 7) (A) and an adult (340 g) (C) rat stained for CO. Depths from bregma were 2.9 mm for the dorsomedial and 3.6 mm for the ventrolateralend of the section shown in (A) and 4.4 mm and 5.85 mm, respectively, for the section shown in (C). Scale bars indicate 0.5 mm (A) and 1.0 mm (C). Arrows: R, rostral; DM, dorsomedial.
VPL
VPMdm
VPMvl
POmVL
Rt
E DC
B A
1 mm
0.5 mm
R
DMA B
C D
VPL
VPMdm
VPMvl
POmVL
Rt
E DC
B A
1 mm
0.5 mm
R
DMA B
C D
Figure 1. Thalamic Nuclei and Recording Sites
(A) and (C) Oblique sections, dorsomedial to ventrolateral at 50° clockwise to the horizontal plane when the right hemisphere was viewed rostrally (Inset in [B]), through the thalamus of a young (postnatal day 7) (A) and an adult (340 g) (C) rat stained for CO. Depths from bregma were 2.9 mm for the dorsomedial and 3.6 mm for the ventrolateralend of the section shown in (A) and 4.4 mm and 5.85 mm, respectively, for the section shown in (C). Scale bars indicate 0.5 mm (A) and 1.0 mm (C). Arrows: R, rostral; DM, dorsomedial.
20
Specificity of Thalamic Responses
We examined the specificity of neuronal responses to active movement and touch
in the three parallel trigeminal pathways by recording from 67 individual neurons
located in their corresponding thalamic stations in urethane-anesthetized rats (Figure
1D): POm (n = 24), VPMvl (n = 13), and VPMdm (n = 30). The facial nerve was
stimulated at 83 Hz for 100 ms to induce protraction (forward movement of all
whiskers), and then left unstimulated for 100 ms to allow passive retraction.
Repetitive whisking movements were thus induced at 5 Hz, which is within the
natural whisking rate, in trains of 2 s and intertrain intervals of 3 s. In the movement
path of the principal whisker of each recorded neuron, a pole of 2-mm diameter was
presented vertically during touch blocks (consisting of 12–24 trains each), at 70%–
90% of the whisker’s length. No object was presented in free-air blocks.
With moving whiskers, object localization would be ambiguous unless the brain
contained an independent signal that described whisker motion. A pure movement
(whisking) signal (W) would report whisking only, i.e., would report whisker
movement in a consistent manner regardless of touch events during the movement. A
pure touch signal (T) would report touch only. We found that a clear dissociation
between W and T signals occurs in the thalamus, in which W and T signals are
conveyed mainly by POm and VPMvl neurons, respectively. Most POm neurons
responded to whisking movements independently of whether the whiskers contacted
an object or not. For example, a neuron recorded from POm responded similarly
(Figure 2A, two top graphs) even though the whisker trajectory during touch (red)
deviated upon contact from that in free air (black) (Figure 2A, bottom graph). In
contrast, most VPMvl neurons responded only when contacting an external object
(see Figure 2B). VPMdm neurons exhibited combined whisking and touch signals,
whose interactions were either additive (W + T) or subtractive (W − T), i.e., in which
touch either added or subtracted spikes from the whisking response (see Figure 2C
and 2D). The two VPMdm examples depict typical W + T and W − T responses. In
both free-air and touch conditions, tonic responses often contained a strong 83-Hz
component (see rhythmic responses in Figure 2C and 2D), locked to the 83-Hz
movement ripple (see whisker angle trajectories in Figure 2), similar to tonic
responses of TG neurons (Szwed et al., 2003). Tonic responses were observed only in
VPMdm, except one case in VPMvl (see response durations in Table S1). A period
21
with reduced response between the phasic and tonic components, such as that
exhibited in Figure 2D during whisking in air, was observed in seven VPMdm
neurons (two W − T and five W + T) during either whisking in air or against an
object.
VPMdmC
Tria
l #
1
72
0
80
160
Spik
es/s
W+T
0 100 200
deg
Time (ms)
B
1
72
POmW
A
Spik
es/s
0
30
60
Tria
l #
0 100 200Time (ms)
deg
VPMdmW-T
1
72
D
Tria
l #
0
40
80
Spik
es/s
2000 100Time (ms)
deg
VPMvlT
0
35
701
72
Tria
l #Sp
ikes
/s
0 100 200Time (ms)
deg
Figure 2. Responses of Individual Neurons during Steady State in the POm, VPMvl, and VPMdm
Examples of responses of individual neurons during steady state are shown for the POm(A), VPMvl (B), and VPMdm (C) and (D) to whisking in air (black) and whisking against an object (red), relative to the time of protraction onset (t = 0). For each cell, the top graph depicts a raster display of single spikes accumulated from three randomly selected cycles from cycles 5–10. Middle graphs depict the PSTH computed for the entire steady-state period (cycles 5–10). Bottom graphs depict the horizontal angle of the principal whisker during a single cycle; ordinate denotes whisker angle, full scale = 50°, and traces were low-pass filtered at 160 Hz. Contact times, during touch trials, areindicated by arrows.
VPMdmC
Tria
l #
1
72
0
80
160
Spik
es/s
W+T
0 100 200
deg
Time (ms)
VPMdmC
Tria
l #
1
72
1
72
0
80
160
0
80
160
Spik
es/s
W+T
0 100 2000 100 200
deg
Time (ms)
B
1
72
POmW
A
Spik
es/s
0
30
60
Tria
l #
0 100 200Time (ms)
deg
1
72
1
72
POmW
A
Spik
es/s
0
30
60
Tria
l #
0 100 2000 100 200Time (ms)
deg
VPMdmW-T
1
72
D
Tria
l #
0
40
80
Spik
es/s
2000 100Time (ms)
deg
VPMdmW-T
1
72
D
Tria
l #
0
40
80
Spik
es/s
2000 100 2000 100Time (ms)
deg
VPMvlT
0
35
701
72
Tria
l #Sp
ikes
/s
0 100 200Time (ms)
deg
VPMvlT
0
35
70
0
35
701
72
1
72
Tria
l #Sp
ikes
/s
0 100 2000 100 200Time (ms)
deg
Figure 2. Responses of Individual Neurons during Steady State in the POm, VPMvl, and VPMdm
Examples of responses of individual neurons during steady state are shown for the POm(A), VPMvl (B), and VPMdm (C) and (D) to whisking in air (black) and whisking against an object (red), relative to the time of protraction onset (t = 0). For each cell, the top graph depicts a raster display of single spikes accumulated from three randomly selected cycles from cycles 5–10. Middle graphs depict the PSTH computed for the entire steady-state period (cycles 5–10). Bottom graphs depict the horizontal angle of the principal whisker during a single cycle; ordinate denotes whisker angle, full scale = 50°, and traces were low-pass filtered at 160 Hz. Contact times, during touch trials, areindicated by arrows.
22
Whisking and touch signals were quantified by measuring the response (spike
count) of cells to whisking in air (SW) and to whisking against an object (ST) during
the first 100 ms of each whisking cycle. These first 100 ms of each cycle contained
only spikes generated during protraction, which comprised most of the spikes
generated by our thalamic neurons in both whisking and touch conditions (see
response durations in Table S1). Thalamic responses in all three nuclei exhibited a
dynamic phase, lasting three to four whisking cycles, during which the response
changed from cycle to cycle, followed by a steady-state phase during which the
response remained stable (see Materials and Methods). The stable steady-state
response, averaged over the last six cycles of each whisking train, was used to classify
the type of thalamic response encountered. The touch component of the response was
estimated as ST − SW, i.e., the response during protraction in touch cycles minus the
response during protraction in free air. Normalized touch responses (touch index [TI]
= [ST − SW]/[ST + SW]) would be 0 for a cell conveying a pure whisking signal (i.e.
response to whisking is the same with and without touch), 1 for a cell conveying a
pure touch signal, and −1 for a cell whose whisking response is completely inhibited
by touch. We classified cells as W if their ST and SW responses did not differ
significantly and their |TI| < 0.2, T if their TI > 0.8, and WT otherwise (see Figure S2
for statistical justification of these thresholds). Distribution of the TIs of individual
thalamic neurons across the thalamus (Figure 3A) revealed a clear anatomical
dissociation of W, T, and WT signals (see Table S1 for details): W signals (TI ~ 0)
are conveyed mostly (94%; 17/18 W cells) via the POm, T signals mostly (75%; 9/12
T cells) via VPMvl, and WT signals (W + T and W − T) mostly (70%; 26/37 WT
cells) via VPMdm. Consistently, the distribution of the TIs in each of the thalamic
nuclei (Figure 3B) shows that POm neurons cluster around TI approximately 0, most
VPMvl neurons cluster around TI approximately 1, and VPMdm neurons distribute
bimodally, with most of the neurons exhibiting 0 < TI < 1 and fewer neurons
exhibiting −1 < TI < 0.
23
Comparison of Latency and Duration of Responses
During touch cycles, latencies (from whisking onset to half-peak response) of W +
T neurons (median = 6.7 ms) were significantly shorter than those of T (median =
17.6 ms) and W (median = 15.6 ms) neurons (p < 0.006, non-parametric Mann-
Whitney test) (Figure 4A). Thus, W + T responses could not result from an
integration of thalamic W and T responses. Latencies of T neurons from the time of
contact were short (median = 6.7 ms), and comparable to the latencies of W + T
neurons from whisking onset (Figure 4B) (p = 0.27, Mann-Whitney test).
Interestingly, latencies of W − T cells (median = 12.4 ms) were longer than those of
W + T neurons (p = 0.05, Mann-Whitney test). Latencies also differed significantly
across nuclei; POm neurons responded with longer latencies than VPMdm and
VPMvl (from contact) (p < 0.001, Mann-Whitney test), and VPMvl responded later
than VPMdm neurons relative to whisking onset (p = 0.04, Mann-Whitney test). The
same relationships were observed when latencies were computed as delays from
stimulus onset to the first spike in a cycle (see Table S1).
In these experiments W and T responses were significantly shorter (p < 0.001,
paired t test) than the duration of protraction, lasting < 40 and 25 ms, respectively (W,
22 ± 7 ms; T, 16 ± 4 ms [excluding one outlier in VPMdm, whose duration was 92
B
TI = (ST - SW)/( ST + SW)
No.
of c
ells
VPMvl
0
10
VPMdm
0
6
0-1 1
POm
0
10
A1
-1
0
TI
Figure 3. Distributions of Normalized Touch Responses in the Thalamus
(A) TI value of each of the recorded neurons is indicated by a color code on its relative location in the canonical thalamic map defined in Figure 1D. (B) Distribution of TI in VPMdm (n = 30), VPMvl (n = 13), and POm (n = 24).
B
TI = (ST - SW)/( ST + SW)
No.
of c
ells
VPMvl
0
10
VPMdm
0
6
0-1 1
POm
0
10
A1
-1
0
TI
B
TI = (ST - SW)/( ST + SW)
No.
of c
ells
VPMvl
0
10
VPMdm
0
6
0-1 1
POm
0
10
TI = (ST - SW)/( ST + SW)
No.
of c
ells
VPMvl
0
10
VPMdm
0
6
0-1 1
POm
0
10
A1
-1
0
TI
1
-1
0
TI
Figure 3. Distributions of Normalized Touch Responses in the Thalamus
(A) TI value of each of the recorded neurons is indicated by a color code on its relative location in the canonical thalamic map defined in Figure 1D. (B) Distribution of TI in VPMdm (n = 30), VPMvl (n = 13), and POm (n = 24).
24
ms]; their durations at half-peak were 13 ± 6 and 7 ± 1 ms, respectively) (Figure 4C).
Thus, W (POm) neurons mainly responded to the initial phase of protraction, when
whisker velocity was highest, similar to most Whisking neurons of the TG (Szwed et
al., 2003). T (VPMvl) neurons mainly reported contact onset, similar to TG Contact
neurons (Szwed et al., 2003). In contrast, response durations of WT (W + T and W −
T) neurons spanned two modes, one brief (< 40 ms; 21 ± 8 ms, 20/37 neurons) and
one long (55–118 ms; 102 ± 16 ms, 17/37 neurons), which together covered the entire
protraction phase (Figure 4C). This bimodal distribution of WT response durations
resembles that of TG Whisking/Touch neurons (Szwed et al., 2003). The distribution
of response durations differed significantly between the three response types (W, T,
WT; p < 0.02, Mann-Whitney test) and between the three nuclei (p < 0.002, Mann-
Whitney test).
Latency from protraction onset or contact (ms)
0
20
0 4 8 12 16 20 24 28 more
No.
of c
ells B
Duration (ms)
No.
of c
ells
0
30
0 20 40 60 80 100 120
C
Latency from protraction onset (ms)
No.
of c
ells
0
15 T
WW+T
W-T
A
0 4 8 12 16 20 24 28 more
Figure 4. Distributions of Thalamic Latencies and Durations According to Response Type
(A) Latencies from protraction onset to half-peak response of all thalamic neurons during touch trials.(B) Latencies from relevant stimulus (contact time for Touch neurons and protraction onset for the rest) to half-peak response of all thalamic neurons during touch trials. (C) Response durations of individual cells during touch trials, measured from the PSTH, as the period during protraction in which the response was above 0.1 of its maximum.
Latency from protraction onset or contact (ms)
0
20
0 4 8 12 16 20 24 28 more
No.
of c
ells B
Duration (ms)
No.
of c
ells
0
30
0 20 40 60 80 100 120
C
Latency from protraction onset (ms)
No.
of c
ells
0
15 T
WW+T
W-T
A
0 4 8 12 16 20 24 28 more
Latency from protraction onset or contact (ms)
0
20
0 4 8 12 16 20 24 28 more
No.
of c
ells BB
Duration (ms)
No.
of c
ells
0
30
0 20 40 60 80 100 120
CC
Latency from protraction onset (ms)
No.
of c
ells
0
15 T
WW+T
W-T
A T
WW+T
W-T
TT
WW+TW+T
W-T
A
0 4 8 12 16 20 24 28 more
Figure 4. Distributions of Thalamic Latencies and Durations According to Response Type
(A) Latencies from protraction onset to half-peak response of all thalamic neurons during touch trials.(B) Latencies from relevant stimulus (contact time for Touch neurons and protraction onset for the rest) to half-peak response of all thalamic neurons during touch trials. (C) Response durations of individual cells during touch trials, measured from the PSTH, as the period during protraction in which the response was above 0.1 of its maximum.
25
Discussion
We showed here that the major active-touch signal conveyed in each of the three
afferent pathways of the whisker system is different: whisking in the paralemniscal
(via POm), contact in the extralemniscal (via VPMvl), and combined whisking–touch
in the lemniscal (via VPMdm) pathway. The three afferent pathways did not respond
synchronously. In each whisking cycle, VPMdm neurons, conveying the combined
signal, fired first whereas POm and VPMvl neurons, conveying isolated whisking and
touch signals, fired later. VPMdm also contained tonic responses that were absent in
the other nuclei. All these observations, together with the known anatomy and
physiology of the system, suggest that VPMdm responses did not result from a
combination of signals transmitted by the POm and VPMvl. Moreover, the orthogonal
response types of POm (W) and VPMvl (T) indicate that each of the three thalamic
nuclei conveys a signal that could not result from a combination of signals transmitted
by the other two nuclei. This, and the fact that similar response types (W, T, and WT)
are exhibited by different classes of TG neurons, strongly suggest that the three
thalamic nuclei are driven primarily by their afferent pathways. (W − T signals,
whose latencies do allow intra-thalamic inhibition and/or cortical feedback (Krupa et
al., 2004) as primary drivers, might be an exception). Thus, our results suggest
parallel afferent processing via the thalamus (Figure 5).
W
WT
VPMvl
object
T
extralemnisca
l
VPMdm
POm
whisking
para
Figure 5. Proposed Scheme of Afferent Conduction of Active-Touch Signals
Whisking signals (W) are conveyed by the paralemniscal pathway (para) via the POm and are proposed to involve whisking control. Touch signals (T) are conveyed by the extralemniscal pathway (extra) via VPMvl, and are proposed to involve processing of object location (‘where’). Combined whisking–touch signals (WT) are conveyed by the lemniscalpathway via VPMdm, and are proposed to involve processing of object identity (‘what’).
W
WT
VPMvl
object
T
extralemnisca
l
VPMdm
POm
whisking
para
W
WT
VPMvl
object
T
extralemnisca
l
VPMdm
POm
whisking
para
Figure 5. Proposed Scheme of Afferent Conduction of Active-Touch Signals
Whisking signals (W) are conveyed by the paralemniscal pathway (para) via the POm and are proposed to involve whisking control. Touch signals (T) are conveyed by the extralemniscal pathway (extra) via VPMvl, and are proposed to involve processing of object location (‘where’). Combined whisking–touch signals (WT) are conveyed by the lemniscalpathway via VPMdm, and are proposed to involve processing of object identity (‘what’).
26
Previously, we showed that POm neurons represent the temporal frequency of
passive whisker movements by latency and spike count (Ahissar et al., 2000), and
suggested that the POm is involved in temporal decoding of signals that encode
whisker movement and of signals that encode the horizontal coordinate of object
location (Ahissar and Arieli, 2001; Ahissar and Zacksenhouse, 2001). Our current
data show that the POm does not convey contact information, and thus cannot resolve
object location by its own. Taken together, our data now suggest that the POm is
involved in temporal processing related to sensory-motor control of whisker
movement, and that the POm and VPMvl together are involved in temporal
processing of object location. In such processing, the POm would convey the
reference signal and VPMvl the contact signal (Kleinfeld et al., 1999).
Why would sensory information flow in parallel pathways in this, or in other
systems (Perl, 1963; Diamond, 1983; Diamond et al., 1992a; Casagrande, 1994; Kim
and Ebner, 1999; Ahissar et al., 2000; Diamond, 2000; He and Hu, 2002; Jones,
2003)? Based on the anatomical and physiological data available, Bishop (Bishop,
1959) suggested that parallel sensory pathways evolved in successive steps, each
adding a larger fiber pathway, and incorporating successively higher brain areas to
implement a novel function. Thus, Bishop suggested that the first spinal
somatosensory pathway to evolve was the spinothalamic, followed by the
neospinothalamic, and then the dorsal column–lemniscal. An order that is analogous
to the paralemniscal, extralemniscal, and then lemniscal in the trigeminal system. The
functional segregation reported here between these pathways, and the evidence
indicating that these three pathways close the sensory-motor loop at different levels of
brain hierarchy, raise the following sensory-motor hypothesis: The paralemniscal
system is involved in a low order motor-sensory-motor loop that controls whisking
velocity and frequency in a servo-like manner (Wiener, 1949), the extralemniscal
system adds a higher level of control based on contact information and object
location, and the lemniscal system adds the highest level of control so far, which is
based on information related to object identity. This proposed functional segregation
(Figure 5) does not imply functional isolation; these parallel loops are expected to
interact such that a higher loop uses, and builds upon, the processing performed by a
lower loop. One example for such interaction is object localization, in which contact
timing (extralemniscal) must interact with whisking information (paralemniscal) to
extract object location. Analysis of object identity requires interaction of detailed
27
spatial information with information about whisker movement and contact (Moore,
2004; Arabzadeh et al., 2005). The high-resolution directional-selective spatial
information (Minnery et al., 2003; Timofeeva et al., 2003) together with whisking
information (WT signals) conveyed by the VPMdm meet this requirement. Object-
identity analysis also involves comparisons with memorized patterns; hence, it
requires significant cortical involvement (Guic-Robles et al., 1992; Hawkins and
Blakeslee, 2004), such as that exhibited by the lemniscal system. Thus, the
paralemniscal, extralemniscal, and lemniscal parallel loops may have evolved
sequentially, as suggested for parallel sensory pathways (Bishop, 1959), by adding
contact detection to movement control, and identity analysis to contact detection.
Our experimental paradigm utilized rats under general anesthesia, which affects
response amplitude, latency, duration, and adaptation in the thalamus and cortex
(Simons et al., 1992; Fanselow and Nicolelis, 1999; Friedberg et al., 1999; Castro-
Alamancos, 2004). However, these effects are quantitative in nature and are expected
to be similar for all thalamic neurons, and thus cannot account for the prominent
differences in response types we report here. The state of thalamic and cortical
neurons during the steady-state response phase, the phase used herein for response
classification in anesthetized rats, is considered to be analogous to the state of
thalamic and cortical neurons during exploratory whisking in awake rats (Castro-
Alamancos, 2002a; Nicolelis and Fanselow, 2002; Castro-Alamancos, 2004).
Consistently, during the steady state, thalamic neurons are hypothesized to function in
their gating, signal-processing mode (Sherman and Guillery, 1996). Nevertheless,
under anesthesia, the intensity and nature of top-down effects, such as those affecting
the thalamus directly, or indirectly (Bokor et al., 2005; Lavallee et al., 2005), are
probably different; the efferent signals that control whisking are lacking; and the
sensory-motor loops that control active touch (Kleinfeld et al., 1999) are practically
opened. Moreover, behaving rats continuously control their whisking according to
context and in reaction to contacts. Thus, although the basic segregation of response
types observed here in anesthetized rats is expected to occur in awake ones, the exact
behavior of thalamic neurons during active touch should be further studied in awake
behaving rats.
28
Supporting Information
Table S1. Response types, magnitudes, latencies and durations in each thalamic nucleus.
Cell typea
Thalamic region W-T W W+T T
Number of cells
POm 2 17 5
VPMdm 8 1 18 3
VPMvl 1 3 9
Responses (spikes per cycle, mean ± SD)
SWb 0.49 ±0.40 0.36 ±0.10 0.36 ±0.21
ST 0.32 ±0.27 0.34 ±0.10 0.58 ±0.28
POm
p < 0.026 > 0.1 < 0.029
SW 1.76 ±1.15 0.28 0.44 ±0.46 0.02 ±0.00
ST 0.63 ±0.74 0.29 0.83 ±0.56 0.29 ±0.04
VPMdm
p < 0.001 0.890 < 0.022c < 0.001
SW 3.10 0.21 ±0.04 0.01±0.01
ST 1.13 0.40 ±0.13 0.36 ±0.20
VPMvl
p 0.000 < 0.001 < 0.001
29
Cell type
Thalamic region W-T W W+T T
Latencies to half-peak(ms)d
POm 17.5-20.3 (18.9)e 12.8-28.9 (15.8) 14.3-28.5 (17.7)
VPMdm 10.0-29.4 (12.2) 7.4 4.5-54.9 (6.6) 5.9-7.3 (7.0)
VPMvl 9.7 5.7-6.7 (6.5) 4.1-14.7 (6.3)
Delay to 1st spikes(ms)d
POm 22-27 (25)e 15-41 (22) 18-42 (30)
VPMdm 11-30 (15) 8 5-55 (7) 6-9 (8)
VPMvl 10 6-7 (7) 5-16 (7)
Duration(ms)f
POm 19-30 (25)e 12-38 (20) 15-38 (27)
VPMdm 13-115(103) 27 15-118 (63) 15-92 (20)
VPMvl 82 12-14 (12) 11-23 (16)
aCells were classified as T if their normalized touch responses (TI) during protraction > 0.8, W if
their ST and SW responses did not differ significantly and |TI|<0.2 (see Figure S1), W-T if ST < SW
and W+T if ST > SW. bSW, spikes per free-air whisk; ST, spikes per touch whisk; p, ranges of p-values, computed for every
cell, which indicates the probability that for that cell single-whisk responses in free-air and touch
whisks belong to the same distribution [two-tailed t-test, number of whisks for each cell was 144
(n=54 cells), 216 (n=13), or 288 (n=3)]. cOne outlier had p = 0.136 (see Figure S1) dDelays and latencies are from contact for T neurons and from protraction onset for all other
neurons. eRange (median) fDurations were measured from the PSTHs, as the period in which the response was above 0.1 of its
maximum during protraction.
30
s1
s2
s3
d3d2 d1
Si
Di
Figure S1. Transformation of recording coordinates from coronal to oblique-horizontal plane.
(A1-3) Coronal sections through the thalamus containing lesions (arrows) in the ventrolateralVPM (A1), dorsomedial VPM (A2) and POm (A3). Slices where counted starting from the rostral end of the VPM. "Slice x/y" indicates that the center of the lesion was found in slice no. x, out of total y slices that spanned the rostrocaudal length of the VPM in that rat. (B) Normalization of recording site in the coronal plane. Each lesion is characterized by a ratio of 2 distances measured along the 50˚ dorsomedial-to-ventrolateral slope; Si, the distance of the lesion from the border between the POm and VPM, and Di, the distance between the POm/VPM border to the VPM/VPL border. (C) Re-mapping of recording sites on the oblique-horizontal plane. The rostro-caudal coordinate is determined by the normalized sequential number of the coronal slice; the normal rostro-caudal length of the VPM is set to 20 coronal slices. Thus, for example, slice 12 out of 18 is transformed to 20*12/18 = 13.3. The dm-vl coordinate of a site i (i=1,2,3) is si = di * Si/Di, where di is the POm/VPM to VPM/VPL distance in the oblique-horizontal plane at the rostrocaudal coordinate of site i, and Si/Di is the ratio obtained for that site in the coronal section (B). s1-3 and d1-3 correspond to the lesions depicted in A1-3. Scale bars are 1 mm in all panels.
s1
s2
s3
d3d2 d1
Si
Di
s1
s2
s3
d3d2 d1
Si
Di
Figure S1. Transformation of recording coordinates from coronal to oblique-horizontal plane.
