new tools for studying cdx neural functionspcr screen but require confirmation by southern blot. the...
TRANSCRIPT
UNIVERSITÉ DU QUÉBEC À MONTRÉAL
NEW TOOLS FOR STUDYING CDX NEURAL FUNCTIONS
THE SIS
PRESENTED
AS PART
OF A MASTER' S IN BIOCHEMlSTRY
BY
DANIEL AGUlRRE MARTINEZ
FEBRUARY 20 16
UNIVERSITÉ DU QUÉBEC À MONTRÉAL Service des bibliothèques
Avertissement
La diffusion de ce mémoire se fait dans le respect des droits de son auteur, qui a signé le formulaire Autorisation de reproduire et de diffuser un travail de recherche de cycles supérieurs (SDU-522 - Rév.0?-2011). Cette autorisation stipule que «conformément à l'article 11 du Règlement no 8 des études de cycles supérieurs, [l 'auteur] concède à l'Université du Québec à Montréal une licence non exclusive d'utilisation et de publication de la totalité ou d'une partie importante de [son] travail de recherche pour des fins pédagogiques et non commerciales. Plus précisément, [l 'auteur] autorise l'Université du Québec à Montréal à reproduire , diffuser, prêter, distribuer ou vendre des copies de [son] travail de recherche à des fins non commerciales sur quelque support que ce soit, y compris l'Internet. Cette licence et cette autorisation n'entraînent pas une renonciation de [la] part [de l'auteur] à [ses] droits moraux ni à [ses] droits de propriété intellectuelle. Sauf entente contraire, [l 'auteur] conserve la liberté de diffuser et de commercial iser ou non ce travail dont [il] possède un exemplaire. »
UNIVERSITÉ DU QUÉBEC À MONTRÉAL
NOUVEAUX OUTILS POUR L'ETUDE DES FONCTIONS NEURALES DES
GENESCDX
MÉMOIRE
PRÉSENTÉ
COMME EXIGENCE PARTIELLE
DE LA MAÎTRISE EN BIOCHIMIE
PAR
DANIEL AGUIRRE MARTINEZ
FÉVRIER 2016
-----------------------------------------------------------------
ACKNOWLEDGEMENTS
I would like to address my sincerest thanks to the foll owing people who have helped
me throughout the last two yea rs of my master' s pro gram:
N ico las Pilon, my supervise r and one of my professo rs at the uni versity, fo r giving
me the opportunity to learn in his lab. Despite ali the piles of books and articles
keeping him busy in hi s offi ce he is always ava il able to talk to.
Ouliana Souchkova for imparting on me va luable technica l and scientific knowledge.
Especiall y at the beginning when she had to do ali that plus deal with my broken
ankle and limited mobility.
A li the members of the lab who have helped me in sorne way or another
My family and fr iends for their continuing love and support
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ........................... .................. ...................... ............... ......... i
FIGURE LIST .. ........ ................................... .... ....... ..................................................... vii
TABLE LIST .. ................................................ ...... ....................................................... ix
ABBREV IATIO , SIG A D ACRO YM LLST .................................................... xi
RÉSUMÉ ... .............. .... ............................................ ...... ... ............................ ........... .. . xiii
ABSTRACT ....................... ............................................ ................................. ....... ..... xv
INTRODUCTION ................... .... .. .. ...... ..... .. ... .... ..... .. ... ..... .................... ......... ... .... ....... 1
CHA PTER I 3 BACKGROUND ........... .. .... .. ......... ........................ ........ ... .... .. .. .. ..... ..... ... ........... .. ........ 3 1.1 Yertebrate embryogenes is .. ................ .. ...... .. .... .... ...... .. ........................................ 3
1.2 Mo lecu lar mechanisms ......................................................................................... 8
1.2. 1 Signais ofthe posterior neuroectoderm .................... .. ............................ lü
1 .3 Gene regulation and NC development ................ .; .. .............. ........ ..................... 18
1.3.1 En hancers ........................................ .. ...... ................................................ 18
1.3.2 Cdx2NSE .. .... .... ............................ ...... .. ... ...... .......... ................................ 22
1.4 Diseases ofNCC origin and neural tube defects .. .... ........ .... .............................. 24
1.4. 1 Neurocristopathies .......................... ......... ............................................... 24
1.4.2 Ne ura l tube defects ........................ .... .. ............. ....................................... 26
1.4.3 Too ls for studying neurocristopathi es and NTDs ................................... 27
CHA PTER JI 29 HYPOTHESES A D OBJECTIYES ........................................ ...... ........................... 29 2. 1 Hypotheses ......................................................................... ............ .. .. ................ 29
IV
2.2 Obj ectives ..... .. ... ........ ..... .... ....... .. ...... ...... ..... ... ..... ... ...... .... ...... ....... ........ ..... ... .... 29
CHAPTER III 3 1
MATERlALS AND METHODS ................... .... ... ........ ....... .. .. ..... .. .. .. .. .... ......... ... .. .... 3 1
3.1 Construction of pROSA26-Cdx 1, ES targeting and verifi cation ... ........... ......... 3 1
3.1. t Vector constructi on ........ ............ ... ...... ........ ... ... .... ..... ... ..... .... ..... .. .......... 3 1
3.1.2 Targeting ES ce li s ........ ... ...... ... ...... ....... ......... ... ....... ..... ..... .. ... ... .. ..... .... .. 3 1
3. 1.3 Southern B lott ing .... ... .. ............ ..... ...... .... .. .............. .... .. ....... .... ....... .. .. .... 32
3.2 Testing the neura l spec ifi e regulation of Cdx2NSE ..... ......... .... .. ..... ........... .... .. . 33
3.2.1 Plasn1ids .... ..................... ......... ... .. .. .... .. ... .. .. ... .. .. ........ ............................. 33
3.2.2 Luc iferase assay conditions .... ...... .... .. .. .. .. ... ........ .. .... ..... ... .. .. ...... .. .. .. .. ... 33
3.2.3 Luc iferase assay test and statisti ca l ana lys is .... .... .... .. .. .. .... .. .... .. .... .... .. ... 35
CHAPTER !V 37 TOWARD THE GENERATlON OF MICE CARRY ING THE CRE-IND UCIBLE
CDX J TRANSGENE ... ... ... .... ... ... ... .... ...... .. ... ...... .. ... ... .. ....... ......... ...... ... ... ... .......... 37 4. 1 Introduction .... ......... .. .. .... .. .. .. .. .. .......... .... .. ... .. ... .... .. .. ...... .. .... .. .. ..... .. .. .. .. ... .. .... ... 37
4.2 Results .. .. ... ......... ... .. .. ..... .... .. .. .. .. ...... .. .. .. ... .. ..... ...... .. .. .................. .. .. .. .. .............. 37
4.2. 1 Construction of the targeti ng vector .... .... .. ...................... .. ...... .. .. .... .. .. .. .. 37
4.2.2 Targeting the ROSA26 locus .................................................................. . 38
CHAPTER V 43 TtSTfi-.J G THE NEU RAL SPECIFIC REGULATiON OF CDX2NSE .. .. ........ ...... .. . 43 5.1 In troduction ...... .... ............. .. ...... ........ ...... .. ........................................... .............. 43
5.2 Resul ts .. .... ...... ...... .. .... ........ ... .......... .. .. ..... ... ... .... .. ........... .......... .. ... .. .. ..... .... .. .... . 44
5.2. 1 Potentia l bind ing acti vators .. .. ............................................. .. .. .. .. .. .. .... ... . 44
5.2 .2 NSE and potentia l TFBS sequence homo logy .. .... .... .... .. .. .... .... ...... .. .. .. .. 46
5.2.3 Directional preference and synergy .. ......... .. .. .... .. .. .... .... .. .. .. .. ... ...... .. .. .. .. 49
5.2 .4 Testi ng the reverse NSE .... .. .. .... .............. .. .. .. ........ .. .. .. ...... .. .......... .. ........ 5 1
CHAPTER VI 53 DISCUSSION .. .. .... .. .. ...... ... .. .... ...... .... .... ..... ..... .... ....... .. .. .... .............. ............ ... .... ...... 53 6. 1 ln pursuit of a cond itiona l Cdxl overexpress ion mouse mode! .......... ........ .... .. . 53
6.2 Insights into Cdx2NSE neural regulation and potentia l app licati on .................. 56