(A1-3) Coronal sections through the thalamus containing lesions (arrows) in the ventrolateralVPM (A1), dorsomedial VPM (A2) and POm (A3). Slices where counted starting from the rostral end of the VPM. "Slice x/y" indicates that the center of the lesion was found in slice no. x, out of total y slices that spanned the rostrocaudal length of the VPM in that rat. (B) Normalization of recording site in the coronal plane. Each lesion is characterized by a ratio of 2 distances measured along the 50˚ dorsomedial-to-ventrolateral slope; Si, the distance of the lesion from the border between the POm and VPM, and Di, the distance between the POm/VPM border to the VPM/VPL border. (C) Re-mapping of recording sites on the oblique-horizontal plane. The rostro-caudal coordinate is determined by the normalized sequential number of the coronal slice; the normal rostro-caudal length of the VPM is set to 20 coronal slices. Thus, for example, slice 12 out of 18 is transformed to 20*12/18 = 13.3. The dm-vl coordinate of a site i (i=1,2,3) is si = di * Si/Di, where di is the POm/VPM to VPM/VPL distance in the oblique-horizontal plane at the rostrocaudal coordinate of site i, and Si/Di is the ratio obtained for that site in the coronal section (B). s1-3 and d1-3 correspond to the lesions depicted in A1-3. Scale bars are 1 mm in all panels.
31
Figure S2. Classification of response types based on steady-state responses: statistical
significance and criteria.
(A) For each cell, the probability (one-tailed t-test, across all steady-state cycles) that it did
not respond to whisking in air (SW=0) is depicted as a function of its TI [Touch Index, i.e.,
the normalized touch responses during protraction]. (B) For each cell, the probability (two-
tailed t-test) that its responses to whisking in air (SW) and whisking against an object (ST)
were identical is depicted, as a function of its TI. (C) Distribution of TI in the trigeminal
thalamus (n=67). Green, POm cells; brown, VPMdm cells; orange, VPMvl cells.
Based on these data, cells with TI>0.8 were classified as T (Touch) cells (dotted box in A).
Cells with |TI|<0.2 and p(SW=ST)> 0.05 were classified as W (Whisking) cells (dotted box in
B).
0
5
10
-1 10.8
No.
of c
ells
C
TI = (ST - SW)/ (ST + SW)
0
0.6
1.2P
(SW
= 0)
A T cells
0
0.6
1.2
P (S
W=
ST)
B W cells
0.2-0.20
5
10
-1 10.8
No.
of c
ells
C
TI = (ST - SW)/ (ST + SW)
0
0.6
1.2P
(SW
= 0)
A T cells
0
0.6
1.2
P (S
W=
ST)
B W cells
0.2-0.2
32
CHAPTER 2: Thalamic coding of active Vibrissal touch
Summary
Rats employ actively rhythmic whisker movements to sample information in their
sensory environment. We previously found that various functional types of first-order
trigeminal neurons encode major events occurring during active vibrissal sensing, and
signals related to active touch are conveyed in parallel via three thalamic nuclei. How
thalamic neuron encodes external information, such as object location and whisking
frequency, is not yet known. To address these issues, we recorded from individual
neurons from the three major thalamic nuclei of the whisker system: POm, VPMvl
and VPMdm – while inducing artificial whisking in anesthetized rats at 5Hz and
modulated frequencies. We found that Touch neurons fired upon contact with object,
encoding the horizontal coordinate of object location by first-spike timing. In contrast,
Whisking-Touch neurons did not convey any information about object location by
their first-spike timing; some Whisking-Touch neurons encoded horizontal object
location by inter-spike intervals and by spike counts. We also examined how thalamic
neurons encode whisking frequency. When whisking frequency was modulated, POm
neurons exhibited both spike count and delay to first spike (and onset latency)
modulations: spike count decreased and delay to first spike increased with frequency
increment during both free air whisking and touch conditions. In contrast, VPMdm
neurons represented whisking frequency by spike count alone during whisking
condition, and this representation was suppressed during touch condition. VPMvl
neurons did not present any modulations to whisker frequencies in both conditions.
These findings are consistent with our previous study which suggested that Touch
cells might be involved in temporal coding of object location, Whisking-Touch cells
convey complex information possibly related to coding of object identity, and POm
cells might be involved in the control of whisking frequency.
33
Backgound
Rodent whisking system is a good model for studying active touch (Ahissar and
Arieli, 2001; Szwed et al., 2003). During tactile exploration, rats actively move their
whiskers across objects or surfaces in repeated rhythmic sweeps (5-12Hz) (Carvell
and Simons, 1990; Fanselow and Nicolelis, 1999; Kleinfeld et al., 1999)..
Somatosensory information from the vibrissae of the rat is first encoded by the
trigeminal ganglion neurons, and then carried to the second-order neurons within the
brainstem trigeminal complex; it reaches to the neocortex via the thalamus in three
parallel pathways: the lemniscal pathway ascends via the dorso-medial sector of the
VPM (VPMdm), the paralemniscal ascends via the POm, and extralemniscal pathway
ascends via the ventro-lateral sector of the VPM (VPMvl) (Pierret et al., 2000; Yu et
al., 2006).
Our knowledge of sensory processing mechanisms is derived primarily from
studies of passive (mechanical) movements of different directions, amplitudes and
velocities on individual whiskers. In nature, when exploring, a rat actively protracts
and retracts all the whiskers in synchrony. Whisking is an active behavior. In an
attempt to approximate more closely this natural movement, Zucker and Welker
firstly introduced an “artificial whisking” in anesthetized rats by stimulating the facial
motor never supplying the whiskers electrically, and observed neuron responses in the
trigeminal ganglion (Zucker and Welker, 1969). Later on, Brown and Waite recorded
thalamic neuron responses in VPM using a similar type of stimulation (Brown and
Waite, 1974). They demonstrated that when an object was introduced into the
whisking field, additional bursts or responsive neurons would be generated during
whisker protraction period and the temporal reliability of the generated spike patterns
was high. However, they both did not conduct quantitative measurement of responses
to active touch and did not study the principle of active encoding. Recently, our group
developed this method and successfully investigated the encoding principle of active
touch by rat trigeminal ganglion neurons (Szwed et al., 2003). Our lab results
suggested that individual first-order trigeminal neurons encoded four specific events:
whisking, contact with object, pressure against object, and detachment from object.
Whisking-responsive neurons fired at specific deflection angles, reporting the actual
whiskers' position with high precision. Touch-responsive neurons encoded the
horizontal coordinate of objects' position by spike timing and the radial dimension
34
primarily by spike count, i.e., total number of spikes emitted by a cell during a
whisking cycle (Szwed et al., 2003; Szwed et al., 2006).
Thalamus serves as a gate that regulates the flow of sensory input to the neocortex.
While inducing artificial whisking in anesthetized rats, our recent study showed that
different sensory signals related to active touch are conveyed separately via the
thalamus by three parallel afferent pathways. The paralemniscal pathway conveys
sensor motion (whisking) signals (via POm), the extralemniscal conveys contact
(touch) signals (via VPMvl), and the lemniscal pathway conveys combined whisking-
touch signals (via VPMdm) (Yu et al., 2006). Here, we used the same set of data from
thalamus to examine how these thalamic neurons encode the horizontal location of the
objects and what the encoding principle of whisker frequency is during active
whisking.
Methods The methods are the same as those described in detail in the chapter 1.
Specially, the responses of thalamic neurons in the three nuclei were collected
during contact with simple objects presented at there different locations along the
horizontal coordinate. Two types of stimuli were applied in blocks, each containing
several trains (trials) of identical parameters. First, a constant-frequency electrical
whisking was induced at 5Hz in blocks of 12 consecutive trains of 2 s (18 or 24 trains
in some cases); with inter-train intervals of 3 s. Whisking movements were obtained
with 100ms protraction followed by 100 ms retraction. Second, frequency-modulated
(FM) stimuli were applied as follows: firstly, the whisker stimulation frequency was
kept constant at 5Hz for 1 s. Then, the stimulus frequency was modulated for an
additional 2.5 s (40% modulation depth, 0.4Hz, initial phase 90°). The constant pulse
width (80ms) was used throughout this 2.5s trains. The entire process was repeated 12
x 2 times for each simultaneous recording.
Results Artificial whisking paradigm
We used the similar experimental and stimulus conditions for the examination of
the coding of object location as in our previous recording from trigeminal ganglion
(NV) (Szwed et al., 2003). Briefly, for each neuron in both VPM and POm
35
populations, trains of 2 s electrical whisking at 5 Hz or modulated frequencies (FM)
were applied by stimulating the facial motor nerve, and whisker movement
trajectories were tracked with a fast digital video camera (1000 frames/s) (Figure 1A).
The movement profiles of the whiskers at 5Hz are depicted in Figure 1B. During the
first three whisking cycles, the shape and velocity of whisker movement are constant,
with a slight increase (<10%) in both resting and protracted positions. Within a
whisking cycle, a small stimulus-locked component (83Hz) was superimposed on the
main protraction trajectory. When an object was placed in the path of whisking,
whiskers touched it, pressed, bent, and then bent back and retracted (Figure 1A).
Times and angles of whisker object contact were measured from recorded video
frames. For the whisking frequency test, a constant frequency at 5Hz was applied first
for 1s, and then an additional 2.5s modulated frequency was induced (see method).
The movement profiles of the whiskers at FM are described in Figure 1C.
Here, the neuronal responses of POm (n=24), VPMvl (n=13), and VPMdm (n=30)
were recorded and analyzed for the coding of object location. Among these cells, 14
POm, 11 VPMvl and 13 VPMdm cells were also examined for the coding of whisking
frequency.
Figure 1. Experimental paradigm
(A) Five video frames taken at 1000fps during a single whisking cycle depict a whisker whisking onset, touching the object, bending, detaching the object and whisking offset. Red line denotes a principal whisker. Numbers indicate timing of elapsing from whisking onset. (B) Whisker trajectory of an entire 5 Hz artificial whisking trial. (C) Whisker trajectory of a whisking trial at modulated frequency.
A
B
Time (ms)
0 20001000
Ang
le ( °)
C
Ang
le ( °)
30000 1000 2000Time (ms)
whisking onset touch bent detach whisking offset
1 ms 66 ms 149 ms 200 ms30 ms
Figure 1. Experimental paradigm
(A) Five video frames taken at 1000fps during a single whisking cycle depict a whisker whisking onset, touching the object, bending, detaching the object and whisking offset. Red line denotes a principal whisker. Numbers indicate timing of elapsing from whisking onset. (B) Whisker trajectory of an entire 5 Hz artificial whisking trial. (C) Whisker trajectory of a whisking trial at modulated frequency.
A
B
Time (ms)
0 20001000
Ang
le ( °)
C
Ang
le ( °)
30000 1000 2000Time (ms)
whisking onset touch bent detach whisking offset
1 ms 66 ms 149 ms 200 ms30 ms1 ms 66 ms 149 ms 200 ms200 ms30 ms
36
Representations of object location during artificial whisking
Our previous study suggested that NV touch neurons encoded the horizontal object
positions by spike timing (i.e. the delay to first spike), rather than spike count and
average inter-spike interval (Szwed et al., 2003). Our findings in the thalamus
suggested that POm mainly contains whisking responsive cells; VPM mainly contains
touch responsive cells: touch cells in VPMvl and whisking-touch cells in VPMdm.
Therefore, we investigated how object positions are encoded by different type of
touch-responsive cells in the VPM. For each cell, an object was introduced at three
different horizontal positions during active whisking. The neuronal variables of
individual cells: delay from whisking onset to the first spike, spike count per cycle
and average inter-spike interval per cycle was examined.
Touch cells Figure 2 depicts two examples of touch cells. These two cells did not
respond to free air whisking, but did respond with short bursts upon contact with the
object. In cell 665 (Fig.2A), both the delay to the first spike (R2=0.95, Fig.2B) and
spike count per cycle (R2 = 0.59, Fig.2C) had correlation with the time of whisker
object touching: delay to the first spike increased and spike count decreased as a
function of contact timing. In contrast, inter-spike interval per cycle did not provide
any information about object positions (R2=0.006, Fig.2D). However, in cell 707
(Fig.2C), object location was represented by the delay to first spike (R2 = 0.83, Fig.
2F), but not by its spike count per cycle and inter-spike interval (R2 = 0.30 and R2 =
0.30, Fig.2G and Fig.2H).
To investigate the encoding scheme of the Touch cell population, the amount of
information conveyed by these three variables was quantified by estimating the
coefficients of determination (R2) for individual cells. Since most of Touch cells
usually had no more than 1 spike per trial responding to touch, the encoding by
average inter-spike interval could be considered only for 4 touch cells here. Figure 3
shows that most of Touch cells (83%, 10/12) could encode object location mainly by
their spike timing: delay to the first spike increase as object was located anteriorly.
The delay to 1st spike contained significantly much more information than spike
count and inter-spike interval (p<0.017, non-parametric Mann-whitney test). To some
extend, object location could also be encoded by spike count in some Touch cells. The
spike count provided more information about object location than inter-spike interval
(p=0.01). No information could be conveyed by inter-spike interval in Touch cells.
37
Figure 2. Encoding of horizontal object position in VPMvl Touch cells– two representative cells(A and C) Raster plots of of VPMvl Touch cell 665 and cell 707 responding to free air whisking (top) and three object positions (from sencond to bottom, posterior to anterior) during 3 randomly selected cycles from cycles 5-10 (72 trials), and steady-state PSTHs of the responses (free air whisking, black; object positions: from posterior to anterior, magenta, green and red) (B-D and F-H) Linear regressions of delay to first spike per cycle against whisker-object contact time, and spike count per cycle and interspike interval per cycle function as whisker-object contact time
A
0 100 200
35
70
0
R2 = 0. 83
0
20
40
5 10 15 20Contact time (ms)D
elay
to 1
stsp
ike (
ms)
R2 = 0.30
0
0.4
0.8
5 10 15 20Contact time (ms)
Spik
es/c
ycle
Time (ms)
Spik
es/s
Tria
l #
E F
R2 = 0. 04
0
40
80
5 10 15 20Contact time (ms)
ISI (
ms)
G
H
Contact time (ms)
R2 = 0. 96
0
25
50
5 10 15 20 25 30Del
ay to
1sts
pike
(ms)
0 100 2000
40
80
Time (ms)
Tria
l #Sp
ikes
/sB
R2 = 0. 01
0
5
10
5 10 15 20 25 30Contact time (ms)
ISI (
ms)
C
D
R2 = 0.59
0
0.4
0.8
5 10 15 20 25 30Contact time (ms)
Spik
es/c
ycle
Figure 2. Encoding of horizontal object position in VPMvl Touch cells– two representative cells(A and C) Raster plots of of VPMvl Touch cell 665 and cell 707 responding to free air whisking (top) and three object positions (from sencond to bottom, posterior to anterior) during 3 randomly selected cycles from cycles 5-10 (72 trials), and steady-state PSTHs of the responses (free air whisking, black; object positions: from posterior to anterior, magenta, green and red) (B-D and F-H) Linear regressions of delay to first spike per cycle against whisker-object contact time, and spike count per cycle and interspike interval per cycle function as whisker-object contact time
A
0 100 200
35
70
0
R2 = 0. 83
0
20
40
5 10 15 20Contact time (ms)D
elay
to 1
stsp
ike (
ms)
R2 = 0.30
0
0.4
0.8
5 10 15 20Contact time (ms)
Spik
es/c
ycle
Time (ms)
Spik
es/s
Tria
l #
E F
R2 = 0. 04
0
40
80
5 10 15 20Contact time (ms)
ISI (
ms)
G
H
Contact time (ms)
R2 = 0. 96
0
25
50
5 10 15 20 25 30Del
ay to
1sts
pike
(ms)
0 100 2000
40
80
Time (ms)
Tria
l #Sp
ikes
/sB
R2 = 0. 01
0
5
10
5 10 15 20 25 30Contact time (ms)
ISI (
ms)
C
D
R2 = 0.59
0
0.4
0.8
5 10 15 20 25 30Contact time (ms)
Spik
es/c
ycle
A
0 100 200
35
70
0
R2 = 0. 83
0
20
40
5 10 15 20Contact time (ms)
R2 = 0. 83
0
20
40
5 10 15 20Contact time (ms)D
elay
to 1
stsp
ike (
ms)
Del
ay to
1st
spik
e (m
s)
R2 = 0.30
0
0.4
0.8
5 10 15 20Contact time (ms)
Spik
es/c
ycle
Time (ms)
Spik
es/s
Tria
l #
E F
R2 = 0. 04
0
40
80
5 10 15 20Contact time (ms)
R2 = 0. 04
0
40
80
5 10 15 20Contact time (ms)
ISI (
ms)
G
H
Contact time (ms)
R2 = 0. 96
0
25
50
5 10 15 20 25 30Del
ay to
1sts
pike
(ms)
Del
ay to
1sts
pike
(ms)
0 100 2000
40
80
Time (ms)
Tria
l #Sp
ikes
/sB
R2 = 0. 01
0
5
10
5 10 15 20 25 30Contact time (ms)
ISI (
ms)
C
D
R2 = 0.59
0
0.4
0.8
5 10 15 20 25 30Contact time (ms)
Spik
es/c
ycle
38
Whisking-Touch cells Two types of Whisking-Touch cells exist in the VPM: one
is W+T cells, in which touch added spikes from the whisking response. The other is
W-T cells, in which touch subtracted spikes from the whisking response. The
encoding scheme of object location was investigated by examining these two types of
Whisking-Touch cells, respectively. Linear regression estimate of the delay to first
spike, spike count per cycle and inter-spike interval against whisker object contact
time was calculated.
Figure 4 shows neuronal representations and encoding scheme of one W+T cell. In
this cell, both spike count per cycle and inter-spike interval per cycle displayed
Touch cells (n=12)
Delay
0 0.5 10
2
4
6
1.00.50.0
Inter-spike interval
spikes/cycles
delay to 1st spike
R2
A
B
0 0.5 10
2
4ISI
R2
Sp
0 0.5 10
2
4
-3 -2 -1 0 1 2 30
5
10
-0.2 -0.1 0 0.1 0.20
4
8
-3 -2 -1 0 1 2 30
2.5
5
Slope
Figure 3. Encoding of horizontal object position by VPM Touch cells– population (n=12)
(A) Box plot of coefficients of determination (R2) for individual Touch cell’s linear regressions of delay to first spike, spike count per cycle and inter-spike interval against whisker-object contact time. Boxes indicate the first (25%) to third (75%) quartile values. Horizontal line indicates range and line inside box denotes median. Outlier (>3 interquartile values from median) is plotted with asterisks (∗).(B) Distribution of R2 (top) and slope (bottom) for delay to first spike (left), spike count (middle) and inter-spike interval (right, n=4) of Touch cells responding to three object positions.
Touch cells (n=12)
Delay
0 0.5 10
2
4
6
1.00.50.0
Inter-spike interval
spikes/cycles
delay to 1st spike
R2
A
B
0 0.5 10
2
4ISI
R2
Sp
0 0.5 10
2
4
-3 -2 -1 0 1 2 30
5
10
-0.2 -0.1 0 0.1 0.20
4
8
-3 -2 -1 0 1 2 30
2.5
5
Slope
Touch cells (n=12)
Delay
0 0.5 10
2
4
6
0 0.5 10
2
4
6
1.00.50.0
Inter-spike interval
spikes/cycles
delay to 1st spike
1.00.50.0 1.00.50.0
Inter-spike interval
spikes/cycles
delay to 1st spike
R2
A
B
0 0.5 10
2
4
0 0.5 10
2
4ISI
R2
Sp
0 0.5 10
2
4
R2
Sp
0 0.5 10
2
4
0 0.5 10
2
4
-3 -2 -1 0 1 2 30
5
10
-0.2 -0.1 0 0.1 0.20
4
8
-3 -2 -1 0 1 2 30
2.5
5
Slope-3 -2 -1 0 1 2 30
5
10
-3 -2 -1 0 1 2 30
5
10
-0.2 -0.1 0 0.1 0.20
4
8
-0.2 -0.1 0 0.1 0.20
4
8
-0.2 -0.1 0 0.1 0.20
4
8
-3 -2 -1 0 1 2 30
2.5
5
-3 -2 -1 0 1 2 30
2.5
5
Slope
Figure 3. Encoding of horizontal object position by VPM Touch cells– population (n=12)
(A) Box plot of coefficients of determination (R2) for individual Touch cell’s linear regressions of delay to first spike, spike count per cycle and inter-spike interval against whisker-object contact time. Boxes indicate the first (25%) to third (75%) quartile values. Horizontal line indicates range and line inside box denotes median. Outlier (>3 interquartile values from median) is plotted with asterisks (∗).(B) Distribution of R2 (top) and slope (bottom) for delay to first spike (left), spike count (middle) and inter-spike interval (right, n=4) of Touch cells responding to three object positions.
39
significant correlation with whisker-object contact time (R2 = 0.66 and R2 = 0.76,
Figure 4C and 4D): Spike count decreased and inter-spike interval increased as the
object moved anteriorly, which suggest that these two variables did convey
information about object location. In contrast, the delay to first spike did not exhibit
any correlation with the time of whisker object touching (R2 = 0.02). Although object
position was presented by two neuronal variables in this cell, the distribution of the
coefficients of determination showed that all of these three codes contained much less
information about object locations across W+T cell population (p>0.09, Mann-
whitney test) (Figure 5).
Figure 4. Encoding of horizontal object position by a W+T cell
(A) Raster plots of VPMdm whisking-touch cell 693 responding to free air whisking (top) and three object positions (from second to bottom, posterior to anterior) during 3 randomly selected cycles from cycles 5-10 (72 trials), and steady-state PSTHs of the responses (free air whisking, black; object positions: from posterior to anterior, magenta, green and red)(B-D) Linear regressions of delay to first spike per cycle, spike count per cycle and interspike interval per cycle against the time of whisker-object contact.
Time (ms)0 100 200
0
80
160
Tria
l #Sp
ikes
/s
A B
R2 = 0. 02
0
10
20
30
4 6 8 10 12Contact time (ms)
Del
ay to
1sts
pike
(ms)
R2 = 0. 66
0
1
2
4 6 8 10 12Contact time (ms)
Spik
es/c
ycle
Contact time (ms)
R2 = 0. 76
0
10
20
4 6 8 10 12IS
I (m
s)C
D
Figure 4. Encoding of horizontal object position by a W+T cell
(A) Raster plots of VPMdm whisking-touch cell 693 responding to free air whisking (top) and three object positions (from second to bottom, posterior to anterior) during 3 randomly selected cycles from cycles 5-10 (72 trials), and steady-state PSTHs of the responses (free air whisking, black; object positions: from posterior to anterior, magenta, green and red)(B-D) Linear regressions of delay to first spike per cycle, spike count per cycle and interspike interval per cycle against the time of whisker-object contact.
Time (ms)0 100 200
0
80
160
0 100 2000
80
160
Tria
l #Sp
ikes
/s
A B
R2 = 0. 02R2 = 0. 02
0
10
20
30
4 6 8 10 12Contact time (ms)
Del
ay to
1sts
pike
(ms)
Del
ay to
1sts
pike
(ms)
R2 = 0. 66
0
1
2
4 6 8 10 12Contact time (ms)
Spik
es/c
ycle R2 = 0. 66R2 = 0. 66
0
1
2
4 6 8 10 12Contact time (ms)
Spik
es/c
ycle
Contact time (ms)
R2 = 0. 76R2 = 0. 76
0
10
20
4 6 8 10 12IS
I (m
s)C
D
40
The encoding of horizontal object position was also examined among W-T cells
using these three neuronal variables. Figure 6 shows one example of W-T cells. This
cell exhibited robust tonic responses during free air whisking, while the response was
strongly inhibited by object touch. Spike count per cycle somewhat correlated with
the timing of object touch, in which spike count increased as a function of the time of
object touch. However, this correlation does not reach statistical significance (R2 =
0.41, Figure 6C). And other two variables did not demonstrate any relationship with
contact timing (R2 = 0.01 and R2 = 0.06, Figure 6B and 6D).
The coefficients of determination of entire W-T cell population show that all the
W-T cells exclude one increased their spike counts as a function of object contact
W+T cells (n=20)
Inter-spike interval
spikes/cycles
delay to 1st spike
1.00.50.0
A
R2
0 0.5 10
9
18
0 0.5 10
4
8
0 0.5 10
5
10Delay Sp ISI
R2
B
-2 -1 0 1 2 3 40
3
6
-0.2 -0.1 0 0.1 0.20
6
12
-3 -2 -1 0 1 2 30
2.5
5
Slope
Figure 5. Encoding of horizontal object position by VPM W+T cells– population (n=20)
(A) Box plot of coefficients of determination (R2) for individual W+T cell’s linear regressions of delay to first spike, spike count per cycle and inter-spike interval against whisker-object contact time. Boxes indicate the first (25%) to third (75%) quartile values. Horizontal line indicates range and line inside box denotes median. Outlier (>3 inter-quartile values from median) is plotted with asterisks (∗).(B) Distribution of R2 (top) and slope (bottom) for delay to first spike (left), spike count (middle) and inter-spike interval (right, n=18) of W+T cells responding to three object positions.
W+T cells (n=20)
Inter-spike interval
spikes/cycles
delay to 1st spike
1.00.50.0
A
R2
0 0.5 10
9
18
0 0.5 10
4
8
0 0.5 10
5
10Delay Sp ISI
R2
B
-2 -1 0 1 2 3 40
3
6
-0.2 -0.1 0 0.1 0.20
6
12
-3 -2 -1 0 1 2 30
2.5
5
Slope
W+T cells (n=20)
Inter-spike interval
spikes/cycles
delay to 1st spike
1.00.50.0
Inter-spike interval
spikes/cycles
delay to 1st spike
1.00.50.0 1.00.50.0
A
R2
0 0.5 10
9
18
0 0.5 10
9
18
0 0.5 10
9
18
0 0.5 10
4
8
0 0.5 10
4
8
0 0.5 10
4
8
0 0.5 10
5
10
0 0.5 10
5
10Delay Sp ISI
R2
B
-2 -1 0 1 2 3 40
3
6
-0.2 -0.1 0 0.1 0.20
6
12
-3 -2 -1 0 1 2 30
2.5
5
Slope-2 -1 0 1 2 3 4
0
3
6
-2 -1 0 1 2 3 40
3
6
-0.2 -0.1 0 0.1 0.20
6
12
-0.2 -0.1 0 0.1 0.20
6
12
-3 -2 -1 0 1 2 30
2.5
5
-3 -2 -1 0 1 2 30
2.5
5
Slope
Figure 5. Encoding of horizontal object position by VPM W+T cells– population (n=20)
(A) Box plot of coefficients of determination (R2) for individual W+T cell’s linear regressions of delay to first spike, spike count per cycle and inter-spike interval against whisker-object contact time. Boxes indicate the first (25%) to third (75%) quartile values. Horizontal line indicates range and line inside box denotes median. Outlier (>3 inter-quartile values from median) is plotted with asterisks (∗).(B) Distribution of R2 (top) and slope (bottom) for delay to first spike (left), spike count (middle) and inter-spike interval (right, n=18) of W+T cells responding to three object positions.
41
timing. Spike count contained significant more information than delay to first spike
and inter-spike interval (p<0.05) (Figure 7).
Overall, out results here suggest that Touch cells in VPM could encode the
horizontal object position primarily by spike timing similar to those in NV. Whisking-
Touch cells could not complete object localization task at least by these three
neuronal codes, which is consistent with our previous hypothesis that Whisking-
Touch cells might involve in object identification, but not localization. Since POm
cells mainly convey sensor motion (whisking) signals (Yu et al., 2006), they might
coordinate with Touch cells to encode the object locations.
R2 = 0.41
0
0.5
1
0 5 10 15Contact time (ms)
Spik
es/c
ycle
R2 = 0. 01
0
60
120
0 5 10 15Contact time (ms)
Del
ay (m
s)R2 = 0.06
0
40
80
0 5 10 15Contact time (ms)
ISI (
ms)
0 100 2000
50
100
Time (ms)
Tria
l #Sp
ikes
/s
A B
C
D
Figure 6. Encoding of horizontal object position by a W-T cell
(A) Raster plots of W-T cell 616 responding to free air whisking (top) and three object positions (from second to bottom, posterior to anterior) during 3 randomly selected cycles from cycles 5-10 (72 trials), and steady-state PSTHs of the responses (free air whisking, black; object positions: from posterior to anterior, magenta, green and red)(B-D) Linear regressions of delay to first spike per cycle, spike count per cycle and interspikeinterval per cycle against the time of whisker-object contact.
R2 = 0.41
0
0.5
1
0 5 10 15Contact time (ms)
Spik
es/c
ycle
R2 = 0. 01
0
60
120
0 5 10 15Contact time (ms)
Del
ay (m
s)R2 = 0.06
0
40
80
0 5 10 15Contact time (ms)
ISI (
ms)
0 100 2000
50
100
Time (ms)
Tria
l #Sp
ikes
/s
A B
C
D
R2 = 0.41R2 = 0.41
0
0.5
1
0 5 10 15Contact time (ms)
Spik
es/c
ycle
R2 = 0. 01R2 = 0. 01
0
60
120
0 5 10 15Contact time (ms)
Del
ay (m
s)R2 = 0.06R2 = 0.06
0
40
80
0 5 10 15Contact time (ms)
ISI (
ms)
0 100 2000
50
100
0 100 2000
50
100
Time (ms)
Tria
l #Sp
ikes
/s
A B
C
D
Figure 6. Encoding of horizontal object position by a W-T cell
(A) Raster plots of W-T cell 616 responding to free air whisking (top) and three object positions (from second to bottom, posterior to anterior) during 3 randomly selected cycles from cycles 5-10 (72 trials), and steady-state PSTHs of the responses (free air whisking, black; object positions: from posterior to anterior, magenta, green and red)(B-D) Linear regressions of delay to first spike per cycle, spike count per cycle and interspikeinterval per cycle against the time of whisker-object contact.
42
Representations of whisking frequency during artificial whisking
Previous study showed that during the passive stimulus, the whisker frequency was
represented by two different coding schemes: response amplitude and spike counts
decrease as a function of the frequency in the VPM; Latencies increased and spike
count decreased as a function of the frequency in the POm (Ahissar et al., 2000). To
investigate how thalamic neurons encode whisking frequency during active touch, the
responses of thalamic neurons during frequency modulated (FM) whisking both in
free air and against objects were examined.
The FM stimulation was similar to those used in passive paradigm (Ahissar et al.,
2001). The whisker stimulation frequency was first kept constant at 5Hz for 1 s. Then,
R2
W-T cells (n=9)
1.00.50.0
Inter-spike interval
spikes/cycles
delay to 1st spike
A
Delay Sp ISI
0 0.5 10
3
6
0 0.5 10
2
4
0 0.5 10
2
4
R2
-4 -3 -2 -1 0 10
2
4
-0.2 -0.1 0 0.1 0.20
3
6
-3 -2 -1 0 1 2 30
1.5
3
Slope
B
Figure 7. Encoding of horizontal object position by VPM W-T cells– population
(A) Box plot of coefficients of determination (R2) for individual W-T cell’s linear regressions of delay to first spike, spike count per cycle and inter-spike interval against whisker-object contact time. Boxes indicate the first (25%) to third (75%) quartile values. Horizontal line indicates range and line inside box denotes median. Outlier (>3 inter-quartile values from median) is plotted with asterisks (∗).(B) Distribution of R2 (top) and slople (bottom) for delay to first spike (left), spike count (middle) and inter-spike interval (right, n=6)of W-T cells responding to three object positions.
R2
W-T cells (n=9)
1.00.50.0
Inter-spike interval
spikes/cycles
delay to 1st spike
1.00.50.0 1.00.50.0
Inter-spike interval
spikes/cycles
delay to 1st spike
A
Delay Sp ISI
0 0.5 10
3
6
0 0.5 10
3
6
0 0.5 10
3
6
0 0.5 10
2
4
0 0.5 10
2
4
0 0.5 10
2
4
0 0.5 10
2
4
0 0.5 10
2
4
0 0.5 10
2
4
R2
-4 -3 -2 -1 0 10
2
4
-0.2 -0.1 0 0.1 0.20
3
6
-3 -2 -1 0 1 2 30
1.5
3
Slope-4 -3 -2 -1 0 1
0
2
4
-4 -3 -2 -1 0 10
2
4
-0.2 -0.1 0 0.1 0.20
3
6
-0.2 -0.1 0 0.1 0.20
3
6
-0.2 -0.1 0 0.1 0.20
3
6
-3 -2 -1 0 1 2 30
1.5
3
-3 -2 -1 0 1 2 30
1.5
3
-3 -2 -1 0 1 2 30
1.5
3
Slope
B
Figure 7. Encoding of horizontal object position by VPM W-T cells– population
(A) Box plot of coefficients of determination (R2) for individual W-T cell’s linear regressions of delay to first spike, spike count per cycle and inter-spike interval against whisker-object contact time. Boxes indicate the first (25%) to third (75%) quartile values. Horizontal line indicates range and line inside box denotes median. Outlier (>3 inter-quartile values from median) is plotted with asterisks (∗).(B) Distribution of R2 (top) and slople (bottom) for delay to first spike (left), spike count (middle) and inter-spike interval (right, n=6)of W-T cells responding to three object positions.
43
it was modulated for an additional 2.5 s (40% modulation depth, 0.4Hz, initial phase
90°). For each cell, an object was placed at three different horizontal positions during
modulated whisking. The responses of each cell were averaged across 24 repetitions
in both whisking and touch trial respectively. The dynamic behavior of neuronal
representation by FM stimulation was investigated across different thalamic nuclei by
computing neuronal variables: delay to first spike (or onset latency) and spike count.
Figure 8 and 9 describe two examples of POm and VPMdm cells. For POm cell 607
(in average touch trial), the effect of the stimulus frequency was different for delay to
first spike and spike count. Delay to first spike increased with the frequency
increment, whereas spike count decreased with the frequency increment during
frequency modulated stage (Figure 8A). Both these variables presented statistically
significant correlation with whisking frequency (R2 = 0.64 and R2 = 0.68,
respectively.). For VPMdm cell 673 (in whisking trial), the stimulus frequencies had
obvious effect on spike count, but not delay to first spike (Figure 9A). Spike count
increased with increasing instantaneous stimulus frequencies (R2 = 0.72, Figure 9B).
Since neuronal codes of individual VPMvl cells did not exhibit any relation with
whisking frequency, no example from VPMvl is presented here.
POm cell
0.5
R2 = 0.680
1
2 4 6 8Frequency (Hz)
sp
Time (ms)
dela
y (m
s)
0
10
20
30
40
50
1 401 801 1146 1473 1973 2578 2976 32850
0.2
0.4
0.6
0.8
1
Spik
e-co
unt
A
B
R2 = 0.6415
25
35
2 4 6 8Frequency (Hz)
dela
y (m
s)
Figure 8. Responses of one POm cell to FM stimuli during object touch
(A) Stimulus frequencies (black), delay to first spike (blue) per cycle and spike counts (yellow) as a function of train time. Mean of 24 trials are presented.(B) Linear regression of delay to 1st spike (left) and spike counts (right) as a function of whisker frequency.
POm cell
0.5
R2 = 0.680
1
2 4 6 8Frequency (Hz)
sp
Time (ms)
dela
y (m
s)
0
10
20
30
40
50
1 401 801 1146 1473 1973 2578 2976 32850
0.2
0.4
0.6
0.8
1
Spik
e-co
unt
A
B
R2 = 0.6415
25
35
2 4 6 8Frequency (Hz)
dela
y (m
s)
POm cell
0.5
R2 = 0.68R2 = 0.680
1
2 4 6 8Frequency (Hz)
sp
Time (ms)
dela
y (m
s)
0
10
20
30
40
50
1 401 801 1146 1473 1973 2578 2976 32850
0.2
0.4
0.6
0.8
1
0
10
20
30
40
50
1 401 801 1146 1473 1973 2578 2976 32850
0.2
0.4
0.6
0.8
1
Spik
e-co
unt
A
B
R2 = 0.64R2 = 0.6415
25
35
2 4 6 8Frequency (Hz)
dela
y (m
s)
Figure 8. Responses of one POm cell to FM stimuli during object touch
(A) Stimulus frequencies (black), delay to first spike (blue) per cycle and spike counts (yellow) as a function of train time. Mean of 24 trials are presented.(B) Linear regression of delay to 1st spike (left) and spike counts (right) as a function of whisker frequency.
44
Local population analysis revealed that POm and VPMdm cells exhibited
transformed responses in somewhat different ways under the active stimulus. As the
stimulus frequency increased, POm cells exhibited increased delay to first spike (and
onset latency) and decreased spike count in both whisking and touch trials, which
were similar to the responses under the passive movement (Figure 10). In the
VPMdm, stimulus frequencies were represented by spike count in whisking trial, but
had no effect on spike count in touch trials (Figure 11). Meanwhile, delay to firs spike
(and onset latency) was not affected by whisking frequency in both whisking and
touch trials. Due to the absence of whisking responses in the VPMvl, cells rarely
display representations of whisking frequency and no correlation appears between
neural codes and frequencies during free air whisking. Moreover, during whisking
against the object, VPMvl Cells exhibited a constant, non-adapting response to the
modulated frequency in touch trial (Figure 12).
In short, POm cells could code whisking frequency by both first spike timing
(latency) and spike count regardless of the presence of object. Differently, VPMdm
cells could present input frequency by rate variable alone in a different way from
R2 = 0.720
1
2
3
2 4 6 8Frequency (Hz)
sp
Time (ms)
Del
ay (m
s)
0
20
40
60
80
1 401 801 1146 1473 1973 2578 2976 32850
0.5
1
1.5
2
2.5
Spik
e-co
unt
VPMdm cell A
B
R2 = 0.3220
30
40
50
2 4 6 8Frequency (Hz)
dela
y (m
s)
Figure 9. Representation of whisker frequency in one VPMdm cell during free air whisking.(A) Stimulus frequencies (black), delay to first spike (blue) per cycle and spike counts (yellow) as a function of train time. Mean of 24 trials are presented.(B) Linear regression of delay to 1st spike (left) and spike counts (right) as a function of whisker frequency.
R2 = 0.72R2 = 0.720
1
2
3
2 4 6 8Frequency (Hz)
sp
Time (ms)
Del
ay (m
s)
0
20
40
60
80
1 401 801 1146 1473 1973 2578 2976 32850
0.5
1
1.5
2
2.5
Spik
e-co
unt
VPMdm cell A
B
R2 = 0.32R2 = 0.3220
30
40
50
2 4 6 8Frequency (Hz)
dela
y (m
s)
Figure 9. Representation of whisker frequency in one VPMdm cell during free air whisking.(A) Stimulus frequencies (black), delay to first spike (blue) per cycle and spike counts (yellow) as a function of train time. Mean of 24 trials are presented.(B) Linear regression of delay to 1st spike (left) and spike counts (right) as a function of whisker frequency.
45
POm during whisking condition. The coding process was suppressed by the presence
of object. VPMvl cells did not involve in the coding of whisking frequency.
Figure 10. Representations of POm cell populations (n=14) to FM stimuli during free air whisking and object touch
(A, B and C) Linear regressions of delay to first spike per cycle, spike count per cycle and onset latency against the stimulus frequency in both whisking trial (black) and touch trial (red)(D) Linear regression of spike counts as a function of delay to first spike in both whisking trial (black) and touch trial (red).
C Frequency (Hz)
R2 = 0. 83
R2 = 0. 91
15
20
25
30
35
2 4 6 8
Del
ay to
1st
spik
e (m
s)
Frequency (Hz)
R2 = 0. 70R2 = 0. 63
0.2
0.3
0.4
0.5
0.6
2 4 6 8
Spik
e-co
unt (
sp/c
ycle
)
BA
R2 = 0. 54R2 = 0. 78
15
20
25
30
35
2 4 6 8Frequency (Hz)
late
ncy
(ms)
R2 = 0. 72R2 = 0. 77
0.2
0.3
0.4
0.5
0.6
20 22 24 26 28 30 32
Delay to 1st spike (ms)
Spik
e-co
unt (
sp/c
ycle
)D
Figure 10. Representations of POm cell populations (n=14) to FM stimuli during free air whisking and object touch
(A, B and C) Linear regressions of delay to first spike per cycle, spike count per cycle and onset latency against the stimulus frequency in both whisking trial (black) and touch trial (red)(D) Linear regression of spike counts as a function of delay to first spike in both whisking trial (black) and touch trial (red).
C Frequency (Hz)
R2 = 0. 83
R2 = 0. 91
R2 = 0. 83
R2 = 0. 91
15
20
25
30
35
2 4 6 8
Del
ay to
1st
spik
e (m
s)
Frequency (Hz)
R2 = 0. 70R2 = 0. 63R2 = 0. 70R2 = 0. 63
0.2
0.3
0.4
0.5
0.6
2 4 6 8
Spik
e-co
unt (
sp/c
ycle
)
BA
R2 = 0. 54R2 = 0. 78R2 = 0. 54R2 = 0. 78
15
20
25
30
35
2 4 6 8Frequency (Hz)
late
ncy
(ms)
R2 = 0. 72R2 = 0. 77R2 = 0. 72R2 = 0. 77
0.2
0.3
0.4
0.5
0.6
20 22 24 26 28 30 32
Delay to 1st spike (ms)
Spik
e-co
unt (
sp/c
ycle
)D
Figure 11. Responses of VPMdm cell populations (n=13) to FM stimuli during free air whisking and object touch
A
R2 = 0. 60R2 = 0. 01
0.4
0.6
0.8
1
2 4 6 8
Frequency (Hz)
Spik
e-co
unt (
sp/c
ycle
)
2 4 6 8
Frequency (Hz)
R2 = 0. 01 R2 = 0. 02
15
20
25
30
35
Del
ay to
1st
spik
e (m
s)
R2 = 0. 15R2 = 0. 21
0
10
20
30
40
2 4 6 8
Frequency (Hz)
late
ncy
(ms)
R2 = 0. 00R2 = 0. 54
0.4
0.6
0.8
1
18 20 22 24 26 28 30 32
Delay to 1st spike (ms)
Spik
e-co
unt (
sp/c
ycle
)C D
B
Figure 11. Responses of VPMdm cell populations (n=13) to FM stimuli during free air whisking and object touch
A
R2 = 0. 60R2 = 0. 01
0.4
0.6
0.8
1
2 4 6 8
Frequency (Hz)
Spik
e-co
unt (
sp/c
ycle
)
2 4 6 8
Frequency (Hz)
R2 = 0. 01 R2 = 0. 02
15
20
25
30
35
Del
ay to
1st
spik
e (m
s)
R2 = 0. 15R2 = 0. 21
0
10
20
30
40
2 4 6 8
Frequency (Hz)
late
ncy
(ms)
R2 = 0. 00R2 = 0. 54
0.4
0.6
0.8
1
18 20 22 24 26 28 30 32
Delay to 1st spike (ms)
Spik
e-co
unt (
sp/c
ycle
)C D
BA
R2 = 0. 60R2 = 0. 01
0.4
0.6
0.8
1
2 4 6 8
Frequency (Hz)
Spik
e-co
unt (
sp/c
ycle
)
2 4 6 8
Frequency (Hz)
R2 = 0. 01 R2 = 0. 02
15
20
25
30
35
Del
ay to
1st
spik
e (m
s)
R2 = 0. 15R2 = 0. 21
0
10
20
30
40
2 4 6 8
Frequency (Hz)
late
ncy
(ms)
R2 = 0. 00R2 = 0. 54
0.4
0.6
0.8
1
18 20 22 24 26 28 30 32
Delay to 1st spike (ms)
Spik
e-co
unt (
sp/c
ycle
)C D
B
R2 = 0. 60R2 = 0. 01R2 = 0. 60R2 = 0. 01
0.4
0.6
0.8
1
2 4 6 8
Frequency (Hz)
Spik
e-co
unt (
sp/c
ycle
)
2 4 6 8
Frequency (Hz)
R2 = 0. 01 R2 = 0. 02R2 = 0. 01 R2 = 0. 02
15
20
25
30
35
Del
ay to
1st
spik
e (m
s)D
elay
to 1
stsp
ike
(ms)
R2 = 0. 15R2 = 0. 21R2 = 0. 15R2 = 0. 21
0
10
20
30
40
2 4 6 8
Frequency (Hz)
late
ncy
(ms)
R2 = 0. 00R2 = 0. 54R2 = 0. 00R2 = 0. 54
0.4
0.6
0.8
1
18 20 22 24 26 28 30 32
Delay to 1st spike (ms)Delay to 1st spike (ms)
Spik
e-co
unt (
sp/c
ycle
)C D
B
46
Discussion
We showed here how thalamic neurons encode the horizontal position of an object
and whisker frequency during active touch in anesthetized rats. Touch neurons in the
VPM encode the horizontal object position by their spike timing relative to whisking
onset. The principal variables used to encode whisker frequency are different across
POm, VPMdm and VPMvl: spike timing and spike count in the POm in both
whisking and touch conditions, and spike count alone in the VPMdm during
whisking condition, while VPMvl neurons do not present any information about
whisker frequencies in both conditions. Our results are similar to previous findings of
Swzed et al (Szwed et al., 2003), who observed temporal coding of horizontal object
location in the NV during artificial whisking, and to some extent to those of Ahissar et
R2 = 0. 01R2 = 0. 225
1525354555
2 4 6 8Frequency (Hz)
Del
ay to
1st
spik
e (m
s)
Figure 12. Representations of VPMvl touch cell populations (n=9) to FM stimuli during free air whisking and again object.
(A, B and C) Linear regressions of delay to first spike per cycle, spike count per cycle and onset latency as function of stimulus frequency in both whisking (black) and touch (red) conditions.(D) Spike counts as a function of delay to first spike in both whisking (black) and touch (red) conditions.
A
C
B
D
R2 = 0. 45R2 = 0. 15
0
0.2
0.4
0.6
2 4 6 8Frequency (Hz)
Spik
es/c
ycle
R2 = 0. 08R2 = 0. 19
0
20
40
60
2 4 6 8Frequency (Hz)
late
ncy
(ms)
R2 = 0. 27
R2 = 0. 00
0
0.2
0.4
0.6
20 30 40 50Delay to 1st spike (ms)
Spik
e-co
unt (
sp/c
ycle
)
(A, B and C) Linear regressions of delay to first spike per cycle, spike count per cycle and onset latency as function of stimulus frequency in both whisking (black) and touch (red) conditions.(D) Spike counts as a function of delay to first spike in both whisking (black) and touch (red) conditions.
R2 = 0. 01R2 = 0. 22 R2 = 0. 01R2 = 0. 225
1525354555
2 4 6 8Frequency (Hz)
Del
ay to
1st
spik
e (m
s)
Figure 12. Representations of VPMvl touch cell populations (n=9) to FM stimuli during free air whisking and again object.
(A, B and C) Linear regressions of delay to first spike per cycle, spike count per cycle and onset latency as function of stimulus frequency in both whisking (black) and touch (red) conditions.(D) Spike counts as a function of delay to first spike in both whisking (black) and touch (red) conditions.
A
C
B
D
R2 = 0. 45R2 = 0. 15R2 = 0. 45R2 = 0. 15
0
0.2
0.4
0.6
2 4 6 8Frequency (Hz)
Spik
es/c
ycle
R2 = 0. 08R2 = 0. 19R2 = 0. 08R2 = 0. 19
0
20
40
60
2 4 6 8Frequency (Hz)
late
ncy
(ms)
R2 = 0. 27
R2 = 0. 00
R2 = 0. 27
R2 = 0. 00
0
0.2
0.4
0.6
20 30 40 50Delay to 1st spike (ms)
Spik
e-co
unt (
sp/c
ycle
)
(A, B and C) Linear regressions of delay to first spike per cycle, spike count per cycle and onset latency as function of stimulus frequency in both whisking (black) and touch (red) conditions.(D) Spike counts as a function of delay to first spike in both whisking (black) and touch (red) conditions.
47
at (Ahissar et al., 2000), who observed both temporal and rate coding of whisking
frequency in the POm, and rate coding alone in the VPM during passive stimuli.
Artificial whisking
The mechanisms of sensory processing in the rat vibrissal system are derived
primarily from studies of passive (mechanical) movements of the whiskers. During
passive stimuli, the object moves to stationary whisker, forces are applied only to a
single point of the whisker's shaft. In nature, rat actively moves their whiskers to
perform a wide variety of tasks, such as discriminate textures (Carvell and Simons,
1990) and localize objects (Brecht et al., 1997; Harvey et al., 2001). Whiskers are
moved back and forth by their muscles. Our artificial whisking paradigm mimics that
which occurs in nature: forces are applied both to the whisker’s follicle and to the
whisker’s shaft. Our previous study in NV indicated that responses of a given neuron
to passive stimuli provide only limited information about its response in the active
whisking. The observed selectivity to major components of active touch in NV could
not be inferred from the responses of the same cells to passive stimuli. Moreover,
even the polarity of responses (i.e., response to contact or to detachment) in active
touch could not be predicted from their directional selectivity with passive stimuli.
The results from NV study suggest that experiments aimed at understanding active
vibrissal touch should be done in the active mode, since input to the vibrissal system
differs qualitatively between the passive and active modes. Although the principle of
muscle-driven whisker movement is similar between artificial and natural whisking,
there are still differences in detail, such as the absence of active retraction, artificial
stimulus-locked component superimposed on protraction and less changeable
whisking pattern across cycles in our artificial whisking. Thus, although the principles
by which movement properties are encoded in neuronal responses are expected to
occur in behavior rats, the exact coding scheme of thalamic neurons during active
touch still need to be further examined in awaked behavior rats.
Encoding of object location
In this chapter, we clarified which variables in the neuronal signals that constitute
the input to the vibrissal system encode the location of external objects and whisker
frequencies. Previous research in NV (Szwed et al., 2003) indicate that horizontal
object position is encoded by (i) coincident firing of individual Whisking and Contact
48
cells, and (ii) the temporal interval between firing onset of Whisking cells and firing
of Contact cells. These data suggest there are two specific encoding-decoding
schemes for horizontal object position in the vibrissal system. The spatiotemporal
encoding scheme, which involves a coincidence/phase detector circuit, suggested that
thalamic VPM neurons provide a spatial code of horizontal object location,
implicating that VPM cells are selective for one specific object location and have no
phase coding. The temporal scheme suggested that horizontal object position is
encoded by the spike count of thalamic POm neurons. However, we found here that
individual thalamic Touch cells precisely encode the horizontal object location by
their first-spike timing similarly to NV cells, which is not consistent with the
spatiotemporal encoding scheme for the VPM. On the other hand, POm neurons
mainly convey whisker-motion signals during active touch, which does not support
the temporal encoding scheme. Although our thalamic results do not support these
two working models, such models still could occur later in downstream brain areas.