v
APPENDlX A .... ... .............. ...... ...... ........... ........... ... .... ........ .... ... ......... ... ... .. ... .. .......... 63 EVOLUTIONARY CONSE RV ATION OF CDX2NSE POTENTIAL TFBS ...... ..... 63
APPENDIX B ........... ... .. .......... ............... ........ .... .................. .. .. ..... ...... .............. ....... .. 65 REGULATION OF PAX3 NCEJ BY CDX2 PROTErN S ....... ....... ..... .... ...... ............ 65
APPENDTX C .. ......... ... ....... .... ......... ....... .... .... ............ .. ..... ... .... ... .. ................... .... .. .. .. 67 CDX2NSE AND FGF SIGNA LLIN G ..... ....... .... .. ... ... ... ... ... ... ........... ........ ........... ... ... 67
REFER ENCES ... .... ......... ...... ........ ......... ......... .. ... .. .... .... ..... ....... .. ..................... .... ...... 69
>
FIGURE LIST
Figure Page
1.1 Steps of neurulat ion ................................................... ......... ... ..... ....... .. .......... .... ... .. . 7
1.2 NCC migration throughout four different leve ls ............... ................ ..... ......... ........ 7
1.3 Neura l crest gene regulatory network ....... ................... .......... ......................... ..... .. . 9
1.4 AP ax is patterning ... .... ..... ...................................................... .... ........................... 11
1.5 The canonicat Wnt signaling pathway . .. ... .... ... .. ............................. ....... ... ..... ....... 11
1.6 The FGF signalling pathway ... ............. .... ... .... ........................................ ... .. ....... .. 13
1.7 The spat ial and temporal di ffe rences of murine Cclx gene express ion ....... ... ........ 17
1.8 Types of cis-regulatory elements ............... .. ...... ........... ................. ........ ...... .. ....... 19
1.9 Techniques for identify ing and testing enhancers .... ....... ... ............... ... ........ .... ..... 2 1
1.10 Identificat ion and testing the Cdx2 neura l spec ifie enhancer (NSE) .... .. ... ... ...... 23
1.11 The enhancer is not entirely neural spec ifie ........................................................ 23
4. 1 Sacll linea rization of pROSA26-Cdx 1 ............................... ................. ............. .... 39
4.2 Depiction of gene target ing at the ROSA26 1ocus ........... ..... ......... ................ ....... .40
VIII
4.3 PC R screen of targeted ES c lones ............................................ .. .. ....... .................. 41
5.1 Potential Lefl1 /Tcfand STEM family binding sites ........................................... .45
5.2 Vertebrate homology ofthe neural specifie enhancer .................. .... .... .. ............... 48
5.3 Cdx2NSE synergy and directional preference ................................................... .. . 50
5.4 Fragmentation results of Cdx2NSErever e ........................................................... 52
TABLE LIST
Tab le Page
1.1 D iversity ofNeural Crest Ce l! derivatives .. .... ... .......... ............ ...... .. ... ... ... .. .... ........ 8
3. 1 Pritners for PCR screen .................. ...... .......... ...... ... ... ... .... .. .. ...... .. .. .. .. .................. 32
3.2 P-va lue guide ... .. .. .... .... .. ......... ...... ..... ..... .. .... ...... ... ................... ........... .. ..... ........... 35
3.3 Oligos to construct Cdx2NSE reverse strand fragme nts . ...... ........ ... .... ... ... ........... 36
AP
BMP
Cd x
CNS
DLHP
DY
EMT
EGFP
FGF
GRN
TRES
LEFl
MEF
MHP
NC
ABBREVTATTON, S!GN AND ACRONYM LJST
Anterior-Posterior
Borre morphogenie protein
Caudal-related homeobox
Central nervous system
Dorsa l-latera l hinge po ints
Dorsa l-Ventral
Epithelial-mesenchymal transition
Enhanced green fluorescence prote in
Fibrob last growth factor
Gene regulatory network
Internai ribosome entry site
Lymphoid enhancer-binding factor 1
Mouse Embryo Fibrob last
Median hinge points
Neural crest
XII
NCC
NCE
NP
NPB
N SE
NT
NTD
PNS
SOX2
TCF
TF
TFBS
Wnt
N eural cres t ce l!