For example, VPMvl mainly contains Touch neurons, thus, the second temporal
circuit might be implemented in a site integrating POm and VPMvl outputs, such as
S1, S2 or M1.
The dimensions of whisker-ralated spatial coordinates are horizontal (i.e., its
location along the anterior-posterior axis, parallel to the whisker rows), radial (i.e.,
from the snout outward) and vertical (i.e., parallel to whisker arcs). Studies in NV
indicated that object location is encoded by different neuronal variables along these
three dimensions: timing of Contact cells (temporal code) for the horizontal, spike
count of Touch cells (rate code) for the radial, and whisker identity (spatial code) for
the vertical coordinate. For integration of information related to these three different
codes, central circuits of the vibrissal system should translate the language of some of
the channels to the language of the others. To control movement, sensory codes
should, at some stage, be translated to motor codes. All these requirements can
probably be satisfied by a single transformation: temporal to rate. This is because
spatial and rate codes are already coded by firing rates, and motor control also appears
to be based on (population) rate coding (Kleinfeld et al., 1999; Salinas et al., 2000;
Wessberg et al., 2000; Zhang and Barash, 2000; Berg and Kleinfeld, 2003a; Carmena
et al., 2003; Ahrens and Kleinfeld, 2004; Brecht et al., 2004; Haiss and Schwarz,
2005). Using passive stimuli, we found that such a temporal-to-rate transformation is
implemented in one of the ascending sensory pathways – the paralemniscal one
49
(Ahissar et al., 2000). A temporal-to-rate code transformation can be implemented by
several circuits, one of them is a neuronal phase-locked loop (NPLL) (Ahissar and
Vaadia, 1990; Ahissar et al., 1997; Ahissar, 1998; Zacksenhouse and Ahissar, 2006).
A non-intuitive critical prediction of an NPLL is that response latency should increase
with the stimulus frequency (Ahissar et al., 1997; Ahissar, 1998; Kleinfeld et al.,
1999). With passive stimuli in whisking-range frequencies (5–11 Hz) (Ahissar et al.,
2000; Sosnik et al., 2001), we found that this is indeed the case in POm, but not in
VPM. This suggests that NPLLs might be implemented in thalamocortical loops of
the paralemniscal system (Ahissar and Arieli, 2001; Moore, 2004), or in other loops
that connect the POm with an oscillatory source. However, frequency-dependent
latency shifts could also be caused by other mechanisms. In particular, thalamic
GABAB inhibition, when combined with specific brainstem activation profiles, could
induce a similar phenomenon (Golomb et al., 2006). Whatever the underlying
mechanism is, a temporal-to-rate code translation would be instrumental for a
sensory-motor loop controlling whisking kinematics. Indeed, our results here suggest
that the POm participates in such a loop. Using active whisking, we found that POm
neurons responded consistently and reliably to whisking movements, irrespective of
whether the whiskers contacted on an object or not. The response latency increased
with the stimulus frequency increment. Overall, our findings suggest that POm is
involved in temporal processing related to sensory-motor control of whisker
movement, possibly using NPLL-like mechanisms.
Encoding of whisking frequency
In the vibrissa sensory system, rat sensory behaviors and stimulus features generate
specific frequency ranges of whisker motion. During active exploration, rats typically
sweep their whiskers at 4-12Hz (Carvell and Simons, 1990; Harvey et al., 2001)
against and over tactual surfaces, and during rest or quiescence, their whiskers are
typically still (<1Hz). Part of the information obtained by rat whiskers is carried by
the frequency of their movements. The precise frequency of whisking may impact the
accuracy of perceptual judgments (Carvell and Simons, 1995; Harvey et al., 2001).
Using passive paradigm, our previous studies indicated that there are two coding
streams representing frequency-dependent changes in the thalamus in anesthetized rat.
In POm, neuronal responses increased in latency and decreased in magnitude as
50
whisker frequency increased. These latency shifts provides a temporal code of
deflection frequency, implementing a phase-locked loop (see discussion above).
On the other hand, frequency was encoded by changes in response magnitude in VPM
(Diamond et al., 1992a; Ahissar et al., 2000; Sosnik et al., 2001). Using artificial
whisking paradigm, our results here further clarified the encoding scheme of whisker
frequencies in the POm, i.e., first-spike timing increased and spike count decreased as
a function of frequency. However, the spike-count increased rather than decreased
with frequency increment was presented in the VPM during active free air whisking.
The Frequency-dependent adaptation observed for free air whisking in VPM did not
generalize to other stimuli. Such as, in localization task, changes in the pattern of
stimulation during touch conditions evoked non-adapted responses that might enhance
detection of this novel stimulus feature (Abbott et al., 1997; Garabedian et al., 2003).
Using passive stimuli, several researches complementarily revealed the frequency-
dependent characteristics of VPM response in anesthetized rats by accessing the
relative adaptation of neuronal firing at the different stimulus frequencies (Castro-
Alamancos, 2002a, b; Chung et al., 2002; Deschenes et al., 2003; Hartings et al.,
2003). When the effect of stimulation frequency on spike rate is measured in a brief
window after vibrissa deflection, VPM neurons demonstrated significant adaptation
(i.e., the reduction of responses with increased input frequency) at higher stimulus
frequencies, which is consistent with the results concluded from our group.
In contrast, when firing rate is calculated instead as the total number of spikes evoked
over an extended stimulation period, the number of spikes increased slightly with
each increase in the frequency of stimulation (Hartings et al., 2003). This is somewhat
consistent with the results we observed in the VPM by using artificial whisking.
Furthermore, during activated states, the relay of high-frequency sensory inputs is
allowed through the thalamus. Consequently, sensory adaptation is mostly absent in
the thalamus during activated states (Fanselow and Nicolelis, 1999; Castro-
Alamancos, 2002a, b).
Previous studies showed that frequency tuning characteristics in VPM are
attributable to synaptic depression of the feedforward excitatory postsynaptic
potentials from the bran stem reticular formation and level of postsynaptic
depolarization of VPM neurons, which work together to serve as a gating mechanism
of sensory information flow. As a result, the delay of high-frequency inputs are
permitted during activated states (Castro-Alamancos, 2002a, b). Increased thalamic
51
depolarization can overcome synaptic depression at this synapse (Castro-Alamancos,
2002b), suggesting that the different adaptation profiles between passive stimuli and
our active stimuli reflect, in part, and the degree of postsynaptic depolarization under
different recording conditions. In addition, VPm firing is affected by inhibitory input
from the thalamic reticular nucleus (RT) (Cox et al., 1996; Pinault and Deschenes,
1998). Depressing this inhibitory input can also facilitate the relay of high-frequency
signals in VPM (Castro-Alamancos, 2002b). Thus, Rt-mediated inhibition can act in
concert with neuromodulator mechanisms to enhance transmission of high-frequency
inputs. Furthermore, previous studies demonstrated that low whisker deflection
velocity at low frequencies ( 2 Hz) evokes few spikes in VPM relative to background
activity. Faster whisker movement velocities at higher frequencies (2-10Hz) are
presented by high firing rates. Thus, the presence of facilitation rather than adaptation
during our active free air whisking may somewhat attributable to the velocity
sensitivity of VPM firing. So far, the mechanisms of frequency-dependent
characteristics of POm are still less well understood.
52
CHAPTER 3: Coding of object location in the secondary
somatosensory cortex during active vibrissal touch2
Summary Sensation is an active process. When the sensors move, both motion and touch
signals are crucial for sensory processing. Recently, we showed that whisker-motion
and contact signals are conveyed in parallel via the thalamus. Yet, it is unknown how
downstream brain areas process these signals to localize objects. Specifically, the
function of the secondary somatosensory cortex (S2), which receives massive
thalamic projections from both whisker-motion and contact pathways, is poorly
understood. Here, neuronal activities from S2 in response to object touch were
recorded using artificial whisking in anesthetized rats. We found that high proportion
of neurons (41%) in deeper layers (layers 4, 5 and 6) of S2 (S2-L46) show significant
selectivity to object location, in distinction with the superficial layers (layers 2/3) of
S2 (20%). Similar to the latter, small fractions (<23%) of neurons in three thalamic
nuclei (POm, VPMdm, VPMvl) and three layers (2/3, 4 and 5a) of the primary
somatosensory cortex (S1) displayed significant selectivity to object location. We
further investigated the dynamics of responses in all these somatosensory sites during
whisking in free air and against objects in various locations. Comparison of the
dynamics of location-selective (S) cells with those of non-selective cells revealed two
differences: a lack of a slow-wave dynamics in S cells, and a stronger response of S
cells to the first cycle in S2-L46, S1-L5a and S1-L23. Comparison of response
dynamics across nuclei revealed two major patterns of cortical dynamics, one evident
in S2-L46 and S1-L5a and another in S2-L23 and S1L4. The analysis so far thus
suggests that cortical processing of active touch is done in parallel in at least two
streams, and that processing of object location involves several brain sites, among
which a significant processing of object location is performed in the deep layers of
S2.
2 Yu, C., Haidarliu, S., Derdikman, D., and Ahissar, E. (in preparation) Coding of object
location in the secondary somatosensory cortex during active vibrissal touch
53
Background
Much of our sensory information is acquired via moving sensors (Ahissar and
Arieli, 2001). In the rat vibrissal system, the whiskers serve as arrays of highly
sensitive detectors. The existence, location and texture of stationary objects are
sampled by whiskers while moving across them (Carvell and Simons, 1990; Brecht et
al., 1997; Fanselow and Nicolelis, 1999; Sachdev et al., 2000; Harvey et al., 2001).
Since the same whisker can be activated by objects located in different locations
along its scanning path, the brain needs to know where the whisker is at the moment
of activation in order to localize the object. Whisker movements are reported to brain
by a set of neurons located in the first input stage of the vibrissal system (Lichtenstein
et al., 1990; Pali et al., 2000; Szwed et al., 2003; Leiser and Moxon, 2007).
Like the visual and auditory systems, which are characterized by multiple
representations within the cerebral cortex, the tactile system of the rat comprises at
least two principal cortical processing domains (Welker, 1971; Walker and Sinha,
1972; Welker, 1976; Akers and Killackey, 1978; Simons, 1978; Land and Simons,
1985a; Koralek et al., 1990; Fabri and Burton, 1991). Tactile information captured by
the whisker is consecutively transmitted to the primary somatosensory cortex (S1),
along anatomically and functionally well-defined pathways (Chmielowska et al.,
1989; Chiaia et al., 1991; Diamond et al., 1992b; Lu and Lin, 1993; Ahissar et al.,
2000; Pierret et al., 2000; Veinante et al., 2000; Yu et al., 2006), and to the secondary
somatosenory cortex (S2), which has been more anatomically, but less functionally
explored (Carvell and Simons, 1987; Spreafico et al., 1987; Koralek et al., 1988;
Remple et al., 2003; Brett-Green et al., 2004; Kwegyir-Afful and Keller, 2004;
Benison et al., 2006; Melzer et al., 2006; Kamatani et al., 2007). A series of
anatomical studies reported that these two cortical regions are reciprocally connected
(Fabri and Burton, 1991; Kim and Ebner, 1999; Alloway et al., 2003; Chakrabarti and
Alloway, 2006). However, like in other species and human (Pons et al., 1987; Murray
et al., 1992; Pons et al., 1992; Turman et al., 1992; Turman et al., 1995; Zhang et al.,
1996; Coleman et al., 1999; Karhu and Tesche, 1999; Zhang et al., 2001b; Barba et
al., 2002; Inui et al., 2004), the functional association of these two somatosensory
cortical areas is still less clearly understood.
Focusing on understanding the neural mechanism of object localization during
active touch in the rat vibrissal system, our recent studies indicated that different
54
sensory signals related to active touch are conveyed separately by different classes of
first-neurons in TG: Whisking, Whisking-touch and Touch, Which could encode
horizontal object locations by spike timing and radial locations by firing rate (Szwed
et al., 2003; Szwed et al., 2006). The information about active touch flows through
three parallel and functional distinctive channels in the thalamus: paralemnical
pathway via the POm, extralemniscal pathway via the VPMvl and lemniscal pathway
via VPMdm (Yu et al., 2006). Studying cortical representations of active touch and
object location, we found that laminar difference in S1 during active touch match the
difference between their thalamic counterparts: S1L5a neurons are primarily sensitive
to whisking involving motion control, similar to POm; Whereas S1L4 bareel neurons
are sensitive to both whisking and touch involving object identification, similar to
those in VPMdm . However, we did not find cells that are sensitive to contact only,
such as those in the VPMvl, which suppose to cooperate with POm involving object
localization (Derdikman et al., 2006a; Yu et al., 2006). Anatomical studies indicate
that the major projections of VPMvl are to deep layers of S2, which is also a target of
POm (Carvell and Simons, 1987; Spreafico et al., 1987; Koralek et al., 1988; Alloway
et al., 2000; Pierret et al., 2000). Thus, S2 is becoming an interestingly potential
region implementing the integration of paralemniscal and lemniscal signals for coding
object locations. However, physiology recordings in S2 cortex of rat have been rarely
investigated (Kwegyir-Afful and Keller, 2004; Melzer et al., 2006), especially during
active whisking. Thus, using the artificial active whisking with extracellular
recordings (Zucker and Welker, 1969; Brown and Waite, 1974; Szwed et al., 2003;
Derdikman et al., 2006a; Szwed et al., 2006; Yu et al., 2006), we attempt to
understand the whisker-related information processing performed in S2 during active
touch. Specifically, we attempt to investigate what external information related to
object localization is conveyed by S2 neurons and how S2 neurons encode object
location compare to other cortical and sub-cortical nuclei during active whisking.
Combined with our previous recordings from SI and thalamus (Derdikman et al.,
2006a; Yu et al., 2006), our results indicated that there is a functional segregation
within S2; the deeper layers of S2 contain high proportion of location-selective
neurons, which can independently encode object location parallel to those in S1.
55
Materials and Methods
Animal preparations and Recording procedures
Experiments on secondary somatosensory cortex were conducted on 19 male
Albino Wistar rats with body weight ranging from200-300 g. All procedures were
approved by the Institutional Animal Care and Use Committee of The Weizmann
Institute of Science. Experimental protocols were similar to those previously
described (Szwed et al., 2003; Derdikman et al., 2006a; Szwed et al., 2006; Yu et al.,
2006). In brief, surgery was performed under general anesthesia with urethane (1.5
g/kg, i.p.). Supplemental doses of anesthetic (10% of initial dose) were administered
when required. Atropine methyl nitrate (0.3 mg/kg, i.m.) was administered to prevent
respiratory complications. Anesthetized animals were mounted in a stereotaxic device
(SR-6; Narishige; Japan) which allows free access to the somatosensory brain
structure and to the whiskers. Body temperature was maintained at 37°C during
experimental manipulations. A craniotomy was made overlying the right secondary
somatosensroy cortex. According to known stereotaxic coordinates of S2, up to 4
tungsten microelectrodes (0.5-1 MΩ; Alpha Omega Engineering, Israel) were lowered
in parallel in each recording until units drivable by manual whisker stimulations were
encountered. Standard methods for single-unit recordings were used (Szwed et al.,
2003). Single-units were sorted online by spike templates. Units were considered
single only if they had homogenous spike shapes that did not overlap with other units
or noise, and if they exhibited refractory periods of > 1 ms in their autocorrelation
histograms. Artefacts produced by electrical stimulation were isolated by an online
spike-sorter (MSD-3.21; Alpha-Omega Engineering) and removed online from unit
recordings. Neuronal recordings from thalamus and primary somatosensory cortex
were similarly performed on 40 and 42 Albino Wistar rats, respectively (Derdikman
et al., 2006a; Yu et al., 2006)
Experimental paradigms and Histology
Artificial whisking was induced as described before in (Szwed et al., 2003;
Derdikman et al., 2006a; Szwed et al., 2006; Yu et al., 2006). Briefly, the facial nerve
was cut and its distal end mounted on a pair of silver electrodes. Bipolar, rectangular
electrical pulses (0.5-4.0 V, 40 μs duration) were applied through an isolated pulse
stimulator (Model 2100; A-M systems, Sequim, WA) at 83 Hz, the lowest frequency
that still produces continuous whisker movement, or 200 Hz which produce more
56
similar trajectory of whisker movement to those observed during natural whisking.
We induced 2s trains (5 Hz, 50% duty cycle) of artificial whisking followed by inter-
train intervals of 3 s in blocks of 12 trains (trials) each. Blocks of free-air artificial
whisking were interleaved with blocks of artificial whisking against an object
positioned in front of individual whiskers within the receptive field, i.e., the whiskers
that evoked the relatively large responses in the recorded cell during manual passive
stimulations. For each whisker, a vertical pole (2 mm diameter) was placed at 70-90%
of the whisker’s length from the skin, and at three different horizontal distances (1 to
9 mm) from the resting position of the whisker. Each of the four whisking conditions
(free-air and three object positions) was repeated in two blocks, interleaved in time.
All the whiskers of the mystacial pad were left intact throughout an experiment to
mimic natural conditions as much as possible. Precautions were taken to verify that
the examined whisker was always the first whisker to contact the object during
protraction. Thus, the other whiskers located in between the examined whisker and
the objects were moved rostral to the object prior to each block of trials. Whisker
movements were recorded at 1000 frames/sec with a fast digital video camera
(MotionScope PCI 1000; Redlake; San Diego, CA). Video recordings were
synchronized with neurophysiological data with 1 ms accuracy (Szwed et al., 2003;
Knutsen et al., 2005)
At the end of each recording session, electrolytic lesions were induced by passing
currents (10μA, 2 x 4s, unipolar) through the tips of the recording electrodes. The
brains were then removed, fixed, sliced coronally, and stained for cytochrome
oxidase. Lesions located in the S2 could be clearly seen. Although laminar location in
S2 is not as well-defined as those in S1, it is still easy to distinguish six layers of S2 in
our histological data. Thus, lesions were classified as being in superficial layers (layer
2/3) and deep layers (layer 4, 5 and 6a) according to their histological sites.
Analysis of neuronal data
Trajectories of whisker movements were analyzed offline, using semi-automatic
image processing software (Knutsen et al., 2005). Whisking onset time was
determined from the video records as the time at which the whiskers started moving.
Raster plots and peri-stimulus-time-histograms (PSTHs; 1 ms bins, smoothed by
convolution with a triangle of area 1 and a base of ±10 ms) were computed and
57
examined for all trains of each cell. Average response latencies were computed from
PSTHs as the delay from protraction onset to 1/2 peak response (Sosnik et al., 2001)
Responses were analyzed during steady-state periods. Cycles 5 – 10 were selected as
those cycles in which virtually all S2 cortical, S1 cortical and thalamic neurons
exhibited stabilized responses (Derdikman et al., 2006a; Yu et al., 2006). Amplitudes
of response peaks were calculated as the maximum of the smoothed PSTH. Statistical
analysis was done with MINITAB (Minitab Inc), except for location selectivity,
which was computed with MATLAB. For non-normally distributed data,
nonparametric Mann-Whitney (Wilcoxon rank sum) tests were used to compare
samples. For significant location selectivity test, a pair-wise bootstrap test was used:
For each cell, the selective index (SI) was calculated from the real responsive PSTHs
for each combination of two out of three positions: SI (selectivity index) = (AT-W –
<AT-W>)/( AT-W + <AT-W>), where AT-W is the amplitude of the subtracted touch
response in one of two selected object positions, and <AT-W> is the is the mean of
touch response amplitudes after subtraction in selected two object locations
The raster plots of single spikes were randomized (1000times) by shuffling the trials
of these two positions (24 trials for each) and selected 24 repetitions with
replacement. The PSTH*s were built and simulated SI*s were calculated. Histogram
of 1000 bootstrap replications of SI* was drawn, and the probability of SI comparing
against SI* series was determined. For adaptation analysis, Spike counts were
calculated for the first 60 ms from the onset of protraction per cycle, and normalized
relative to the spike count in the first cycle: SPn = (SPi – SP1)/ (SPi + SP1), where SPi
is the spike count in cycle i, and SP1 is the spike count in cycle 1.Thus, the normalized
spike count attained positive values between 0 and 1 for facilitation and negative
values from 0 to –1 for depression.
Results The specificity of neuronal responses to active movement and touch in S2 were
examined by recording from 55 individual neurons located in the superficial (Layer 2
and 3) and deeper layers (layer 4,5 and 6a) of S2 in urethane anesthetized rats : S2L23
(n = 33) and S2L46 (n = 22) . Similar to our previous studies (Szwed et al., 2003;
Derdikman et al., 2006a; Szwed et al., 2006; Yu et al., 2006), repetitive whisking
movements were induced by stimulating the facial motor nerve at 5Hz, which is
58
within the natural whisking rate. Thus, artificial whisking produced in our
experiments contained 100ms active protraction superimposing a ripple component at
83 Hz or 200Hz, and un-stimulated 100 ms passive retraction. In the movement path
of the selected whisker, a pole of 2 mm diameter was presented vertically during
touch blocks, at 70-90% of the whisker’s length. No object was presented in free-air
whisking blocks.
As previously reported (Carvell and Simons, 1987; Kwegyir-Afful and Keller,
2004), S2 neurons responded to multi-whiskers. The median receptive field size was 9
whiskers (mean, 8.9±2.4 whiskers) in S2L23 and 10 whiskers (mean, 10.3±2.7
whiskers) in S1L46, no significant difference between two groups (p=0.062, non-
parametric Mann-Whitney Test ). Since rarely identified single principal whisker for
neurons encountered in SII during manual passive stimulation, each unit was recorded
by positioning the object in front of more than one whisker in the receptive field
during active whisking. The principal whisker was determined as the whisker that
caused the maximal response during active whisking.
The responses of POm (n=24), VPMvl (n=13), VPMdm (n=30), S1L23 (n=32), S1
L4 (n=24), and S1L5a (n=23) neurons, which were recorded under identical
conditions, and whose responses had been presented previously (Derdikman et al.,
2006a; Yu et al., 2006), were also subjected to extensive analysis for the comparison
with S2 neurons.
Coding of object locations
To examine how S2 neurons encode object location during active touch, whisking
and touch signals were quantified by measuring the response amplitude of cells to
whisking in air and to whisking against an object. Similar to thalamic and S1 cortical
responses, neuronal responses in S2 exhibit a dynamic phase: the responses change
from cycle to cycle during the first 3-5 cycles until a steady-state phase, in which the
responses remain stable (Derdikman et al., 2006a; Yu et al., 2006). The responses
averaged over the last six cycles of each whisking train during steady state are used to
specify S2 responses encountered.
For each cell, an object was introduced at three different horizontal positions during
active whisking. To quantity the touch responses, the average PSTH of the entire
population in each nucleus, in free air whisking cycle was subtracted from the average
PSTH of the entire population in each touch cycle. Figure 1A depicts the average
59
PSTHs (dash line) and subtracted PSTHs (solid line) of S2L46 neurons during
whisking and touch cycles. PSTHs for different conditions are indicated by different
colors. The earliest zero-crossing time point of the subtracted responses across the
population in each nucleus is defined as the touch response onset of the entire
population. For S2L46 cell populations, the earliest time of crossing point is 20 ms at
the most posterior position. Since all touch responses complete after 40 ms from touch
onset, the touch response amplitude of individual neurons after subtraction is
computed within 40ms from the zero-crossing time. Thus, the time window for touch
response in S2L46 is 20 ms to 60 ms. Under the same principle, the time window for
S2L23 is 25 ms to 65 ms. The time windows computed for POm, VPMvl, VPMdm,
S1L23, S1 L4, and S1L5a are 22-62, 10-50, 13-53, 16-56, 13-53 and 24-64 ms,
correspondingly.
To examine the specificity of touch response in each location, normalized touch
responses within selected time window in each location is estimated as SI (selectivity
index) = (AT-W – <AT-W>)/( AT-W + <AT-W>), where AT-W is the amplitude of the
subtracted touch response in one of three object positions, and <AT-W> is the is the
mean of touch response amplitudes after subtraction in all three object locations. Two
examples of selective responses in S2L46 are depicted in Figure 1B-E. These two
cells exhibit strongest responses either to most posterior location (P1) or to middle
location (P2) in response to whisking and touches , but almost no touch responses to
other two locations (P2 and P3 or P1 and P3). The SIs of the most responsive
locations are 0.35 and 0.30, which significantly higher than others (For cell 890, SI1 =
-0.31 and SI3 = -0.44, p<0.0001; for cell 872, SI2 = -0.24 and SI3 = -0.29, p<0.0001,
pair-wise bootstrap test).
To examine and compare selective touch response properties in each nucleus, the
maximum SI of individual neurons was calculated across S2, S1 and thalamus.
Individual neurons in both S2L23 and S2L46 show mostly high selectivity (Figure 2).
However, SI of S2L46 neurons is significant higher than those in POm, VPMdm,
VPMvl, S1L4 and S1L5a (p<0.018, non-parametric Mann-Whitney Test), whereas SI
of neurons in S2L23 are only significant higher than those in S1L5a (p=0.015, Mann-
Whitney Test), and have no difference from others (p>0.076, Mann-Whitney Test).
Relatively, neurons in S1L23 also exhibit high SI, but only significantly higher than
those in VPMvl (p=0.047) and S1L5a (p=0.028).
60
Figure 1. Responses of S2L46 neurons at steady state during active whisking.