N eura l cres t enhancer
N eural Pl ate
Neura l Pl ate Border
Neura l Spec ifi e E nhancer
Neura l Tube
Ne ura l T ube Defects
Peripheral nervo us system
SRY (sex determining reg ion Y)-box 2
T-cell spec ifi e transcripti on factor
Transcripti on fac tor
Transcripti on factor binding site
Win g less-re lated in tegrati on site
RÉSUM É
Le système nerveux est un composant fo nd amenta l chez les vertébrés qui implique un réseau compl exe de gènes régul ateurs (GRN). Sa mi se en pl ace commence avec l' inducti on du neuroectoderme par l' acti v ité des s ignaux Wnt, FGF et BM P le long de l' axe antéro-postéri eur nouve ll ement fo rmé. Plusieurs signaux transcripti onne ls et morphogén iques coopèrent pour indu ire l' in vagi nati on du neuroectoderme ventra lement afin de fo rmer le tube neural et la crête neura le qui donneront na issance, respecti veme nt, a u système nerveux central et a u système nerveux périphérique . Le neuroectoderme postérieur possède son unique lot de molécules de signa li sati on et toute perturbati on dans les évènements précoce de son déve loppement entra îne un groupe de maladi es et ma lfo rmations congénita les di st inct. L ' utilisation de modè les transgéniques murins est une faço n idéa le de comprendre les acti v ités transcriptionnell es et morphogéniques dans cette région. No us proposons deux modèles basés sur les facteurs de transcription de la fa mill e Cdx- essenti e ls pour le développement postéri eur de 1 ' embryo n, et des acteurs majeurs dans le développement de la crête neura le (Sanchez-Ferras et a l. , 2014; Sanchez-Ferras et a l. , 201 2). Le premier sera it un modè le pour l' anoma li e qui pourra it être entra înée par la surexpress ion condi tionne lle du gène Cdxl et nous permettra it a ins i de mi eux comprendre le rô le de ce gène dans les cancers déri vés des cellules de la crête neurale. A ce j our, un criblage par PCR nous a permi s d ' identifie r des ce llul es souches po si ti v es pour 1' in tégration d ' une cassette Cdx 1 ma is une confi rmat ion par southern blot es t nécessa ire. Concernant la deuxième souri s transgénique, e lle représente un modèle unique pour étudi er les évènements précoces dans le neuroectoderme postér ieur. En effet, e lle est basée sur 1' uti 1 isatio n de 1 ' enhancer (NSE) de Cdx2 qui permet de générer une li gnée transgénique avec une inducti on spécifique au tube neura l via la technique Cre-LoxP. Cependant, Cdx2N SE s ' es t révélée n' être que partie ll ement acti ve dans le mésoderme. Cette acti v ité mésoderm ale étant mineure, il sera it poss ibl e de reti re r les é léments non-neuraux de NSE afin d ' obtenir une acti v ité enti èrement neurale. Pour déterminer la nature de l' acti v ité des di fférents é léments deNSE, des essa is luc ifé rase ont été effectués par co-transfecti on du facteur pro-neura l Sox2 et d ' une proté ine de fusion Lefl -~caténine
agissant comme effecteur de la vo ie Wnt, ces deux derniers étant importants pour le mainti en de l' identité neuroectoderm ale postéri eure. Les résul tats ont montré que NSE est régul é par Wnt et Sox2 de manière sy nerg istique, ma is les é léments nonneuraux reste nt à identifie r.
Mots-clés: GRN, neuroectoderme, postér ieur, déve loppement, Cdx, régul ati on, outil s
----- --------
>
·x:
AB STRACT
The nervo us system is a fundamental feature of vertebrates whose fo rmation invo lves a compl ex gene regulato ry netwo rk (GRN). The process can be sa id to comm ence w ith the inducti on of the neuroectoderm by Wnt, FGF and BMP activ ity a long the newly fo rmed anterior-poster ior ax is. Severa! transcriptiona l and morphogeni e s igna is in concerted action interna lize the neuroectoderm ventra lly where it beco mes the neura l tube and neural crest, the precursors of the central and peripheral nervo us system, respecti ve ly. The posterio r end of the neuroectoderm has its own unique set of s igna lling mo lecules and di sruptions in the ea rly events in thi s region engend er a di stinct set of congenita l di seases and ma lfo rmations. One way to und erstand the transcriptional and morphogenie acti vity in thi s poste rior reg ion is through the use of mo use mode ! s. Described are two proposed mo use systems based on the genes of the Cdx fa mil y of transcription factors- important fo r posteri or embryonic deve lopment, and key players in neura l c rest deve lopment (Sanchez-Ferras et a l. , 2014 ; SanchezFerras et a l. , 2012). The first would serve as a mode! for the potenti al pathogenes is brought on by Cdxl conditi ona l overexpress ion, in particul ar its potenti al ro le in the oncogenes is of neural crest ce l! de ri ved cancers. So fa r, pos iti ve embryoni c stem ce l! c lones w ith the Cdx l overexpress ion cassette have been identifi ed v ia a preliminary PCR screen but require confirmation by southern blot. T he second mouse mode! would take advantage of a modifi ed neural spec ifie enhancer of Cdx2 to generate a t ruly neural spec ifi e Cre driver line fo r the posteri or ne uroectoderm, a unique and useful mode ! for study ing the ea rly events in the posteri or neuroectoderm. Tniti a lly considered as entirely neura l spec ifi e (Wang and Shashikant, 2007), a Cre mouse line dri ven by the Cdx2NSE later revea led parti a l activ ity in the mesoderm as we il (Coutaud and Pilon, 201 3). S ince gene express ion was minor in the mesoderm, it seems conce ivable that the mesoderm express ion can be removed, resulting in true neura l spec ifie express ion. To do thi s the non-neura l specifi e e lements in Cdx2N SE sequence have to be ide ntifi ed and e ither removed or made non-functi ona l. To thi s end, lucife rase assays were carried out via co-transfecting ne ura l Sox2 and Wnt effectors Lefl-~catenin . Both are important for mainta ining posteri or neuroectoderm identity. Results show the N SE is regul ated by the synergy of Wnt acti v ity and Sox2 . However, the non-neura l e lements have yet to be identified.
Keywords: GRN, neuroectoderm , posterior, deve lopment, Cdx, regul ation, too ls
INTRODUCTION
A hallmark of verte brate embryogenesis is the development of the nervo us system.
T he process begins with the induction of the neuroectoderm across the midline. T he
new structure is not only physica lly and behav ioura lly unique from the surrounding
epidenna l ectoderm but has itse lf two di st inct doma ins: the neural plate bo rder
(NPB), and the neura l plate (NP). T he NPBs, located on the latera l edges of the
neuroectoderm , conta in the precursors of the neura l crest ce ll s (NCC). T he neura l
pl ate, located between the N PBs, encompasses a relative ly large piece of
neuroepithelium that becomes the nascent central nervous system via the fo rm ation of
the neura l tube (NT) . T he neuroectoderm eventua lly invaginates ventra lly to fo rm the
NT and NCC populati on. T he NCCs are a multipotent mi gratory group of ce ll s that
can become peripheral neurons, g li a, pigment ce ll s, cartil age ce ll s and others (Garnett
et a l., 20 12; Perri s, 1997; Stuh 1 mi lier and Garc ia-Castro, 20 12) .
T he transit ion from ectoderm to neura l plate to multipotent ne ural crest ce lls requires
a fin e ly tuned neura l crest-gene regul ato ry network (NC-G RN) whose interactions are
constantly be ing updated, as they are made known (Amore et a l. , 2003; Hinman et a l. ,
2003; Olive ri et a l. , 2003; Van Otterl oo et a l. , 20 13). Work in rev is ing and updating
th is regul atory gene netwo rk is va luable not just to improve our understanding of
neurogenes is in genera l but how changes to this netwo rk can affect deve lopment. One
way to shed sorne light on thi s gene network is through transgenic mouse mode ls.