A, Average PSTHs (dash line) and subtracted PSTHs (solid line) of S2L46 neurons responding to free air whisking (black) and three horizontal locations (from posterior to anterior, red, green and blue, respectively) during protraction phase. Vertical lines denote time window (40ms) for touch responses. Time of touch response onset is 20ms.B,D, PSTHs of S2L46 cell 890 and 872 response at steady state during protraction phase. Insets show subtracted PSTHs of these two cells. C,E, SI values of these two cells at three object positions (from posterior to anterior, P1, P2 and P3) are indicated by the color codes.
A
D
P1
P2
P3
ES2L46 cell 872
0 50 1000
25
50
0 50 100
0
30
Time (ms)
Spik
es/s
CB
0 50 1000
50
100S2L46 cell 890
0 50 100
0
40
-1
0
1 P1
P2
P3
Time (ms)
Spik
es/s
0 20 50 60 100-10
0
35
Time (ms)
Spik
es/s
Free airP1P2P3
Figure 1. Responses of S2L46 neurons at steady state during active whisking.
A, Average PSTHs (dash line) and subtracted PSTHs (solid line) of S2L46 neurons responding to free air whisking (black) and three horizontal locations (from posterior to anterior, red, green and blue, respectively) during protraction phase. Vertical lines denote time window (40ms) for touch responses. Time of touch response onset is 20ms.B,D, PSTHs of S2L46 cell 890 and 872 response at steady state during protraction phase. Insets show subtracted PSTHs of these two cells. C,E, SI values of these two cells at three object positions (from posterior to anterior, P1, P2 and P3) are indicated by the color codes.
A
D
P1
P2
P3
ES2L46 cell 872
0 50 1000
25
50
0 50 100
0
30
Time (ms)
Spik
es/s
CB
0 50 1000
50
100S2L46 cell 890
0 50 100
0
40
-1
0
1 P1
P2
P3
Time (ms)
Spik
es/s
0 20 50 60 100-10
0
35
Time (ms)
Spik
es/s
Free airP1P2P3
A
D
P1
P2
P3
ES2L46 cell 872
0 50 1000
25
50
0 50 100
0
30
Time (ms)
Spik
es/s
CB
0 50 1000
50
100S2L46 cell 890
0 50 100
0
40
-1
0
1 P1
P2
P3
Time (ms)
Spik
es/s
0 20 50 60 100-10
0
35
Time (ms)
Spik
es/s
A
D
P1
P2
P3
P1
P2
P3
ES2L46 cell 872
0 50 1000
25
50
0 50 100
0
30
0 50 100
0
30
Time (ms)
Spik
es/s
CB
0 50 1000
50
100S2L46 cell 890
0 50 100
0
40
0 50 100
0
40
-1
0
1 P1
P2
P3
-1
0
1 P1
P2
P3
Time (ms)
Spik
es/s
0 20 50 60 100-10
0
35
0 20 50 60 100-10
0
35
Time (ms)
Spik
es/s
Free airP1P2P3
Free airP1P2P3
61
To further statistically clarify the location selectivity of cells in each nucleus, the
pair-wise bootstrap test was used (see Methods). High proportion of location selective
(S) cells is found in S2L46, while much less is presented in S2L23 and other nuclei.
41% of neurons (9/22) in S2L46 display significant selectivity to one of three
horizontal object locations (Figure 3A, p<0.05, pair-wise bootstrap test).
Distinctively, only 19% of neurons (6/31) in S2L23 show significant location
selectivity. Similar to S2L23, less location-selectivity is observed in POm (4%, 1/24),
VPMvl (23%, 3/13), VPMdm (20%, 6/30), S1L23 (13%, 4/32), S1L4 (8%, 2/24) and
S1L5a (13%, 3/23). The proportion of location-selective cells in S2L46 is
significantly higher than those in other nuclei ((p=0.002, exact binomial test). No
salient difference was observed when location selectivity was checked across different
Figure 2. Comparison of maximum selective index (SI) of cells at three object locations in the S2, S1 and thalamus
A, Distribution of maximum SIs of cells in the S2L23 (n=33), S2L46 (n=22), POm (n=24), VPMdm (n=30), VPMvl (n=13), S1L23 (n=31), S1L4 (n=24) and S1L5a (n=23).B, Mean maximum SI of S2, S1 and thalamic cells.
B
nuclei
0.0
0.1
0.2
0.3
S2L23S2L
46 PomVPMvl
VPMdmS1L
23 S1L4S1L
5a
Mea
n of
max
-SI
0
5
10 VPMdm
SI (max)
A S2L23
0
5
10
0
5
10
0
5
10
0 0.1 0.2 0.3 0.4 0.5 0.6
S1L23
S1L4
S1L5a
S2L46
POm
0
5
10
0
5
10
0
5
10
0
5
10 VPMvl
0 0.1 0.2 0.3 0.4 0.5 0.6
SI (max)
No.
of c
ells
Figure 2. Comparison of maximum selective index (SI) of cells at three object locations in the S2, S1 and thalamus
A, Distribution of maximum SIs of cells in the S2L23 (n=33), S2L46 (n=22), POm (n=24), VPMdm (n=30), VPMvl (n=13), S1L23 (n=31), S1L4 (n=24) and S1L5a (n=23).B, Mean maximum SI of S2, S1 and thalamic cells.
B
nuclei
0.0
0.1
0.2
0.3
S2L23S2L
46 PomVPMvl
VPMdmS1L
23 S1L4S1L
5a
Mea
n of
max
-SI
0.0
0.1
0.2
0.3
S2L23S2L
46 PomVPMvl
VPMdmS1L
23 S1L4S1L
5a
Mea
n of
max
-SI
0
5
10
0
5
10 VPMdm
SI (max)
A S2L23
0
5
10
0
5
10
0
5
10
0 0.1 0.2 0.3 0.4 0.5 0.6
S1L23
S1L4
S1L5a
S2L46
POm
0
5
10
0
5
10
0
5
10
0
5
10 VPMvl
0 0.1 0.2 0.3 0.4 0.5 0.6
0
5
10
0
5
10
0
5
10
0
5
10
0
5
10
0
5
10
0 0.1 0.2 0.3 0.4 0.5 0.60 0.1 0.2 0.3 0.4 0.5 0.60.6
S1L23
S1L4
S1L5a
S2L46
POm
0
5
10
0
5
10
0
5
10
0
5
10
0
5
10
0
5
10
0
5
10
0
5
10 VPMvl
0 0.1 0.2 0.3 0.4 0.5 0.60 0.1 0.2 0.3 0.4 0.5 0.60.6
SI (max)
No.
of c
ells
62
type of thalamic cells (W, 0%; W+T, 19%; W-T, 18% and T, 25%). The high
proportion (46%) of location-selective cells (S-cells) in S2L46 is also observed when
touch response was computed by spike-count (Figure 3B).
Overall, neurons of S2L23 and S2L46 present different response properties in the
processing of coding object location during active whisking, in which high percentage
of S2L46 neurons display statistically significant higher selectivity to object location,
which is distinctive from S2L23 neurons, other cortical and sub-cortical nuclei.
Comparison of response dynamics
To investigate the specific characteristics of location-selective (S) cells, we
compared the response dynamics of S-cells and non-S cells in free air whisking and
object touch (selective location for S-cell and average of three locations for non-S
cell). The comparison was also examined between S2L23 and S2L46, and across all
other nuclei.
The population responses of S cells and non-S cells within each nucleus are
depicted in Figure 4. The dynamics of response along the 2 s of artificial whisking are
stereotypic for both S cells and non-S cells in each of the 8 nuclei, in which the
Figure 3. Percentages of location-selective cells in S2, S1 and thalamus.
A, Percentages of location-selective cells computed from amplitude of touch response. Dash line denotes less than 20% of S-cells were found in most of nuclei.B, Percentages of location-selective cells computed from spike counts of touch response.
B
nuclei
POm
VPMVl
VPMdmS1L
23S1L
4S1L
5aS2L
23S2L
460
1020304050
S-ce
ll (%
)
nuclei
A
01020304050
S-ce
ll (%
)
POm
VPMVl
VPMdmS1L
23S1L
4S1L
5aS2L
23S2L
46
Figure 3. Percentages of location-selective cells in S2, S1 and thalamus.
A, Percentages of location-selective cells computed from amplitude of touch response. Dash line denotes less than 20% of S-cells were found in most of nuclei.B, Percentages of location-selective cells computed from spike counts of touch response.
B
nuclei
POm
VPMVl
VPMdmS1L
23S1L
4S1L
5aS2L
23S2L
46POm
VPMVl
VPMdmS1L
23S1L
4S1L
5aS2L
23S2L
460
1020304050
S-ce
ll (%
)
nuclei
01020304050
S-ce
ll (%
)
nuclei
A
01020304050
S-ce
ll (%
)
POm
VPMVl
VPMdmS1L
23S1L
4S1L
5aS2L
23S2L
46POm
VPMVl
VPMdmS1L
23S1L
4S1L
5aS2L
23S2L
46
63
transition to steady state occurs roughly around the 4th or 5th cycle both in free air
whisking and object touch for both cell types and in all sites except S cells in S2L23
(cycles 6-7) and POm (cannot be defined). Yet, the actual response dynamics of S and
non-S cell leading to the steady state responses are different during whisking and
touch trials. The first cycle responses are similar in whisking and touch cycles for
both S-cells and non-S cells across all the nuclei. However, the protraction responses
of S-cells during the consequent cycles are stronger in touch cycles than those in
whisking cycles, while the responses of non-S cells were not affected by the presence
of the object, except those of non-S cells in VPMvl and S1L4, which are also stronger
in touch than in whisking at the later cycles.
A S cell
0
20
40 S2L23 (n=6)
0
30
60S2L46 (n=9)
90
0
45POm (n=1)
0
25
50 VPMvl (n=3)
0
30
60 VPMdm(n=6)
40
0
20S1L23 (n=4)
0
25
50S1L4 (n=2)
40
0 200 400 600 800 1000 1200 1400 1600 1800 2000
0
20S1L5a (n=3)
Spik
es/s
Time (ms)
A S cell
0
20
40 S2L23 (n=6)
0
30
60S2L46 (n=9)
90
0
45POm (n=1)
0
25
50 VPMvl (n=3)
0
30
60 VPMdm(n=6)
40
0
20S1L23 (n=4)
0
25
50S1L4 (n=2)
40
0 200 400 600 800 1000 1200 1400 1600 1800 2000
0
20S1L5a (n=3)
A S cell
0
20
40
0
20
40 S2L23 (n=6)
0
30
60
0
30
60S2L46 (n=9)
90
0
45
0
45POm (n=1)
0
25
50
0
25
50 VPMvl (n=3)
0
30
60
0
30
60 VPMdm(n=6)
40
0
20
40
0
20
0
20S1L23 (n=4)
0
25
50
0
25
50S1L4 (n=2)
40
0 200 400 600 800 1000 1200 1400 1600 1800 2000
0
20
40
0 200 400 600 800 1000 1200 1400 1600 1800 2000
0
20
0 200 400 600 800 1000 1200 1400 1600 1800 2000
0
20S1L5a (n=3)
Spik
es/s
Time (ms)
64
Another difference is obvious when the responses of S and non-S cells in the first
cycle are compared in both conditions (Figure 5). The responses of S and non-S
neurons in S2 and S1 exhibit differential strengths in the first cycle. In S2L46, the
responses of S-cells are much stronger than those of non-S cells, but there is no
difference between S-cells and non-S cells in S2L23. Similar to S2L46, S-cells in
Figure 4. Dynamics of neuronal responses in S2, S1 and thalamus.
A, Average PSTHs of location-selective neurons during free air whisking (black) and object touch at selective location (red) across nuclei. Black bars indicate stimulus time within each cycle.B, Average PSTHs of non-selective neurons for whisking (black) and touch (magenta) at each
station.
B non-S cell
S1L4 (n=22)
0
25
50
0 200 400 600 800 1000 1200 1400 1600 1800 20000
20
40
0
30
60
0
20
40 S2L23 (n=27)
0
30
60
0
45
90
0
25
50
0
20
40S1L23 (n=28)
S1L5a (n=20)
S2L46 (n=13)
VPMvl (n=10)
VPMdm (n=24)
POm (n=23)
Spik
es/s
Time (ms)
Figure 4. Dynamics of neuronal responses in S2, S1 and thalamus.
A, Average PSTHs of location-selective neurons during free air whisking (black) and object touch at selective location (red) across nuclei. Black bars indicate stimulus time within each cycle.B, Average PSTHs of non-selective neurons for whisking (black) and touch (magenta) at each
station.
B non-S cell
S1L4 (n=22)
0
25
50
0 200 400 600 800 1000 1200 1400 1600 1800 20000
20
40
0
30
60
0
20
40 S2L23 (n=27)
0
30
60
0
45
90
0
25
50
0
20
40S1L23 (n=28)
S1L5a (n=20)
S2L46 (n=13)
VPMvl (n=10)
VPMdm (n=24)
POm (n=23)
Spik
es/s
Time (ms)
B non-S cell
S1L4 (n=22)
0
25
50
0 200 400 600 800 1000 1200 1400 1600 1800 20000
20
40
0
30
60
0
20
40 S2L23 (n=27)
0
30
60
0
45
90
0
25
50
0
20
40S1L23 (n=28)
S1L5a (n=20)
S2L46 (n=13)
VPMvl (n=10)
VPMdm (n=24)
POm (n=23)
Spik
es/s
Time (ms)
65
S1L23 and S1L5a display stronger responses than non-S cells at first cycle, while
there are weaker responses of S cells in S1L4 in both whisking and touch trials. In
thalamus, S-cells and non-S cells mostly show similar responses at the first cycle.
Accompanying with response propagation to their stereotypic patterns, a slow
response wave appears in non-S cell populations, but weak or none slow wave is
found in S cell populations (Figure 6). To quantify the slow wave component, a low-
pass filter with a cutoff frequency of 2Hz was used to isolate slow wave from the fast
response. For non-S cell populations in both whisking and touch conditions, the slow
wave is strongly exhibited in both S2L23 and S2L46, and also appeared in all three
cortical layers in SI, which is consistent with previous report (Derdikman et al.,
2006a). Interestingly, a clear slow wave in VPMvl and a weaker slow wave in
VPMdm are also observed, but no slow wave appears in POm; the VPM waves
peaked significantly later than S1 and S2 waves. For S cell populations, the weaker
slow waves are observed in S2L23 during whisking and touch conditions, but almost
no temporally-localized slow wave were found in S2L46, thalamic nuclei and SI
cortical layers in both conditions. Instead, slow frequency modulations were evident
all along the whisking train.
S cell in T
non-S cell in T
S cell in W
non-S cell in W
Figure 5. Comparison of responses of S-cells and non-S cells in first cycle during free air whisking and object touch at each station. S-cell responses in first cycle are indicated by black for free whisking and red for touch; non-S cell responses are marked by gray for free whisking and magenta for touch.
Spik
es/s
0
15
30
0
20
40
POm
0
15
30
0
40
20
VPMvl
S2L46
S2L23
0 20 40 60Time (ms) Time (ms)
0
15
30 VPMdm
0
20
40 S1L23
0
10
20 S1L4
40
0 20 40 600
20
S1L5a
Spik
es/s
S cell in T
non-S cell in T
S cell in W
non-S cell in WS cell in T
non-S cell in T
S cell in W
non-S cell in W
Figure 5. Comparison of responses of S-cells and non-S cells in first cycle during free air whisking and object touch at each station. S-cell responses in first cycle are indicated by black for free whisking and red for touch; non-S cell responses are marked by gray for free whisking and magenta for touch.
Spik
es/s
0
15
30
0
20
40
POm
0
15
30
0
40
20
VPMvl
S2L46
S2L23
0 20 40 60Time (ms) Time (ms)
0
15
30 VPMdm
0
20
40 S1L23
0
10
20 S1L4
40
0 20 40 600
20
S1L5a
Spik
es/s
Time (ms)
0
15
30 VPMdm
0
20
40 S1L23
0
10
20 S1L4
40
0 20 40 600
20
S1L5a
Spik
es/s
66
To quantify the progression of slow wave, the latency to the peak of slow wave is
computed for non-S cells. The latency varied across different cortical layers in SI and
SII, and thalamic nuclei. S2L23 and S2L46 slow-wave peaks exhibited similar
latencies in both whisking and touch conditions (p>0.200, non-parametric Mann-
whitney test), and both of them are similar to S1L4 and S1L5a (P>0.05, S1L4 vs
S1L5a; p>0.097, S2 vs S1, Mann-whitney test), but longer than S1L23 (p<0.03,
Mann-whitney test). Interestingly, the latencies of slow wave in VPMvl were shorter
than those in VPMdm, but the difference between them is not significant in both
conditions (P>0.087, Mann-whitney test), while both of them are significantly longer
Figure 6. comparison of slow wave responses between location-selective (S) cells and non-selective (non-S) cells in each nucleus.
A, Slow waves of non-S cell populations for both whisking (gray) and touch (magenta) cycles. Arrows denote the peaks of slow wave in both conditions (black for whisking, magenta for touch) B, Slow waves of S cell populations during whisking (black) and touch (red) cycles.
10
0
5
10 POm0
6
12 S2L46
0
5
10 VPMvl
0
5VPMdm
0 400 800 1200 1600 20000
5
10S1L5a
0
5
10 S1L230
5
10 VPMdm
0
5
10 S2L23
0
6
12 S2L46
0
5
10POm
0
5
10 VPMvl
0
5
10 S1L23
0
5
10 S1L4
0 400 800 1200 1600 20000
5
10S1L5a
A B S cell
0
5
10 S2L23
non-S cell
Time from train onset Time from train onset
Low
pas
s filt
er o
f spi
ke-r
ate
0
5
10 S1L4
Figure 6. comparison of slow wave responses between location-selective (S) cells and non-selective (non-S) cells in each nucleus.
A, Slow waves of non-S cell populations for both whisking (gray) and touch (magenta) cycles. Arrows denote the peaks of slow wave in both conditions (black for whisking, magenta for touch) B, Slow waves of S cell populations during whisking (black) and touch (red) cycles.
10
0
5
10 POm0
6
12 S2L46
0
5
10 VPMvl
0
5VPMdm
0 400 800 1200 1600 20000
5
10S1L5a
0
5
10 S1L230
5
10 VPMdm
0
5
10 S2L23
0
6
12 S2L46
0
5
10POm
0
5
10 VPMvl
0
5
10 S1L23
0
5
10 S1L4
0 400 800 1200 1600 20000
5
10S1L5a
A B S cell
0
5
10 S2L23
non-S cell
Time from train onset Time from train onset
Low
pas
s filt
er o
f spi
ke-r
ate
0
5
10 S1L4
10
0
5
10 POm0
6
12 S2L46
0
5
10 VPMvl
0
5VPMdm
0 400 800 1200 1600 20000
5
10S1L5a
0
5
10 S1L230
5
10 VPMdm
0
5
10 S2L23
0
6
12 S2L46
0
5
10POm
0
5
10 VPMvl
0
5
10 S1L23
0
5
10 S1L4
0 400 800 1200 1600 20000
5
10S1L5a
A B S cell
0
5
10 S2L23
non-S cell
Time from train onset Time from train onset
Low
pas
s filt
er o
f spi
ke-r
ate
0
5
10 S1L4
67
than cortical layers in S2 and S1 (p<0.007, Mann-whitney test). This latency
progression is exactly opposite to the initial progression of peripherally-originated
information within thalamocortical circuits: these slow waves advance from cortex to
thalamus.
Comparison of adaptation
Our previous study demonstrated that different cortical layers in S1 exhibit
different response dynamics which might indicate involvement in different processing
circuits (Derdikman et al., 2006a). In order to see whether S2 laminar modules and
thalamic nuclei also exhibit module-specific dynamics, and whether different modules
share similar dynamic patterns, we compared the response dynamics between S and
non-S cells in all these sites. For quantitative assessments, the spike count during the
first 60ms of each cycle was computed and normalized, such that the normalized
spike count at cycle i, SPNi was: SPn = (SPi – SP1)/ (SPi + SP1), where SPi is the
spike count in cycle i, and SP1 is the spike count in cycle 1. The progression of spike
counts of the entire population across cortical layers in whisking and touch cycles are
depicted in Figure 7. In S2L23 and S2L46, the facilitations from the first cycle to
steady state are evident for both S-cells and non-S cells in both whisking and touch.
Interestingly, S-cells in S2L23 show larger facilitations than non-S cells in both
whisking and touch conditions, and this is contrary to S-cells in S2L46, which show
less facilitation than non-S cells in both conditions. In S1, adaptations of non-S cell
populations are consistent with previous reports (Derdikman et al., 2006a), in which,
S1L23 responses are depressed and S1L5a responses are facilitated during both
whisking and touch conditions, S1L4 responses have no clear facilitations in touch
and whisking conditions. Compare to non-S cells, S cells in S1L23 exhibit stronger
depressions and S cells in S1L4 show stronger facilitations in both conditions, while
facilitations are completely inhibited for S cells in S1L5a. As comparison, both S2L23
and S1L4 show stronger facilitations for S cells and weaker or no facilitations for
non-S cells. Contrarily, S2L46 and S1L5a present less or no facilitation for S cells and
stronger facilitations for non-S cells. The output layer S1L23 shows strong depression
in both two populations. This phenomenon is not observed in S2. Interestingly, cells
in thalamic nuclei show quite different (complex) adaptation patterns with those in
the cortical nuclei, such as, the responses of S cells in all thalamic sites are stronger
facilitated than those of non-S cells during touch trials, which is similar to those in
68
S2L23 and S1L4. However, it is obviously opposite during free air whisking trials in
POm and VPMdm, which is similar to S2L46 and S1L5a. Although S cells and non-S
cells in VPMvl show similar facilitation pattern (difference) as those in S2L23 and
S1L4 during whisking trials, the facilitation of S cells in W trials don't show clearly
separation with those of non-S cells in T trials , which is distinct from those in S2L23
and S1L46. Briefly, there are two major patterns of response dynamics in cortex: in
S1L4 and S2L23 S cells facilitate more than non-S cells, while in S1L5a and S2L46
the opposite is observed; S1L23 might be related to the latter group despite its general
pattern of depression. Differently, these two patterns do not appear in the thalamic
nuclei.
Comparison of response latencies
So far, the serial versus parallel processing in the cortical somatosensory network is
still under debate. Specially, whether S2 is primarily activated by direct sub-cortical
Figure 7. comparison of adaptation across all nuclei. Dynamics of spike count of S cells and non-S cells in both whisking (black for S cells, gray for non-S cells) and touch (red for S cells, magenta for non-S cells) conditions.
Nor
mal
ized
spik
e-co
unt
-1
0
1
S2L46
-1
0
1
S2L23
-1
0
1
Pom
-1
0
1
VPMvl
-1
0
1
VPMdm
-1
0
1
S1L23
1
-1
0
S1L4
0 5 10-1
0
1
S1L5a
0 5 10Cycle # Cycle #
S cell in T
non-S cell in T
S cell in W
non-S cell in W
Figure 7. comparison of adaptation across all nuclei. Dynamics of spike count of S cells and non-S cells in both whisking (black for S cells, gray for non-S cells) and touch (red for S cells, magenta for non-S cells) conditions.
Nor
mal
ized
spik
e-co
unt
-1
0
1
S2L46
-1
0
1
S2L23
-1
0
1
Pom
-1
0
1
VPMvl
-1
0
1
VPMdm
-1
0
1
S1L23
1
-1
0
S1L4
0 5 10-1
0
1
S1L5a
0 5 10Cycle # Cycle #
Figure 7. comparison of adaptation across all nuclei. Dynamics of spike count of S cells and non-S cells in both whisking (black for S cells, gray for non-S cells) and touch (red for S cells, magenta for non-S cells) conditions.
Nor
mal
ized
spik
e-co
unt
-1
0
1
S2L46
-1
0
1
S2L23
-1
0
1
Pom
-1
0
1
VPMvl
-1
0
1
VPMdm
-1
0
1
S1L23
1
-1
0
S1L4
0 5 10-1
0
1
S1L5a
0 5 10Cycle # Cycle #
Nor
mal
ized
spik
e-co
unt
-1
0
1
S2L46
-1
0
1
S2L23
-1
0
1
Pom
-1
0
1
VPMvl
-1
0
1
VPMdm
-1
0
1
S1L23
1
-1
0
S1L4
0 5 10-1
0
1
S1L5a
0 5 10Cycle # Cycle #
S cell in T
non-S cell in T
S cell in W
non-S cell in WS cell in T
non-S cell in T
S cell in W
non-S cell in W
69
input or corticocortical input from S1 is not yet known during active whisking. To
address this issue, we compared the neuronal response latencies across thalamic, S1
and S2 cortical domains. During free-air whisking cycles, modal latencies (to half-
peak response) at steady-state in S2L23 (median=15.7ms) and S2L46 (15.0ms) are
shorter than those in S1L23 (17.0ms), S1L4 (16.6ms) and S1L5a (21.0ms), and longer
than those in VPMdm and VPMvl (6.3ms and 6.9ms), but similar to those in POm
(15.3ms) (Figure 8). To further verify the information flowing within thalamocortical
circuits, the onset latency to the first stimulus cycle was also examined. As seen in
Figure 9A and 9B, onset latencies in POm are bimodal distributed, in which one
group neurons respond before those in S1 and S2 cortical layers, the others respond
later. This bimodal distribution is consistent with previous report on POm neurons
using passive stimulation (Trageser and Keller, 2004), although the entire population
latencies are much shorter in our active whisking paradigm. As seen in Figure 9B-C,
the neuronal responses in S2L23 (median=11ms) and S2L46 (median=11ms) have
longer latencies than those in S1L4 (median=9ms), and similar to those in S1L23 and
S1L5a (median=10ms and 11ms, respectively) at the first stimulus cycle. Meanwhile,
neurons in S2L23 and S2L46 show longer latencies than those in VPMdm (4ms),
VPMvl (4ms) and POm (4ms in one group). Thus, the onset of S2 activity can not be
driven directly by the onset of S1 activity, but by thalamic activity during active
whisking.