Mouse systems can prov id e in vivo ev idence on NC-GRN processes both spatia lly
and temporally. Two proposed mouse models to be descr ibed in th is memo ire have
the ir ori g in s in the cauda l end of deve lopment, as they are based on the genes
encoding the caudally-restri cted Cdx family of transcriptio n fac tors. Cdx transcription
2
fac tors are known fo r the ir impo1tance in poste rior embryonic deve lopment, but have
just recentl y been implicated in the initi a l steps of the caudal NC-GRN , acting as
medi ators, inducers and co-operators of known players of the NC regulatory c ircuit
(Sanchez-Ferras et a l. , 201 4; Sanchez-Ferras et a l. , 201 2) .
The first mo use mode! wou ld e lucidate the effects of conditionally overexpress ing
Cdxl . Overexpress ion targetin g NCC deve lopment is be li eved to result in the
pathogenes is of NCC deri ved cancers, such as neuroblastoma. So far, potentia lly
pos iti ve embryonic stem ce ll c lones possess ing the tra nsge ne construct have been
identifi ed via a pre liminary PCR screen but verification by so uthern blot has of yet
been inconclusive. Once veriti ed, mi cro injecti on and rea ring of transgeni c mice
capable of inducible Cdx l overexpress ion co uld proceed . T he second mode! is
env isioned to serve as a Cre dri ver line fo r Cre-LoxP sys tems a iming to study the
early events of the posteri or neuroectoderm . Deve loped in the lab by Co utaud and
Pilon (201 3), Cdx2N SE-Cre was des igned to prov ide Cre express ion so le ly in the
posteri or neuroectoderm . Initia lly characterized as a neural specifie enhancer (NSE)
by Wang and Shashikant (2007) it was later revealed to be parti a ll y active in non
neura l mesoderma l ti ssue (Coutaud and Pilon, 20 13). To address this problem, it was
imperative that the non-neura l cha racteri stics of the enhancer be identiti ed and
removed or made non-functionai thereby preserving a true NSE capabie of be ing
empl oyed as a Cre dri ver line for true neural spec ifi e induction in Cre-Lox P system s.
To tes t the absence or presence of a neura l identity, luc ife rase assays were carried o ut
on a Cdx2NSE-L uc reporte r, co-transfecting neura l Sox2 and Wnt sig na lling effecto rs
Lefl-~catenin . Wnt is important fo r dete rmining posterior structures and regulating
Cdx genes (Pi lo n et a l. , 2006; Prinos et a l. , 2001 ), and Sox2 is a neura l marker of the
neuroectoderm (G raham et a l. , 2003; Uchikawa et a l. , 20 Il ). So fa r, results show the
NSE is regul ated in synergy by combined W nt act ivity and Sox2 . However, we have
not been able to identify the spec ifie areas of the NSE architecture responsibl e fo r the
non-neural activity .
CH APTER 1
BAC KG ROUND
1. 1 Vertebrate embryogenes is
Fo llowing fe tt ilizat ion and c leavage, the third and fo urth major events in vertebrate
development are gastrul ati on and neurul ation. Prior to gastrul ati on, the deve lo ping
embryo is co mposed of a spheri cal layer of ce ll s. Gastrulat ion tra nsforms the s ing le
layer into three di stinct germ layers. K nown as the ectoderm, mesoderrn , and
endoderm , each of the three germ layers esta bi ishes the framewo rk of ali future ·
ti ssues . Moreover, gastrul ati on beg in s to shape the ante ri or-poste rio r (AP) axis and
dorsa l midline (Tam and Behringer, 1997). T hese axes are defin ed by the fo rmati on
of the mesoderm a l der ived notocho rd (Purves and Wi Il iams, 200 1 ).
Neurulation (F igure l.l a) begins w ith the induct ion of the neural plate a lo ng the A P
axis, fo llowed by spec ifica ti on of the neural pl ate borders (NPB). Housed in the N PB
are the precursors of the neural crest ce ll s (NCC) . T he NP B separates the non-neural
ectoderm (NNE) from the neuroectoderm . Next, the neuroectoderm beg ins to bend
inward creating neura l fo lds at the lateral edges. Bending co ntinues inward as the
neura l folds e levate; such is the degree of the bending that the resulting U-shape is
ca ll ed the ne ura l groove. As the g roove lowers the neura l fo lds ri se, meet, fuse and
c lose the loop. T he c losed loop, nowa cy linder of neuroepithe lia l ce ll s known as the
neura l tube (NT) , is freed from the overly ing ectode rm . T he NT is co mpleted first in
the middle of the AP axis, next extending crania ll y and cauda ll y; in bi rds and
4
mamma ls the NT fo nns first in crani a l then trunca l leve ls (F igure 1.1 b) (Co las and
Schoenwo lf, 2001 ; Duband, 201 0) .