A B
0 10 20 30 40 50 600
5
10
15
20
25
30
0 10 20 30 40 50 600
0.5
1
Time (ms)
Spik
es/s
Prob
abili
ty
Time (ms)
S2L23S2L46POmVPMdmVPMvlS1L23S1L4S1L5aS1
S2Thalamus
Figure 8. Comparison of response onset at steady states during free air whisking.
A, Average PSTHs of the entire population during steady states in S2L23 (magenta), S2L46 (red), POm (dark green), VPMvl (green, dash), VPMdm (light green), S1L23 (blue, dash), S1L4 (cyan) and S1L5a (dark blue).B, the distribution of response latencies at steady states indicated by cumulative probability plots indicating.
A B
0 10 20 30 40 50 600
5
10
15
20
25
30
0 10 20 30 40 50 600
0.5
1
Time (ms)
Spik
es/s
Prob
abili
ty
Time (ms)
S2L23S2L46S2L23S2L46POmVPMdmVPMvlS1L23S1L4S1L5a
S1L23S1L4S1L5aS1
S2Thalamus
Figure 8. Comparison of response onset at steady states during free air whisking.
A, Average PSTHs of the entire population during steady states in S2L23 (magenta), S2L46 (red), POm (dark green), VPMvl (green, dash), VPMdm (light green), S1L23 (blue, dash), S1L4 (cyan) and S1L5a (dark blue).B, the distribution of response latencies at steady states indicated by cumulative probability plots indicating.
70
Discussion Our findings suggest that the deep layers of the rat secondary somatosensory cortex
(S2) play a significant role in object localization. These results further suggest that S2
is not a unitary computational unit and that its superficial and deep layers are parts of
parallel processing streams; the superficial layers (S2L23) exhibit dynamics that is
similar to that exhibited by layer 4 barrels of primary sematosensory cortex (S1L4),
Figure 9. Comparison of response onset in the first cycle during free air whisking.
A, The distribution of response latencies in the first cycle during free air whisking. B, The distribution of response latencies of first cycle indicated by cumulative probability curves C, Average PSTHs of the entire population during the first cycle. Here, POmneurons were spitted into two groups: S-lat (short latencies, dark green, dash line) and L-lat (longer latencies, dark green)
B
A
0
0
0
6
12
No.
of c
ells
0 4 8 12 16 20 24 28 more
6
12S2L23
0
6
12S2L46
6
12POm
VPMvl
Latency from protraction onset0 4 8 12 16 20 24 28 more
Latency from protraction onset
0
6
12
0
6
12VPMdm
0
6
12S1L23
0
6
12S1L4
S1L5a
0 5 10 15 20 25 300
25
Time (ms)
Spik
es/s
C
0 5 10 15 20 25 300
0.2
0.4
0.6
0.8
1
Time (ms)
Prob
abili
ty
S2L23S2L46POm
VPMdmVPMvlS1L23S1L4S1L5a
Figure 9. Comparison of response onset in the first cycle during free air whisking.
A, The distribution of response latencies in the first cycle during free air whisking. B, The distribution of response latencies of first cycle indicated by cumulative probability curves C, Average PSTHs of the entire population during the first cycle. Here, POmneurons were spitted into two groups: S-lat (short latencies, dark green, dash line) and L-lat (longer latencies, dark green)
B
A
0
0
0
6
12
No.
of c
ells
0 4 8 12 16 20 24 28 more0 4 8 12 16 20 24 28 more
6
12S2L23
0
6
12S2L46
6
12POm
6
12POm
VPMvl
Latency from protraction onset0 4 8 12 16 20 24 28 more
Latency from protraction onset
0
6
12
0
6
12VPMdm
0
6
12S1L23
0
6
12S1L4
S1L5a
0
6
12
0
6
12VPMdm
0
6
12S1L23
0
6
12S1L4
0
6
12S1L4
S1L5a
0 5 10 15 20 25 300
25
Time (ms)
Spik
es/s
C
0 5 10 15 20 25 300
0.2
0.4
0.6
0.8
1
Time (ms)
Prob
abili
ty
S2L23S2L46POm
VPMdmVPMvlS1L23S1L4S1L5a
S2L23S2L46POm
VPMdmVPMvlS1L23S1L4S1L5a
71
whereas the deep layers (S2L46) exhibit dynamics that is similar to that exhibited by
layer 5a of S1 (S1L5a). Location-selective cells where not distributed evenly across
the somatosensory stations studied; the highest proportion (41%) was found in S2L46,
the lowest (4%) was found in POm , and proportions around 20% were found at the
rest. Location-selective cells exhibited a lack of a slow-wave dynamics that differed
from that of non-selective cells in all stations (no slow-waves in POm), and a stronger
response to the first cycle in S2L46, S1L5a and S1L23 (Table 1).
Functional segregation within S2
Unlike S1 cortex, which shows a unique anatomic-functional organization and have
been studied intensively over the years, little is known about the functional
organization within S2, and very few physiological studies have examined the
response properties in S2 (Kwegyir-Afful and Keller, 2004; Melzer et al., 2006). In
those studies, no differences were found between layers within S2 using passive
paradigm, neurons in all the layers exhibit consistent response properties, such as the
receptive field, latencies, angular tuning, kinetics, etc. Using active whisking, we
found here a clear functional segregation between the superficial and deeper layers of
S2. S2L23 and S2L46 represent diverse response properties. Neurons in S2L46
exhibit higher ability for coding object locations with a specific selectivity. S and
non-S Cells in S2L23 and S2L46 exhibit different response strength in the first cycle,
Table 1 comparison of location-selective cells and non-location-selective cells across all nuclei.
a a weaker slow wave dynamics in S cells of S2L23b a depressive process in S1L23 Y, the presence of slow wave; N, the absence of slow wave.~=, similar; >, stronger; <, weaker
Nuclei Response Strength of 1st cycle
adaptation strength
S cell nS cell S vs nS S vs nS
Slow Wave
S2L23 Ya
Y ~= >
S2L46 N Y > <
Pom N N ~= ~=
VPMvl N Y ~= ~=
VPMdm N Y ~= ~=
S1L23 N Y > <b
S1L4 N Y < >
S1L5a N Y > <
Table 1 comparison of location-selective cells and non-location-selective cells across all nuclei.
a a weaker slow wave dynamics in S cells of S2L23b a depressive process in S1L23 Y, the presence of slow wave; N, the absence of slow wave.~=, similar; >, stronger; <, weaker
Nuclei Response Strength of 1st cycle
adaptation strength
S cell nS cell S vs nS S vs nS
Slow Wave
S2L23 Ya
Y ~= >
S2L46 N Y > <
Pom N N ~= ~=
VPMvl N Y ~= ~=
VPMdm N Y ~= ~=
S1L23 N Y > <b
S1L4 N Y < >
S1L5a N Y > <
Nuclei Response Strength of 1st cycle
adaptation strength
S cell nS cell S vs nS S vs nS
Slow Wave
S2L23 Ya
Y ~= >
S2L46 N Y > <
Pom N N ~= ~=
VPMvl N Y ~= ~=
VPMdm N Y ~= ~=
S1L23 N Y > <b
S1L4 N Y < >
S1L5a N Y > <
72
and distinctive patterns of adaptation, which resembled those appeared in S1L4 and
S1L5a, respectively.
Anatomical and physiological studies report that S2 receives dense thalamic input
from VPMvl and POm, which convey extralemnical and paralemnical signals,
respectively, and to a lesser extent, also from the VPMdm, which conveyes leminical
signals (Carvell and Simons, 1987; Spreafico et al., 1987; Koralek et al., 1988;
Alloway et al., 2000; Pierret et al., 2000). Moreover, studies show that S2 receives
projections from both the barrel- and the septa-related circuits of S1 cortex
(Hoeflinger et al., 1995; Kim and Ebner, 1999; Hoffer et al., 2003; Chakrabarti and
Alloway, 2006).These two circuits convey different types of information related to
whisking and touch, involving in lenmical and paralemniscal pathways separately.
Consistent with these multiple inputs, studies under passive whisking show that
neurons in S2 exhibit large receptive fields similar to cells in POm and S1 septa, as
well as angular preferences and representations of frequency modulations that are
similar to those exhibited by neurons in VPM and S1 barrels (Diamond et al., 1992a;
Ahissar et al., 2000; Bruno et al., 2003; Timofeeva et al., 2003; Kwegyir-Afful and
Keller, 2004; Lee and Simons, 2004; Melzer et al., 2006). Although it is unclear
whether these inputs project onto different layers of S2, both thalamocortical and
corticocortical connections suggest that different schemes related to whisker
information could be conducted in S2.
It is not yet known whether the S2L23 and S2L46 receive the input components
from separate brain regions and distribute it to different brain areas, except that an
anatomical tracing study indicated that S2L46 receive robust input from both VPMvl
and POm (Pierret et al., 2000). However, the distinction we found here suggests that
the S2L23 and S2L46 differentially process whisker-related information and imply
specific computational and behavioral functions. S1L23 might mostly involve in
controlling the kinematics of whisker movements, such as frequency, angle, velocity,
etc, and transmitting sensory information to M1 or other higher brain area engaged in
motor control of whisker movement, be a part of the paralemniscal system, to some
extent, combining with lemniscal system. More effective location-selective cells are
seen in the S2L46, which may reflect the fact that POm and VPMvl together are
involved in object localization task, the integration of the POm reference signal and
VPMvl contact signal could take place in S2L46, since S2L46 is the major target of
POm and VPMvl projection (Kleinfeld and Delaney, 1996; Kwegyir-Afful and Keller,
73
2004; Yu et al., 2006). The specific coding characteristic of S cells in S2L46, may
also reflect some kind of functional interaction of thalamocortical (Carvell and
Simons, 1987; Spreafico et al., 1987; Koralek et al., 1988; Alloway et al., 2000;
Pierret et al., 2000). and corticocortical processing (Carvell and Simons, 1987; Fabri
and Burton, 1991; Hoeflinger et al., 1995; Kim and Ebner, 1999; Hoffer et al., 2003;
Chakrabarti and Alloway, 2006), which cause active selection of relevant touch
information occur in S2L46 during processing of whisker-related information. Thus,
S2L46 might partially present converged responses of paralemniscal and
extralemniscal systems, involve in processing information related to object
localization.
The functional specificity of cortical areas depends primarily on afferent and
efferent connections. The strength difference between S cells and non-S cells in each
nucleus may reflect differences in the origin of thalamic inputs for different cortical
regions (Koralek et al., 1988; Diamond et al., 1992b; Lu and Lin, 1993; Pierret et al.,
2000) (Chmielowska et al., 1989; Ahissar et al., 2000) , and difference in the efficacy
or number of thalamic inputs for different type of neurons within each region
(Simons and Carvell, 1989; Brumberg et al., 1999; Bruno and Simons, 2002; Pinto et
al., 2003). In addition, the intra-cortical inhibition may be also a responsible source
for the strength difference (Keller and White, 1987; Kyriazi et al., 1994; Brumberg et
al., 1999; Pinto et al., 2003; Schubert et al., 2007). Thus, the responses of cortical
neurons may be determined by the interplay between direct thalamocortical excitation
and strong, locally generated cortical inhibition (Miller et al., 2001) Previous studies
showed that there are more regular-spiking than fast-spiking neurons in S2, which
suggest that less intracortical inhibition exists in S2 than that in S1 (Kwegyir-Afful
and Keller, 2004; Melzer et al., 2006). The fact that neurons in both S2L23 and
S2L46 are strongly facilitated during repetitive stimulation may somewhat reflect less
pronounced inhibition in S2.Thus, the inputs for all cortical areas could be from the
same type of thalamic neurons, but different cortical area has different thalamic
source, and might have different inhibition strength on different type of cells, this
might cause response strength difference in different cortical region. Therefore,
stronger responses of S cells in S2L46 and S1L5a might because those neurons
receive more efficient inputs from either the combination of paralemniscal (POm)
and extralemniscal (VPMvl) pathways or paraleminscal pathways alone, and has less
intra-cortical inhibition (S1L5a receives strongly innervations from excitatory
74
neurons of S1L4 (Feldmeyer et al., 2005)). Whereas, S1L4 barrel neurons are driven
directly by thalamocortical afferent fibers from VPMdm (Chmielowska et al., 1989;
Lu and Lin, 1993), and contain numerous local inhibitory neurons (Keller and White,
1987; Armstrong-James et al., 1991) and stronger surround inhibition evoked by
adjacent whisker stimuli (Simons and Carvell, 1989). Thus, thalamic input of lemnical
pathway with more pronounced intro-cortical inhibition may cause weaker response
in S cells than those in non-S cells in S1L4. The anatomical connections of S2L23 are
yet unclear. Probably, S cells and non-S cells in S2L23 receive thalamic and cortical
inputs from paraleminiscal, to some extent, may also from lemniscal pathway
(Hoeflinger et al., 1995; Kim and Ebner, 1999; Pierret et al., 2000; Kwegyir-Afful and
Keller, 2004; Chakrabarti and Alloway, 2006), the combination of paralemnical and
leminical inputs with similar inhibition, may cause similar response strength in
S2L23. Since the output layer S1L23 has stronger reciprocal connections with S1L5a
in spite of receiving inputs from both S1L4 and S1L5a (Bernardo et al., 1990;
Gottlieb and Keller, 1997; Kim and Ebner, 1999; Bureau et al., 2006; Petreanu et al.,
2007), it might be more related to S1L5a and show similar response strength as S1L5a
and S2L46.
Parallel cortical processing of active touch
It is still ambiguous that whisker-related information is serially or parallely
processed in the primary and secondary cortex of the rat. A series of anatomical
evidences show that S2 receive thalamic projection from both VPM and POm in
parallel to S1 (Carvell and Simons, 1987; Spreafico et al., 1987; Koralek et al., 1988;
Alloway et al., 2000; Pierret et al., 2000; Brett-Green et al., 2003) . S2 and S1 are
reciprocal and topographically interconnected (Carvell and Simons, 1987; Fabri and
Burton, 1991; Hoeflinger et al., 1995; Kim and Ebner, 1999; Chakrabarti and
Alloway, 2006). Here, we provide electrophysiological evidence for a separate
sensory pathway for cortical processing of object location. Previous studies in the
thalamus suggest that S2 could be optimally tuned for object localization since it
receives majority of projections from POm and VPMvl, which convey the reference
signal and contact signal of object location, respectively (Pierret et al., 2000; Yu et al.,
2006). Indeed, our observations here demonstrate that S2L46 contains a larger
proportion of neurons with strong location selectivity, which indicate that S2L46
neurons could extract selective information of object locations and perform a distinct
75
coding scheme. Comparably, this selective coding scheme was not obviously seen in
S1. This may reflect the fact that S1L4 barrels mainly receive thalamic inputs through
lemnical system processing information related to object identification during active
touch, and S1L5a mainly process thalamic input trough paralemincal system related to
whisker motion control (Derdikman et al., 2006a; Yu et al., 2006). Thus, the ability
and specificity for encoding object location in S2L46 suggest that S2 play
functionally distinct roles in the processing of active touch, largely independent to the
processing performed in S1.
During active touch, response properties in S2 exhibit noticeably similarities with
those in S1. Response latencies in S2 are comparable to those in S1 layers in the first
responsive cycle and later steady states under our active paradigm, which are
consistent with previous report from passive whisking (Kwegyir-Afful and Keller,
2004). S2 and S1 latencies are significantly longer than those of their thalamic inputs
(Ahissar et al., 2000; Kwegyir-Afful and Keller, 2004). Combining with anatomical
and electrophysiological recordings (Carvell and Simons, 1987; Spreafico et al., 1987;
Koralek et al., 1988; Alloway et al., 2000; Pierret et al., 2000; Brett-Green et al.,
2003; Kwegyir-Afful and Keller, 2004), our results here further reveal that the onset
of S2 activity is driven directly by their thalamic counterpart during active whisking,
parallel to those in S1. In addition, the timing of slow wave peak of cells in both
S2L23 and S2L46 is statistically similar to those observed in S1 input layers: S1L4
and S1L5a, which are engaged in leminical and paraleminscal systems, respectively.
Moreover, the pattern of response strength in the first cycle is similarly observed in
S2L46 and S1L5a, and two different patterns of response adaptation are clearly seen
in both S2 and S1, one pattern is obvious in S2L46 and S1L5a, the other appears in
S2L23 and S1L4. Although the functional role of these specific properties in the
processing of active touch is yet unknown, these similarities between S2 and S1
cortical input layers further suggest that S2 may contribute to tactile behavior in an
independent way with S1. In consistent with a parallel processing scheme, our
observations here suggest that sensory information relate to touch are processed
concurrently by S1 and S2 during active whisking.
Summary
A large portion of neurons (41%) in deeper layers of S2 (S2L46) preferentially
respond to one of object locations, while small portion of neurons (19%) in superficial
76
layers (S2L23) show such location preference. Comparatively, less location-selective
neurons (<23%) are found in POm, VPMdm, VPMvl and S1 cortical layers. Focusing
on response dynamics, we find that the transient responses to the steady states within
first few cycles are accompanied by strong slow wave in non-location-selective
neuron populations during free air whisking and object touch across all the nuclei.
Differently, slow wave is not seen in location-selective neuron populations except
those in S2L23. Comparing the first cycle responses, we find that S neurons display
much stronger response than non-S neurons in S2L46, in analogy with S1L23 and
S1L5a, but no strength difference was observed in S2L23 and thalamic nuclei. In
addition, we show that S neurons present weaker facilitation than non-S neurons in
S2L46, but display stronger facilitation than non-S neurons in S2L23. This strength
difference is also observed within SI cortical layers: weaker facilitation for S neurons
in S1L5a and stronger facilitation for S neurons in S1L4. Overall, these differences
between S2L23 and S2L46 and between S2, SI and thalamus suggest that neurons in
different layers of S2 could process tactile information using different coding schemes
during active whisking, cortical processing of active touch could be conveyed in
parallel in at least two streams, and a specific processing of object location is
performed in S2L46.
77
DISCUSSION
In this thesis, I showed that the major active-touch signal conveyed in each of the
three afferent pathways of the whisker system is different: whisking in the
paralemniscal (via POm), contact in the extralemniscal (via VPMvl), and combined
whisking-touch in the lemniscal (via VPMdm) pathway. Accordingly, POm whisking
cells can encode whisking frequency by both spike timing and spike rate; VPM Touch
cells can encode horizontal object location by spike timing; Whereas Whisking-Touch
cells convey complex information possibly related to coding of object identity.
Furthermore, I found that the secondary somatosensory cortex (S2), which is the
major target of both VPMvl and POm, plays a significant role in object localization in
parallel to S1. The deep layers (S2-L46) contain high proportion location selective
cells that are clearly distinctive from its superficial layers as well as other thalamic
and cortical nuclei. On the other hand, The S2 deep and superficial layers exhibited
different response dynamics, indicating that S2 is not a unitary computational unit and
that its superficial and deep layers are parts of parallel processing streams.
Parallel thalamic and cortical processing of active touch My results in the thalamus indicated that the three thalamic afferent pathways did
not respond synchronously. In each whisking cycle, VPMdm neurons, conveying the
combined signal, fired first while POm and VPMvl neurons, conveying isolated
whisking and touch signals, fired later. VPMdm also contained tonic responses that
were absent in the other nuclei. All these observations, together with the known
anatomy and physiology of the system, suggest that VPMdm responses did not result
from a combination of signals transmitted by the POm and VPMvl. Moreover, the
orthogonal response types of POm (W) and VPMvl (T) indicate that each of the three
thalamic nuclei conveys a signal that could not result from a combination of signals
transmitted by the other two nuclei. This, and the fact that similar response types (W,
T, and WT) are exhibited by different classes of TG neurons, strongly suggests that
the three thalamic nuclei are driven primarily by their afferent pathways. (W-T
signals, whose latencies do allow intra-thalamic inhibition and/or cortical feedback
(Ahissar and Arieli, 2001) as primary drivers, might be an exception. Thus, our results
suggest parallel thalamic processing.
78
It is still under debate whether whisker-related information is processed in a serial
or parallel fashion in the primary and secondary cortex of the rat. Here, I provide
electrophysiological evidence for a separate sensory pathway for cortical processing
of object location. My observations demonstrate that S2L46 contains a large
proportion of location-selective neurons, which indicate that S2L46 neurons might be
actively involved in the decoding of object location. In contrast, S1 contained
significantly less location-selective neurons, which might indicate a smaller
involvement of S1 in object localization. This may reflect the fact that S1L4 barrels
mainly receive thalamic inputs through lemniscal system, which presumably
processes information related to object identification during active touch, and S1L5a
mainly process thalamic input of the paralemniscal system, presumably related to
whisker motion control (Derdikman et al., 2006a; Yu et al., 2006). Thus, the ability
and specificity for encoding object location in S2L46 suggest that S2 plays specific
roles in the processing of active touch, largely independent to the processing
performed in S1.
During active touch, response properties in S2 exhibit noticeable similarities with
those in S1. Response latencies in S2 are comparable to those in S1 in the first
responsive cycle as well as in later, steady states cycles. This is consistent with a
previous report using passive stimuli (Kwegyir-Afful and Keller, 2004). S2 and S1
latencies are significantly longer than those of their thalamic inputs (Ahissar et al.,
2000; Kwegyir-Afful and Keller, 2004). Combined with anatomical and
electrophysiological recordings (Carvell and Simons, 1987; Spreafico et al., 1987;
Koralek et al., 1988; Alloway et al., 2000; Pierret et al., 2000; Brett-Green et al.,
2003; Kwegyir-Afful and Keller, 2004), my results further reveal that the onset of S2
activity is driven directly by their thalamic drivers during active whisking, parallel to
those in S1. In addition, the timing of slow wave peak of cells in both S2L23 and
S2L46 is statistically similar to those observed in S1 input layers S1L4 and S1L5a,
which are engaged in leminical and paraleminscal systems, respectively. Moreover,
the pattern of response strength in the first cycle is similarly observed in S2L46 and
S1L5a, and two different patterns of response adaptation are clearly seen in both S2
and S1, one pattern is obvious in S2L46 and S1L5a, the other appears in S2L23 and
S1L4. Although the functional role of these specific properties in the processing of
active touch is yet unknown, these similarities between S2 and S1 cortical input layers
further suggest that S2 may contribute to tactile behavior in a way that is independent
79
to S1. Consistent with a parallel processing scheme, my observations here suggest that
sensory information related to touch are processed concurrently by S1 and S2 during
active whisking.
Sensory motor loops In active sensation, sensory information is acquired via movements of sensory
organs; rats move their whiskers repetitively to scan the environment, thus detecting,
localizing, and identifying objects. Sensory information is then processed by an
interconnected series of neuronal circuits in the brain, which eventually affects future
motor movements. Anatomical studies have shown that the circuitry involved in the
vibrissae sensorimotor system is configured as a series of feedback loops that form a
closed topology at the level of brainstem up through loops that close at the level of the
neocortex (reviewed in (Kleinfeld et al., 1999; Kleinfeld et al., 2006)). Such scheme
of information flow contrasts the classical notion of “bottom-up” and “top-down”
processes with a different point of view. Bottom-up and top-down streams are two
components of closed-loop systems. There is no one bottom-up and one top-down
system, but rather multiple such streams organized within a net of nested loops
comprising a sensory-motor system. The different loops interact by interconnections
between them.
The brainstem sensorimotor loop is the first point of interaction between sensory
input and motor output. It involves input that is relayed by the trigeminal ganglion to
trigeminal nuclei that in turn project to the facial motor nucleus, which control motion
of vibrissae. The brainstem loop is a positive-feedback loop participating in rhythmic
pattern generation (Nguyen and Kleinfeld, 2005). The fact that the patterned
activation of whisking can operate without patterned inputs from higher brain centers
or sensory feedback (Lovick, 1972; Gao et al., 2003), has risen the searching for a
sub-cortical central pattern generator (CPG) within the brainstem loop.
In addition, there are several higher-order sensorimotor loops, such as midbrain,
thalamic and cortical loops, which provide feedback pathways parallel to the pathway
from the trigeminal nuclei to the facial nucleus.
Neurons of the superior colliculus (SC) mediate the midbrain loop in the vibrissal
system, receiving input from trigeminal nuclei and projecting to the vibrissa
motoneurons (Kleinfeld et al., 1999). In addition, The SC neurons receive different
80
direct efferents from various areas, such as S1 and S2, the motor cortex (M1),
cerebellum and basal ganglia (Miyashita et al., 1994; Miyashita and Mori, 1995;
Hattox et al., 2002; May, 2005). They have been shown to play a major role in
orientating behavior, including control of the motion of relevant sensory and body
organs. However, the contribution of SC neurons in the interplay of sensory inputs
and motor commands is still unknown.
The functional segregation I reported in this thesis between three thalamic
pathways, and the evidence indicating that these three pathways close the sensory-
motor loop at different levels of brain hierarchy, raise the following sensory-motor
hypothesis: the paralemniscal system is involved in a basic motor-sensory-motor loop
that controls whisking velocity and frequency in a servo-like manner (Wiener, 1949),
the extralemniscal system adds a higher-level of control based on contact information
and object location, and the lemniscal system adds the highest level of control so far,
which is based on information related to object identity. This proposed functional
segregation does not imply functional isolation; these parallel loops are expected to
interact, such that a higher loop uses, and builds upon, the processing performed by a
lower loop. For example, the paralemniscal loop might interact with the brainstem
loop to optimize whisking control. Another example for such interaction is object
localization, where contact timing (extralemniscal) must interact with whisking
information (paralemniscal) to extract object location. Analysis of object-identity
requires interaction of detailed spatial information with information about whisker
movement and contact (Moore, 2004; Arabzadeh et al., 2005). The high-resolution
directional-selective spatial information (Minnery et al., 2003; Timofeeva et al., 2003)
together with whisking information (WT signals) conveyed by the VPMdm meet this
requirement. Object-identity analysis also involves comparisons with memorized
patterns; hence, it requires significant cortical involvement (Guic-Robles et al., 1992;
Hawkins and Blakeslee, 2004), such as that exhibited by the lemniscal system. Thus,
the paralemniscal, extralemniscal, and lemniscal parallel loops may have evolved
sequentially, as suggested for parallel sensory pathways (Bishop, 1959), by adding
contact detection to movement control, and identity analysis to contact detection
(Figure 1).
These three thalamic nuclei send projections to vibrissae related S1 and S2 cortices
that in turn project to M1. The sensorimotor loops are closed by descending
projections from M1 either directly to facial motor nucleus (Grinevich et al., 2005) or
81
indirectly through SC transformation (Miyashita and Mori, 1995). In addition, S1 and
S2 could also directly project to SC and transformed to the facial nucleus, completing
loops in a different way.
When whiskers move, data acquisition is performed actively at all levels, with
internal expectations being compared with input data, then utilizing in further
modulation of the motor pattern, thereby closing the loops. It is highly probable that
each of these loops contains an active element generating expectations about the
incoming signal at its specific level. With these active closed-loop processes, the
further exploration of the control of whisker movements and sensorimotor integration
(Kleinfeld et al., 2002; Hattox et al., 2003; Brecht et al., 2004; Haiss and Schwarz,
Figure 1. Hierarchy of nested sensory-motor loops.Sensory signals related to active vibrissal touch are conveyed by multiple pathways trough the sensory regions of the brain, and feed-back to motor circuits. In the thalamus, Whisking signals (W) conveyed by the paralemniscal pathway via the POm and involve whisking control; Contact signals (T) conveyed by the extralemniscal pathway via VPMvl and involve processing of object location; and combined whisking-touch signals (WT) conveyed by the lemniscal pathway via VPMdm and involve processing of object identity.
BrainstemLoop
MotorSe
nsor
y
FacialNucleus
-
Zona Incerta
extr
alem
nisc
al
lem
nisc
al
para
lem
nisc
al
VLThalamicNuclei
Thalamus
Cortex
TrigeminalGanglion
PrimaryMotor Cortex
+
RedNucleus
ReticularFormation
Brainstem Reticular
Nucleus
Pontine
Secondary
SuperiorColliculus
PrimarySensory Cortex
VibrissaeTrigeminalGanglion
TrigeminalNuclei
VPM-vl
VPM-dmPOm
CerebellaryOlivary
Figure 1. Hierarchy of nested sensory-motor loops.Sensory signals related to active vibrissal touch are conveyed by multiple pathways trough the sensory regions of the brain, and feed-back to motor circuits. In the thalamus, Whisking signals (W) conveyed by the paralemniscal pathway via the POm and involve whisking control; Contact signals (T) conveyed by the extralemniscal pathway via VPMvl and involve processing of object location; and combined whisking-touch signals (WT) conveyed by the lemniscal pathway via VPMdm and involve processing of object identity.
BrainstemLoop
MotorSe
nsor
y
FacialNucleus
-
Zona Incerta
extr
alem
nisc
al
lem
nisc
al
para
lem
nisc
al
VLThalamicNuclei
Thalamus
Cortex
TrigeminalGanglion
PrimaryMotor Cortex
+
RedNucleus
ReticularFormation
Brainstem Reticular
Nucleus
Pontine
Secondary
SuperiorColliculus
PrimarySensory Cortex
VibrissaeTrigeminalGanglion
TrigeminalNuclei
VPM-vl
VPM-dmPOm
CerebellaryOlivary
BrainstemLoop
MotorSe
nsor
y
FacialNucleusFacialNucleus
-
Zona IncertaZona Incerta
extr
alem
nisc
al
lem
nisc
al
para
lem
nisc
al
VLThalamicNuclei
VLThalamicNuclei
Thalamus
Cortex
TrigeminalGanglion
PrimaryMotor Cortex
+
RedNucleus
RedNucleus
ReticularFormation
Brainstem Reticular
Formation
Brainstem Reticular
Nucleus
PontineReticular
Nucleus
PontineReticular
Nucleus
Pontine
SecondarySecondary
SuperiorColliculusSuperiorColliculus
PrimarySensory Cortex
PrimarySensory Cortex
VibrissaeTrigeminalGanglion
TrigeminalGanglion
TrigeminalNuclei
TrigeminalNuclei
VPM-vl
VPM-dm
VPM-vl
VPM-dmPOmPOm
CerebellaryOlivary
CerebellaryOlivary
82
2005; Cramer et al., 2007)) is crucial in the understanding of active sensory
processing. Recent study in behavior rats suggests that the whisking pattern
generation is under strong feedback control has important implications from
understanding the nature of the signals reaching upstream neural processes
(Mitchinson et al., 2007).
Coding of horizontal object location If whiskers did not move, a stimulus could be defined in space simply according to
the identity of the activated whisker receptors and determined by receptor density.
Temporal variables would carry only information about the temporal changes of the
stimulus. In active sensing, the position of the whiskers changes over time. Thus,
temporal variables are crucial and can represent spatial characteristics of the stimulus,
such as relative location. To interpret sensory input, the rat needs to know not only
which whiskers have touched the object, but also when these whiskers touched, i.e.
what was the position of the whiskers at the moment of touch. The resulting temporal
information has to be subsequently extracted at higher brain stations.
The location of an object in 3-dimensions can be defined according to head and
whisker centered coordinate systems by three orthogonal axis: horizontal (i.e., its
location along the anterior-posterior axis, parallel to the whisker rows), radial (i.e.,
from the snout outward) and vertical (i.e., parallel to whisker arcs). Studies in TG
indicated that object location is encoded by different neuronal variables (codes) along
these three dimensions: the horizontal coordinate is encoded by timing of Contact
cells (temporal code), the radial coordinate by spike count of Touch cells (rate code),
and the vertical coordinate by whisker identity (spatial code).
Rats can accurately localize objects along the horizontal axis with just a single
whisker intact (Knutsen et al., 2006; Mehta et al., 2007). Thus, contact and
proprioceptive, but not spatial, information is available for the animal to process. The
temporal encoding of horizontal object location must rely on a comparison between
contact-time signal and a whisker-position reference signal. Two such reference
signals are known that may interact with a contact signal in order for a read-out circuit
to decode the temporal code of horizontal location: the timing of whisker protraction
onset and the whisker position at the time of contact. Such signals are conveyed by
the primary afferent Whisking cells (Szwed et al., 2003), and exist in both the
83
somatosensory thalamus (Derdikman et al., 2006b) and barrel cortex (Fee et al., 1997;
Brecht, 2006; Crochet and Petersen, 2006; Derdikman et al., 2006b; Golomb et al.,
2006). The exact location and identity of the neurons that implement the read-out
circuit for the temporal code of horizontal location is not yet known.
Previous research in TG (Szwed et al., 2003) suggest that horizontal object position
could be encoded by (i) coincident firing of individual Whisking and Contact cells,
and (ii) the temporal interval between firing onset of Whisking cells and firing of
Contact cells. These data suggest there are two specific encoding-decoding schemes
for horizontal object position in the vibrissal system. The spatiotemporal encoding
scheme, which involves a coincidence/phase detector circuit, suggested that thalamic
VPM neurons provide a spatial code of horizontal object location, implicating that
VPM cells are selective for one specific object location and have no phase coding.
The temporal scheme suggested that horizontal object position is encoded by the spike
count of thalamic POm neurons. However, I found here that individual thalamic
Touch cells precisely encode the horizontal object location by their first-spike timing
similarly to TG cells, which is not consistent with the spatiotemporal encoding
scheme for the VPM. On the other hand, POm neurons mainly convey whisker-
motion signals during active touch, which does not support the temporal encoding
scheme. Although our thalamic results do not support these two working models, such
models still could occur later in downstream brain areas. For example, VPMvl mainly
contains Touch neurons and POm mainly conveys Whisking signals. As in any
temporal coding scheme, the timing of response events must be measured with respect
to some reference event. In the case of horizontal location coding, this reference event
could be the onset of whisker protraction onset. Thus, the horizontal coordinate of an
object could be reliably read out by comparison between precisely timed reference
and contact signals in a site integrating POm and VPMvl outputs, such as S1, S2 or
M1. My results from S2 recordings show that the neurons in deeper layers have
selective responses to object locations in the horizontal coordinate, suggesting that S2
might be involved in the read out of spatial information, to some extend match the
spatiotemporal scheme.
For integration of information related to three different codes, central circuits of
the vibrissal system should translate the language of some of the channels to the
language of the others. To control movement, sensory codes should, at some stage, be
translated to motor codes. All these requirements could probably be satisfied by a
84
single transformation: temporal to rate. This is because spatial and rate codes are
already coded by firing rates, and motor control also appears to be based on
(population) rate coding (Kleinfeld et al., 1999; Salinas et al., 2000; Wessberg et al.,
2000; Zhang and Barash, 2000; Berg and Kleinfeld, 2003a; Carmena et al., 2003;
Ahrens and Kleinfeld, 2004; Brecht et al., 2004; Haiss and Schwarz, 2005). Using
passive stimuli, I found that such a temporal-to-rate transformation is implemented in
one of the ascending sensory pathways – the paralemniscal one (Ahissar et al., 2000).
A temporal-to-rate code transformation can be implemented by several circuits, one of
them is a neuronal phase-locked loop (NPLL) (Ahissar and Vaadia, 1990; Ahissar et
al., 1997; Ahissar, 1998; Zacksenhouse and Ahissar, 2006). A non-intuitive critical
prediction of an NPLL is that response latency should increase with the stimulus
frequency (Ahissar et al., 1997; Ahissar, 1998; Kleinfeld et al., 1999). With passive
stimuli in whisking-range frequencies (5–11 Hz) (Ahissar et al., 2000; Sosnik et al.,
2001), I found that this is indeed the case in POm, but not in VPM. This suggests that
NPLLs might be implemented in thalamocortical loops of the paralemniscal system
(Ahissar and Arieli, 2001; Moore, 2004), or in other loops that connect the POm with
an oscillatory source. However, frequency-dependent latency shifts could also be
caused by other mechanisms. In particular, thalamic GABAB inhibition, when
combined with specific brainstem activation profiles, could induce a similar
phenomenon (Golomb et al., 2006). Whatever the underlying mechanism is, a
temporal-to-rate code translation would be instrumental for a sensory-motor loop
controlling whisking kinematics. Indeed, our results here suggest that the POm
participates in such a loop. Using active whisking, I found that POm neurons
responded consistently and reliably to whisking movements, irrespective of whether
the whiskers contacted on an object or not. The response latency increased with the
stimulus frequency increment. Overall, my findings suggest that POm is involved in
temporal processing related to sensory-motor control of whisker movement, possibly
using NPLL-like mechanisms.
Artificial whisking The mechanisms of sensory processing in the rat vibrissal system are derived
primarily from studies of passive (mechanical) movements of the whiskers. During
passive stimuli, the object moves to stationary whisker, forces are applied only to a
85
single point of the whisker's shaft. However, these passive stimuli become less
relevant when the rat is actively whisking. In nature, rat actively moves their whiskers
to perform a wide variety of tasks, such as discriminate textures (Carvell and Simons,
1990) and localize objects (Brecht et al., 1997; Harvey et al., 2001). Whiskers are
moved back and forth by their muscles. Our artificial whisking paradigm mimics that
which occurs in nature: the muscle force acts on the base of the whisker, while the
object reaction acts on a point along its external shaft. The combination of these two
forces causes the whisker to curve while hitting the object. As a result, active whisker
touch is often accompanied by appreciably curving of the whisker shaft (Knutsen et
al., 2005). The curvature that develops during active touch represents the moment at
the base of the whisker (Solomon and Hartmann, 2006). Such curvatures never
develop with passive stimulations, as these stimulations are applied only to the
external shaft of the whisker. Furthermore, passive stimulations induce unnatural
whisker-follicle interactions, which mechanoreceptors are extremely sensitive to;
minute (< 0.1 deg) passive deflections may elicit significant sensory responses
(Gibson and Welker, 1983a). As a result, passive stimulations activate almost all
afferent neurons indiscriminately (Szwed et al., 2003). Our previous study in TG
indicated that responses of a given neuron to passive stimuli provide only limited
information about its response in the active whisking. During active touch, TG
neurons are primarily characterized by their selectivity to the components of active
touch; the observed selectivity could not be inferred from the responses of the same
cells to passive stimuli. Moreover, even the polarity of responses (i.e., response to
contact or to detachment) in active touch could not be predicted from their directional
selectivity with passive stimuli. The results from TG study suggest that experiments
aimed at understanding active vibrissal touch should be done in the active mode, since
input to the vibrissal system differs qualitatively between the passive and active
modes.
Although the principle of muscle-driven whisker movement is similar between
artificial and natural whisking, the mechanics of whisker movement in this artificial
whisking mode differs in several respects from natural whisking. In awake animals,
both whisker protraction and retraction can be active, using intrinsic and extrinsic
muscles (Berg and Kleinfeld, 2003a), while during artificial whisking retraction is
always passive. Electrically-driven whisking often includes a small stimulus-locked
component superimposed on the main protraction trajectory. Also, artificially-
86
generated whisking cycles are nearly identical, while natural whisking patterns
animals may vary considerable from cycle to cycle (Knutsen et al., 2006).
Furthermore, sympathetic and parasympathetic activation fills the follicle with blood
during natural whisking, possibly affecting movement parameters and
mechanoreceptor activation (Fundin et al., 1997): it is not known to what extent such
effects are also present during artificial whisking. Thus, although the principles by
which movement properties are encoded in neuronal responses are expected to occur
in behavior rats, the exact coding scheme of thalamic and cortical neurons during
active touch still need to be further examined in awaked behavior rats.
Effects of anesthesia My experimental paradigm utilized rats under general anesthesia, which affects
response amplitude, latency, duration, and adaptation in the thalamus and cortex
(Simons et al., 1992; Fanselow and Nicolelis, 1999; Friedberg et al., 1999; Castro-
Alamancos, 2004). However, these effects are mild and are expected to be similar for
neurons in all thalamic nuclei or all cortical nuclei, and thus cannot account for the
prominent differences in response types in the thalamus or touch response properties
among all thalamic and cortical nuclei we report here. The state of thalamic and
cortical neurons during the steady-state response phase, the phase used herein for
response classification in anesthetized rats, is considered to be analogous to the state
of thalamic and cortical neurons during exploratory whisking in awake rats (Castro-
Alamancos, 2002a; Nicolelis and Fanselow, 2002; Castro-Alamancos, 2004).
Consistently, during steady-state thalamic neurons are hypothesized to function in
their gating, signal processing mode (Sherman and Guillery, 1996). Nevertheless,
under anesthesia the intensity and nature of top-down effects, such as those affecting
the thalamus directly, or indirectly (Bokor et al., 2005; Lavallee et al., 2005), are
probably different, the efferent signals that control whisking are lacking, and the
sensory-motor loops that control active touch (Kleinfeld et al., 1999) are practically
opened. Thus, although the basic segregation of response types observed in the
thalamus here in anesthetized rats is expected to occur in awaked ones, the exact
behavior of thalamic and cortical neurons during active touch should be further
studied in awaked behaving rats.
87
REFERENCES Abbott LF, Varela JA, Sen K, Nelson SB (1997) Synaptic depression and cortical gain
control. Science 275:220-224. Ahissar E (1998) Temporal-code to rate-code conversion by neuronal phase-locked
loops. Neural Comput 10:597-650. Ahissar E, Vaadia E (1990) Oscillatory activity of single units in a somatosensory
cortex of an awake monkey and their possible role in texture analysis. Proc Natl Acad Sci USA 87:8935-8939.
Ahissar E, Arieli A (2001) Figuring space by time. Neuron 32:185-201. Ahissar E, Zacksenhouse M (2001) Temporal and spatial coding in the rat vibrissal
system. Prog Brain Res 130:75-87. Ahissar E, Haidarliu S, Zacksenhouse M (1997) Decoding temporally encoded
sensory input by cortical oscillations and thalamic phase comparators. Proc Natl Acad Sci U S A 94:11633-11638.
Ahissar E, Sosnik R, Haidarliu S (2000) Transformation from temporal to rate coding in a somatosensory thalamocortical pathway. Nature 406:302-306.
Ahissar E, Sosnik R, Bagdasarian K, Haidarliu S (2001) Temporal frequency of whisker movement. II. Laminar organization of cortical representations. J Neurophysiol 86:354-367.
Ahrens KF, Kleinfeld D (2004) Current flow in vibrissa motor cortex can phase-lock with exploratory rhythmic whisking in rat. J Neurophysiol 92:1700-1707.
Akers RM, Killackey HP (1978) Organization of corticocortical connections in the parietal cortex of the rat. J Comp Neurol 181:513-537.
Alloway KD, Hoffer ZS, Hoover JE (2003) Quantitative comparisons of corticothalamic topography within the ventrobasal complex and the posterior nucleus of the rodent thalamus. Brain Res 968:54-68.
Alloway KD, Mutic JJ, Hoffer ZS, Hoover JE (2000) Overlapping corticostriatal projections from the rodent vibrissal representations in primary and secondary somatosensory cortex. J Comp Neurol 426:51-67.
Arabzadeh E, Zorzin E, Diamond ME (2005) Neuronal encoding of texture in the whisker sensory pathway. PLoS Biol 3:e17.
Armstrong-James M, Callahan CA, Friedman MA (1991) Thalamo-cortical processing of vibrissal information in the rat. I. Intracortical origins of surround but not centre-receptive fields of layer IV neurones in the rat S1 barrel field cortex. J Comp Neurol 303:193-210.
Barba C, Frot M, Mauguiere F (2002) Early secondary somatosensory area (SII) SEPs. Data from intracerebral recordings in humans. Clin Neurophysiol 113:1778-1786.
Barth DS, Goldberg N, Brett B, Di S (1995) The spatiotemporal organization of auditory, visual, and auditory-visual evoked potentials in rat cortex. Brain Res 678:177-190.
Benison AM, Rector DM, Barth DS (2006) Hemispheric Mapping of Secondary Somatosensory Cortex in the Rat. J Neurophysiol.
Berg RW, Kleinfeld D (2003a) Vibrissa movement elicited by rhythmic electrical microstimulation to motor cortex in the aroused rat mimics exploratory whisking. J Neurophysiol 90:2950-2963. Epub 2003 Aug 2956.
Berg RW, Kleinfeld D (2003b) Rhythmic whisking by rat: retraction as well as protraction of the vibrissae is under active muscular control. J Neurophysiol 89:104-117.
88
Bermejo R, Vyas A, Zeigler HP (2002) Topography of rodent whisking--I. Two-dimensional monitoring of whisker movements. Somatosens Mot Res 19:341-346.
Bernardo KL, McCasland JS, Woolsey TA, Strominger RN (1990) Local intra- and interlaminar connections in mouse barrel cortex. J Comp Neurol 291:231-255.
Bishop GH (1959) The relation between nerve fiber size and sensory modality: Phylogenetic implications of the afferent innervation of cortex. J Nerv Ment Dis 128:89-114.
Bokor H, Frere SG, Eyre MD, Slezia A, Ulbert I, Luthi A, Acsady L (2005) Selective GABAergic control of higher-order thalamic relays. Neuron 45:929-940.
Brecht M (2006) Cortical commands in active touch. Novartis Found Symp 270:38-48; discussion 48-58.
Brecht M (2007) Barrel cortex and whisker-mediated behaviors. Curr Opin Neurobiol 17:408-416.
Brecht M, Sakmann B (2002a) Whisker maps of neuronal subclasses of the rat ventral posterior medial thalamus, identified by whole-cell voltage recording and morphological reconstruction. J Physiol 538:495-515.
Brecht M, Sakmann B (2002b) Dynamic representation of whisker deflection by synaptic potentials in spiny stellate and pyramidal cells in the barrels and septa of layer 4 rat somatosensory cortex. J Physiol 543:49-70.
Brecht M, Preilowski B, Merzenich MM (1997) Functional architecture of the mystacial vibrissae. Behav Brain Res 84:81-97.
Brecht M, Schneider M, Sakmann B, Margrie TW (2004) Whisker movements evoked by stimulation of single pyramidal cells in rat motor cortex. Nature 427:704-710.
Brett-Green B, Fifkova E, Larue DT, Winer JA, Barth DS (2003) A multisensory zone in rat parietotemporal cortex: intra- and extracellular physiology and thalamocortical connections. J Comp Neurol 460:223-237.
Brett-Green B, Paulsen M, Staba RJ, Fifkova E, Barth DS (2004) Two distinct regions of secondary somatosensory cortex in the rat: topographical organization and multisensory responses. J Neurophysiol 91:1327-1336.
Brown AW, Waite PM (1974) Responses in the rat thalamus to whisker movements produced by motor nerve stimulation. J Physiol 238:387-401.
Bruce LL, McHaffie JG, Stein BE (1987) The organization of trigeminotectal and trigeminothalamic neurons in rodents: a double-labeling study with fluorescent dyes. J Comp Neurol 262:315-330.
Brumberg JC, Pinto DJ, Simons DJ (1999) Cortical columnar processing in the rat whisker-to-barrel system. J Neurophysiol 82:1808-1817.
Bruno RM, Simons DJ (2002) Feedforward mechanisms of excitatory and inhibitory cortical receptive fields. J Neurosci 22:10966-10975.
Bruno RM, Khatri V, Land PW, Simons DJ (2003) Thalamocortical angular tuning domains within individual barrels of rat somatosensory cortex. J Neurosci 23:9565-9574.
Bureau I, von Saint Paul F, Svoboda K (2006) Interdigitated paralemniscal and lemniscal pathways in the mouse barrel cortex. PLoS Biol 4:e382.
Burton H, Fabri M, Alloway K (1995) Cortical areas within the lateral sulcus connected to cutaneous representations in areas 3b and 1: a revised interpretation of the second somatosensory area in macaque monkeys. J Comp Neurol 355:539-562.
89
Cadusseau J, Roger M (1990) Distribution of the efferent projections of the rat posterior thalamic nucleus as revealed by Phaseolus vulgaris immunohistochemistry. J Hirnforsch 31:459-465.
Carmena JM, Lebedev MA, Crist RE, O'Doherty JE, Santucci DM, Dimitrov DF, Patil PG, Henriquez CS, Nicolelis MA (2003) Learning to control a brain-machine interface for reaching and grasping by primates. PLoS Biol 1:E42.
Carvell GE, Simons DJ (1987) Thalamic and corticocortical connections of the second somatic sensory area of the mouse. J Comp Neurol 265:409-427.
Carvell GE, Simons DJ (1990) Biometric analyses of vibrissal tactile discrimination in the rat. J Neurosci 10:2638-2648.
Carvell GE, Simons DJ (1995) Task- and subject-related differences in sensorimotor behavior during active touch. Somatosens Mot Res 12:1-9.
Casagrande VA (1994) A third parallel visual pathway to primate area V1. Trends Neurosci 17:305-310.
Castro-Alamancos MA (2002a) Different temporal processing of sensory inputs in the rat thalamus during quiescent and information processing states in vivo. J Physiol 539:567-578.
Castro-Alamancos MA (2002b) Properties of primary sensory (lemniscal) synapses in the ventrobasal thalamus and the relay of high-frequency sensory inputs. J Neurophysiol 87:946-953.
Castro-Alamancos MA (2004) Absence of rapid sensory adaptation in neocortex during information processing states. Neuron 41:455-464.
Chakrabarti S, Alloway KD (2006) Differential origin of projections from SI barrel cortex to the whisker representations in SII and MI. J Comp Neurol 498:624-636.
Chiaia NL, Rhoades RW, Bennett-Clarke CA, Fish SE, Killackey HP (1991) Thalamic processing of vibrissal information in the rat. I. Afferent input to the medial ventral posterior and posterior nuclei. J Comp Neurol 314:201-216.
Chmielowska J, Carvell GE, Simons DJ (1989) Spatial organization of thalamocortical and corticothalamic projection systems in the rat SmI barrel cortex. J Comp Neurol 285:325-338.
Chung S, Li X, Nelson SB (2002) Short-term depression at thalamocortical synapses contributes to rapid adaptation of cortical sensory responses in vivo. Neuron 34:437-446.
Coleman GT, Zhang HQ, Murray GM, Zachariah MK, Rowe MJ (1999) Organization of somatosensory areas I and II in marsupial cerebral cortex: parallel processing in the possum sensory cortex. J Neurophysiol 81:2316-2324.
Cox CL, Huguenard JR, Prince DA (1996) Heterogeneous axonal arborizations of rat thalamic reticular neurons in the ventrobasal nucleus. J Comp Neurol 366:416-430.
Cramer NP, Li Y, Keller A (2007) The whisking rhythm generator: a novel mammalian network for the generation of movement. J Neurophysiol 97:2148-2158.
Crochet S, Petersen CC (2006) Correlating whisker behavior with membrane potential in barrel cortex of awake mice. Nat Neurosci.
Derdikman D, Yu C, Haidarliu S, Bagdasarian K, Arieli A, Ahissar E (2006a) Layer-specific touch-dependent facilitation and depression in the somatosensory cortex during active whisking. J Neurosci 26:9538-9547.
90
Derdikman D, Szwed M, Bagdasarian K, Knutsen PM, Pietr M, Yu C, Arieli A, Ahissar E (2006b) Active construction of percepts about object location. Novartis Found Symp 270:4-14; discussion 14-17, 51-18.
Deschenes M, Veinante P, Zhang ZW (1998) The organization of corticothalamic projections: reciprocity versus parity. Brain Res Brain Res Rev 28:286-308.
Deschenes M, Timofeeva E, Lavallee P (2003) The relay of high-frequency sensory signals in the Whisker-to-barreloid pathway. J Neurosci 23:6778-6787.
Di S, Brett B, Barth DS (1994) Polysensory evoked potentials in rat parietotemporal cortex: combined auditory and somatosensory responses. Brain Res 642:267-280.
Diamond IT (1983) Parellel pathways in the auditory, visual and somatic systems. In: Somatosensory integration in the thalamus (Macchi G, Rustioni A, Spreafico R, eds), pp 251-272. Amsterdam: Elsevier.
Diamond ME (2000) Parallel sensing. Nature 406:245-247. Diamond ME, Armstrong-James M, Ebner FF (1992a) Somatic sensory responses in
the rostral sector of the posterior group (POm) and in the ventral posterior medial nucleus (VPM) of the rat thalamus. J Comp Neurol 318:462-476.
Diamond ME, Armstrong-James M, Budway MJ, Ebner FF (1992b) Somatic sensory responses in the rostral sector of the posterior group (POm) and in the ventral posterior medial nucleus (VPM) of the rat thalamus: dependence on the barrel field cortex. J Comp Neurol 319:66-84.
Ebara S, Kumamoto K, Matsuura T, Mazurkiewicz JE, Rice FL (2002) Similarities and differences in the innervation of mystacial vibrissal follicle-sinus complexes in the rat and cat: a confocal microscopic study. J Comp Neurol 449:103-119.
Erzurumlu RS, Killackey HP (1980) Diencephalic projections of the subnucleus interpolaris of the brainstem trigeminal complex in the rat. Neuroscience 5:1891-1901.
Fabri M, Burton H (1991) Ipsilateral cortical connections of primary somatic sensory cortex in rats. J Comp Neurol 311:405-424.
Fanselow EE, Nicolelis MA (1999) Behavioral modulation of tactile responses in the rat somatosensory system. J Neurosci 19:7603-7616.
Fee MS, Mitra PP, Kleinfeld D (1997) Central versus peripheral determinants of patterned spike activity in rat vibrissa cortex during whisking. J Neurophysiol 78:1144-1149.
Feldmeyer D, Roth A, Sakmann B (2005) Monosynaptic connections between pairs of spiny stellate cells in layer 4 and pyramidal cells in layer 5A indicate that lemniscal and paralemniscal afferent pathways converge in the infragranular somatosensory cortex. J Neurosci 25:3423-3431.
Friedberg MH, Lee SM, Ebner FF (1999) Modulation of receptive field properties of thalamic somatosensory neurons by the depth of anesthesia. J Neurophysiol 81:2243-2252.
Fundin BT, Bergman E, Ulfhake B (1997) Alterations in mystacial pad innervation in the aged rat. Exp Brain Res 117:324-340.
Gao P, Hattox AM, Jones LM, Keller A, Zeigler HP (2003) Whisker motor cortex ablation and whisker movement patterns. Somatosens Mot Res 20:191-198.
Garabedian CE, Jones SR, Merzenich MM, Dale A, Moore CI (2003) Band-pass response properties of rat SI neurons. J Neurophysiol 90:1379-1391.
91
Garraghty PE, Florence SL, Tenhula WN, Kaas JH (1991) Parallel thalamic activation of the first and second somatosensory areas in prosimian primates and tree shrews. J Comp Neurol 311:289-299.
Ghazanfar AA, Nicolelis MA (1997) Nonlinear processing of tactile information in the thalamocortical loop. J Neurophysiol 78:506-510.
Gibson JJ (1962) Observations on active touch. Psychol Rev 69:477-491. Gibson JM, Welker WI (1983a) Quantitative studies of stimulus coding in first-order
vibrissa afferents of rats. 2. Adaptation and coding of stimulus parameters. Somatosens Res 1:95-117.
Gibson JM, Welker WI (1983b) Quantitative studies of stimulus coding in first-order vibrissa afferents of rats. 1. Receptive field properties and threshold distributions. Somatosens Res 1:51-67.
Golomb D, Ahissar E, Kleinfeld D (2006) Coding of stimulus frequency by latency in thalamic networks through the interplay of GABAB-mediated feedback and stimulus shape. J Neurophysiol 95:1735-1750.
Gopal V, Hartmann MJ (2007) Using hardware models to quantify sensory data acquisition across the rat vibrissal array. Bioinspir Biomim 2:S135-145.
Gottlieb JP, Keller A (1997) Intrinsic circuitry and physiological properties of pyramidal neurons in rat barrel cortex. Exp Brain Res 115:47-60.
Grinevich V, Brecht M, Osten P (2005) Monosynaptic pathway from rat vibrissa motor cortex to facial motor neurons revealed by lentivirus-based axonal tracing. J Neurosci 25:8250-8258.
Guic-Robles E, Jenkins WM, Bravo H (1992) Vibrissal roughness discrimination is barrelcortex- dependent. Behav Brain Res 48:145-152.
Guillery RW, Sherman SM (2002) The thalamus as a monitor of motor outputs. Philos Trans R Soc Lond B Biol Sci 357:1809-1821.
Haidarliu S, Ahissar E (2001) Size gradients of barreloids in the rat thalamus. J Comp Neurol 429:372-387.
Haiss F, Schwarz C (2005) Spatial segregation of different modes of movement control in the whisker representation of rat primary motor cortex. J Neurosci 25:1579-1587.
Hartings JA, Simons DJ (2000) Inhibition suppresses transmission of tonic vibrissa-evoked activity in the rat ventrobasal thalamus. J Neurosci 20:RC100.
Hartings JA, Temereanca S, Simons DJ (2003) Processing of periodic whisker deflections by neurons in the ventroposterior medial and thalamic reticular nuclei. J Neurophysiol 90:3087-3094.
Harvey MA, Bermejo R, Zeigler HP (2001) Discriminative whisking in the head-fixed rat: optoelectronic monitoring during tactile detection and discrimination tasks. Somatosens Mot Res 18:211-222.
Hattox A, Li Y, Keller A (2003) Serotonin regulates rhythmic whisking. Neuron 39:343-352.
Hattox AM, Priest CA, Keller A (2002) Functional circuitry involved in the regulation of whisker movements. J Comp Neurol 442:266-276.
Hawkins J, Blakeslee S (2004) On Inteligence. New York: Henry Holt and Company. He J, Hu B (2002) Differential distribution of burst and single-spike responses in
auditory thalamus. J Neurophysiol 88:2152-2156. Hoeflinger BF, Bennett-Clarke CA, Chiaia NL, Killackey HP, Rhoades RW (1995)
Patterning of local intracortical projections within the vibrissae representation of rat primary somatosensory cortex. J Comp Neurol 354:551-563.
92
Hoffer ZS, Hoover JE, Alloway KD (2003) Sensorimotor corticocortical projections from rat barrel cortex have an anisotropic organization that facilitates integration of inputs from whiskers in the same row. J Comp Neurol 466:525-544.
Horev G, Benjamini Y, Sakov A, Golani I (2007) Estimating wall guidance and attraction in mouse free locomotor behavior. Genes Brain Behav 6:30-41.
Inui K, Wang X, Tamura Y, Kaneoke Y, Kakigi R (2004) Serial processing in the human somatosensory system. Cereb Cortex 14:851-857.
Ito M (1988) Response properties and topography of vibrissa-sensitive VPM neurons in the rat. J Neurophysiol 60:1181-1197.
Jones EG (2003) Chemically defined parallel pathways in the monkey auditory system. Ann N Y Acad Sci 999:218-233.
Kamatani D, Hishida R, Kudoh M, Shibuki K (2007) Experience-dependent formation of activity propagation patterns at the somatosensory S1 and S2 boundary in rat cortical slices. Neuroimage 35:47-57.
Karhu J, Tesche CD (1999) Simultaneous early processing of sensory input in human primary (SI) and secondary (SII) somatosensory cortices. J Neurophysiol 81:2017-2025.
Katz D (1989, Original work published 1925) The world of touch (L.E. Krueger trans.). Hillsdale, NJ: Erlbaum.
Keller A, White EL (1987) Synaptic organization of GABAergic neurons in the mouse SmI cortex. J Comp Neurol 262:1-12.
Kim U, Ebner FF (1999) Barrels and septa: separate circuits in rat barrels field cortex. J Comp Neurol 408:489-505.
Kleinfeld D, Delaney KR (1996) Distributed representation of vibrissa movement in the upper layers of somatosensory cortex revealed with voltage-sensitive dyes. J Comp Neurol 375:89-108.
Kleinfeld D, Berg RW, O'Connor SM (1999) Anatomical loops and their electrical dynamics in relation to whisking by rat. Somatosens Mot Res 16:69-88.
Kleinfeld D, Ahissar E, Diamond ME (2006) Active sensation: insights from the rodent vibrissa sensorimotor system. Curr Opin Neurobiol 16:435-444.
Kleinfeld D, Sachdev RN, Merchant LM, Jarvis MR, Ebner FF (2002) Adaptive filtering of vibrissa input in motor cortex of rat. Neuron 34:1021-1034.
Knutsen PM, Derdikman D, Ahissar E (2005) Tracking whisker and head movements in unrestrained behaving rodents. J Neurophysiol 93:2294-2301.
Knutsen PM, Pietr M, Ahissar E (2006) Haptic object localization in the vibrissal system: behavior and performance. J Neurosci 26:8451-8464.
Koralek KA, Jensen KF, Killackey HP (1988) Evidence for two complementary patterns of thalamic input to the rat somatosensory cortex. Brain Res 463:346-351.
Koralek KA, Olavarria J, Killackey HP (1990) Areal and laminar organization of corticocortical projections in the rat somatosensory cortex. J Comp Neurol 299:133-150.
Krupa DJ, Matell MS, Brisben AJ, Oliveira LM, Nicolelis MA (2001) Behavioral properties of the trigeminal somatosensory system in rats performing whisker-dependent tactile discriminations. J Neurosci 21:5752-5763.
Krupa DJ, Wiest MC, Shuler MG, Laubach M, Nicolelis MA (2004) Layer-specific somatosensory cortical activation during active tactile discrimination. Science 304:1989-1992.
93
Kwegyir-Afful EE, Keller A (2004) Response properties of whisker-related neurons in rat second somatosensory cortex. J Neurophysiol 92:2083-2092.
Kyriazi HT, Carvell GE, Simons DJ (1994) OFF response transformations in the whisker/barrel system. J Neurophysiol 72:392-401.
Land PW, Simons DJ (1985a) Cytochrome oxidase staining in the rat SmI barrel cortex. J Comp Neurol 238:225-235.
Land PW, Simons DJ (1985b) Metabolic and structural correlates of the vibrissae representation in the thalamus of the adult rat. NeurosciLett 60:319-324.
Land PW, Buffer SA, Jr., Yaskosky JD (1995) Barreloids in adult rat thalamus: three-dimensional architecture and relationship to somatosensory cortical barrels. J Comp Neurol 355:573-588.
Lavallee P, Urbain N, Dufresne C, Bokor H, Acsady L, Deschenes M (2005) Feedforward inhibitory control of sensory information in higher-order thalamic nuclei. J Neurosci 25:7489-7498.
Lee SH, Simons DJ (2004) Angular tuning and velocity sensitivity in different neuron classes within layer 4 of rat barrel cortex. J Neurophysiol 91:223-229.
Leiser SC, Moxon KA (2007) Responses of trigeminal ganglion neurons during natural whisking behaviors in the awake rat. Neuron 53:117-133.
Lichtenstein SH, Carvell GE, Simons DJ (1990) Responses of rat trigeminal ganglion neurons to movements of vibrissae in different directions. Somatosens Mot Res 7:47-65.
Lovick TA (1972) The behavioural repertoire of precollicular decerebrate rats. J Physiol 226:4P-6P.
Lu SM, Lin RC (1993) Thalamic afferents of the rat barrel cortex: a light- and electron-microscopic study using Phaseolus vulgaris leucoagglutinin as an anterograde tracer. Somatosens Mot Res 10:1-16.
May PJ (2005) The mammalian superior colliculus: laminar structure and connections. Prog Brain Res 151:321-378.
Mehta SB, Whitmer D, Figueroa R, Williams BA, Kleinfeld D (2007) Active spatial perception in the vibrissa scanning sensorimotor system. PLoS Biol 5:e15.
Melzer P, Champney GC, Maguire MJ, Ebner FF (2006) Rate code and temporal code for frequency of whisker stimulation in rat primary and secondary somatic sensory cortex. Exp Brain Res 172:370-386.
Menzel RR, Barth DS (2005) Multisensory and secondary somatosensory cortex in the rat. Cereb Cortex 15:1690-1696.
Miller KD, Pinto DJ, Simons DJ (2001) Processing in layer 4 of the neocortical circuit: new insights from visual and somatosensory cortex. Curr Opin Neurobiol 11:488-497.
Minnery BS, Bruno RM, Simons DJ (2003) Response transformation and receptive-field synthesis in the lemniscal trigeminothalamic circuit. J Neurophysiol 90:1556-1570.
Mitchinson B, Martin CJ, Grant RA, Prescott TJ (2007) Feedback control in active sensing: rat exploratory whisking is modulated by environmental contact. Proc Biol Sci 274:1035-1041.
Miyashita E, Mori S (1995) The superior colliculus relays signals descending from the vibrissal motor cortex to the facial nerve nucleus in the rat. Neurosci Lett 195:69-71.
Miyashita E, Keller A, Asanuma H (1994) Input-output organization of the rat vibrissal motor cortex. Exp Brain Res 99:223-232.
94
Moore CI (2004) Frequency-dependent processing in the vibrissa sensory system. J Neurophysiol 91:2390-2399.
Murray GM, Zhang HQ, Kaye AN, Sinnadurai T, Campbell DH, Rowe MJ (1992) Parallel processing in rabbit first (SI) and second (SII) somatosensory cortical areas: effects of reversible inactivation by cooling of SI on responses in SII. J Neurophysiol 68:703-710.
Nguyen QT, Kleinfeld D (2005) Positive feedback in a brainstem tactile sensorimotor loop. Neuron 45:447-457.
Nicolelis MA, Fanselow EE (2002) Thalamocortical [correction of Thalamcortical] optimization of tactile processing according to behavioral state. Nat Neurosci 5:517-523.
Pali J, Rencz B, Hamori J (2000) Innervation of a single vibrissa in the whisker-pad of rats. Neuroreport 11:849-851.
Perl ER (1963) Somatosensory mechanisms. Annu Rev Physiol 25:459-492. Petersen CC (2007) The functional organization of the barrel cortex. Neuron 56:339-
355. Petreanu L, Huber D, Sobczyk A, Svoboda K (2007) Channelrhodopsin-2-assisted
circuit mapping of long-range callosal projections. Nat Neurosci 10:663-668. Pierret T, Lavallee P, Deschenes M (2000) Parallel streams for the relay of vibrissal
information through thalamic barreloids. J Neurosci 20:7455-7462. Pinault D, Deschenes M (1998) Projection and innervation patterns of individual
thalamic reticular axons in the thalamus of the adult rat: a three-dimensional, graphic, and morphometric analysis. J Comp Neurol 391:180-203.
Pinto DJ, Brumberg JC, Simons DJ (2000) Circuit dynamics and coding strategies in rodent somatosensory cortex. J Neurophysiol 83:1158-1166.
Pinto DJ, Hartings JA, Brumberg JC, Simons DJ (2003) Cortical damping: analysis of thalamocortical response transformations in rodent barrel cortex. Cereb Cortex 13:33-44.
Pons TP, Garraghty PE, Mishkin M (1992) Serial and parallel processing of tactual information in somatosensory cortex of rhesus monkeys. J Neurophysiol 68:518-527.
Pons TP, Garraghty PE, Friedman DP, Mishkin M (1987) Physiological evidence for serial processing in somatosensory cortex. Science 237:417-420.
Remple MS, Henry EC, Catania KC (2003) Organization of somatosensory cortex in the laboratory rat (Rattus norvegicus): Evidence for two lateral areas joined at the representation of the teeth. J Comp Neurol 467:105-118.
Rhoades RW, Belford GR, Killackey HP (1987) Receptive-field properties of rat ventral posterior medial neurons before and after selective kainic acid lesions of the trigeminal brain stem complex. J Neurophysiol 57:1577-1600.
Rice FL, Mance A, Munger BL (1986) A comparative light microscopic analysis of the sensory innervation of the mystacial pad. I. Innervation of vibrissal follicle-sinus complexes. J Comp Neurol 252:154-174.
Sachdev RN, Egli M, Stonecypher M, Wiley RG, Ebner FF (2000) Enhancement of cortical plasticity by behavioral training in acetylcholine-depleted adult rats. J Neurophysiol 84:1971-1981.
Salinas E, Hernandez A, Zainos A, Romo R (2000) Periodicity and firing rate as candidate neural codes for the frequency of vibrotactile stimuli. J Neurosci 20:5503-5515.
95
Schubert D, Kotter R, Staiger JF (2007) Mapping functional connectivity in barrel-related columns reveals layer- and cell type-specific microcircuits. Brain Struct Funct 212:107-119.
Sherman SM, Guillery RW (1996) Functional organization of thalamocortical relays. J Neurophysiol 76:1367-1395.
Shuler MG, Krupa DJ, Nicolelis MA (2002) Integration of bilateral whisker stimuli in rats: role of the whisker barrel cortices. Cereb Cortex 12:86-97.
Simons DJ (1978) Response properties of vibrissa units in rat SI somatosensory neocortex. J Neurophysiol 41:798-820.
Simons DJ, Carvell GE (1989) Thalamocortical response transformation in the rat vibrissa/barrel system. J Neurophysiol 61:311-330.
Simons DJ, Carvell GE, Hershey AE, Bryant DP (1992) Responses of barrel cortex neurons in awake rats and effects of urethane anesthesia. Exp Brain Res 91:259-272.
Solomon JH, Hartmann MJ (2006) Biomechanics: robotic whiskers used to sense features. Nature 443:525.
Sosnik R, Haidarliu S, Ahissar E (2001) Temporal frequency of whisker movement. I. Representations in brain stem and thalamus. J Neurophysiol 86:339-353.
Spreafico R, Barbaresi P, Weinberg RJ, Rustioni A (1987) SII-projecting neurons in the rat thalamus: a single- and double-retrograde-tracing study. Somatosens Res 4:359-375.
Staiger JF (2006) Immediate-early gene expression in the barrel cortex. Somatosens Mot Res 23:135-146.
Szwed M, Bagdasarian K, Ahissar E (2003) Encoding of vibrissal active touch. Neuron 40:621-630.
Szwed M, Bagdasarian K, Blumenfeld B, Barak O, Derdikman D, Ahissar E (2006) Responses of trigeminal ganglion neurons to the radial distance of contact during active vibrissal touch. J Neurophysiol 95:791-802.
Timofeeva E, Merette C, Emond C, Lavallee P, Deschenes M (2003) A map of angular tuning preference in thalamic barreloids. J Neurosci 23:10717-10723.
Trageser JC, Keller A (2004) Reducing the uncertainty: gating of peripheral inputs by zona incerta. J Neurosci 24:8911-8915.
Turman AB, Morley JW, Zhang HQ, Rowe MJ (1995) Parallel processing of tactile information in cat cerebral cortex: effect of reversible inactivation of SII on SI responses. J Neurophysiol 73:1063-1075.
Turman AB, Ferrington DG, Ghosh S, Morley JW, Rowe MJ (1992) Parallel processing of tactile information in the cerebral cortex of the cat: effect of reversible inactivation of SI on responsiveness of SII neurons. J Neurophysiol 67:411-429.
Veinante P, Jacquin MF, Deschenes M (2000) Thalamic projections from the whisker-sensitive regions of the spinal trigeminal complex in the rat. J Comp Neurol 420:233-243.
Walker C, Sinha MM (1972) Somatotopic organization of Smll cerebral neocortex in albino rat. Brain Res 37:132-136.
Wallace MT, Ramachandran R, Stein BE (2004) A revised view of sensory cortical parcellation. Proc Natl Acad Sci U S A 101:2167-2172.
Welker C (1971) Microelectrode delineation of fine grain somatotopic organization of (SmI) cerebral neocortex in albino rat. Brain Res 26:259-275.
Welker C (1976) Receptive fields of barrels in the somatosensory neocortex of the rat. J Comp Neurol 166:173-189.
96
Wessberg J, Stambaugh CR, Kralik JD, Beck PD, Laubach M, Chapin JK, Kim J, Biggs SJ, Srinivasan MA, Nicolelis MA (2000) Real-time prediction of hand trajectory by ensembles of cortical neurons in primates. Nature 408:361-365.
Wiener N (1949) Cybernetics. New York: John Wiley & Sons. Williams MN, Zahm DS, Jacquin MF (1994) Differential foci and synaptic
organization of the principal and spinal trigeminal projections to the thalamus in the rat. Eur J Neurosci 6:429-453.
Woolsey TA, Van der Loos H (1970) The structural organization of layer IV in the somatosensory region (SI) of mouse cerebral cortex. The description of a cortical field composed of discrete cytoarchitectonic units. Brain Res 17:205-242.
Yu C, Derdikman D, Haidarliu S, Ahissar E (2006) Parallel thalamic pathways for whisking and touch signals in the rat. PLoS Biol 4:e124.
Zacksenhouse M, Ahissar E (2006) Temporal decoding by phase-locked loops: unique features of circuit-level implementations and their significance for vibrissal information processing. Neural Comput 18:1611-1636.
Zhang HQ, Zachariah MK, Coleman GT, Rowe MJ (2001a) Hierarchical equivalence of somatosensory areas I and II for tactile processing in the cerebral cortex of the marmoset monkey. J Neurophysiol 85:1823-1835.
Zhang HQ, Murray GM, Turman AB, Mackie PD, Coleman GT, Rowe MJ (1996) Parallel processing in cerebral cortex of the marmoset monkey: effect of reversible SI inactivation on tactile responses in SII. J Neurophysiol 76:3633-3655.
Zhang HQ, Murray GM, Coleman GT, Turman AB, Zhang SP, Rowe MJ (2001b) Functional characteristics of the parallel SI- and SII-projecting neurons of the thalamic ventral posterior nucleus in the marmoset. J Neurophysiol 85:1805-1822.
Zhang M, Barash S (2000) Neuronal switching of sensorimotor transformations for antisaccades. Nature 408:971-975.
Zucker E, Welker WI (1969) Coding of somatic sensory input by vibrissae neurons in the rat's trigeminal ganglion. Brain Res 12:138-156.
97
LIST OF PUBLICATIONS
Refereed Journals:
1. Yu, C., Haidarliu, S., Derdikman, D., and Ahissar, E. (in preparation) Coding
of object location in the secondary somatosensory cortex during active vibrissal
touch
2. Haidarliu, S., Yu, C., Rubin, N., and Ahissar E. (in review) Lemniscal and
extraleminiscal compartments in the VPM of the rat. Submitted to Brain Struct
Funct.
3. Yu, C., Derdikman, D., Haidarliu, S., and Ahissar, E. (2006) Parallel thalamic
pathways for whisking and touch signals in the rat. PLoS Biol, 4(5): e124.
4. Derdikman, D., Yu, C., Haidarliu, S., Bagdasarian, K., Arieli, A., and Ahissar,
E. (2006) Layer-specific touch-dependent facilitation and depression in the
somatosensory cortex during active whisking. J Neurosci, 26(37): 9538-47
Book chapters:
1. Derdikman D, Szwed M, Bagdasarian K, Knutsen PM, Pietr M, Yu C, Arieli A,
Ahissar E (2006) Active construction of percepts about object location. Novartis
Found Symp 270:4-14; discussion 14-17, 51-18.
2. Derdikman, D., Szwed, M., Knutsen, P.M., Pietr, M., Yu, C., Bagdasarian, K.,
and Ahissar, E. Temporal processing in active and passive touch. In: Temporal
Processing in Sensory Systems, (E. Covey, ed.), Kluwer Academic, in press.
98
Declaration of Independence of Ph.D work I, Chunxiu Yu, hereby declare that this doctoral thesis is a compilation of
independent work carried out by myself in the laboratory of Professor Ehud Ahissar,
Department of Neurobiology, Weizmann Institute of Science.