A t the same time the NT fonns , NCCs are freed from what used to be the N PB and
they begin to mi grate. NCCs exhibit an epithe lia l to mesenchyma l trans iti on (EMT)
a ll owing them to de laminate fro m the neuroep ithe lium, and migrate on pathways that
w ill spec ify their lineage. De laminati on and migrat ion occurs in a rostra l to cauda l
wave, coinciding w ith NT formation , w ith the wave of migration ultimate ly fi !ling up
four levels : crania l, cardi ac/vagal, truncal and sacra l (F igure 1.2) (Huang and Sa int
Jeannet, 2004; Mason , 2007). In the trunk, NCCs mi grate a round the NT and through,
and sometimes between, somites * and in the head through rhombomeres'i" (Ghysen,
2003; Go mez et a l. , 2008; Guthrie and Lum sden, 199 1 ). Ventral regions are filled -up
first before more dorsa l regions (Weston and Butler, 1966) and depending o n w here
they end up, NCCs ca n g ive ri se- to a diverse set of de ri vatives of mesenchyma l,
neurona l, secretory or p igmented identity (Table 1.1 ) (S imoes-Costa and Bronner,
20 15; Smith and Schoenwo lf, 1997).
There is sti Il some controversy as to w hether NCCs are fate -restricted prior to
mi grat ion, during migrat ion or if they ma inta in multipotency (Jessen and Mirsky,
2005; Kr isp in et a l. , 20 10; Mayor and Theveneau, 2013 ; McKinney et a l. , 20 13). In
vitro experiments have demo nstrated the multipotent capac ity of NCCs to become
multipl e derivatives such as neurons, osteoblasts and me lanocytes (Dupin et a l. , 20 1 0;
Dupin and Sommer, 20 12) . Moreover, recent advances in in vivo cell-label ing and
time-lapse imaging show that pre-migratory NCCs from any location a long the dorsa l
• Segments of paraxial mesoderm th at wi ll form the skeleta l musc les, vertebraes and ribs
t Neuromere segments that become the future hindbrai n
5
NT may contribute to multiple NC targets, and that microenvironment eues along the
migration pathway may be responsible for specifying NCC fate (McKinney et al.,
2013). There is also recent evidence suggesting that most migratory NCCs are
multipotent and only a few in the population are fate-restricted (Baggiolini et al.,
2015).
A
( 1 )
(•)
17
cU
ys
(2)
{b)
19
cla
ys
(cl
20
da
ya
Neu
r.l
lold
(3)
(d)
22
d.y
a
Neu
ral
pla
te
Neu
ral
Pri
mit
ive
str
uk
Cu
t~
of
am
nio
n
told
B Chi
ck a
an
ial r
egio
n (M
idbr
aln
)
Yol
k sa
c:
• ._
_ N
oloc
hO<
d
24 h
(3
s)
28 h
(5
s)
Chi
ck tr
unca
l reg
ion
(Bra
cllla
l)) 4
2 h
(13
s)
48 h
(1
7 s)
52 h
(2
0 s)
54
h
(23
s)
68h
(30
s)
80 h
(3
8 s)
7
Figure 1.1 Steps of neurulation. (A) Dorsal and transverse view of human neurulation. Neurulation entails at !east 3 spatially and temporally distinct stages: (1) formation and shaping of the neural plate (NP); (2) bending of the NP into the neural groove; (3) closure of the neural groove into the neural tube (NT) and subsequent neural crest cell migration. Image retrieved from Marieb and Hoehn (2007). (B) Chick cranial and truncal temporal differences in NT closure. Birds and mammals exhibit NT closure temporal differences depending on the region along the midline. Yellow: presumptive NCCs; Green: delaminating NCCs; Blue: migrating NCCs. Image retrieved from Du band (20 1 0)
CRANIAL ,-:;:::r-~---~==~cranlofaclal cartilage and bona
connective tissue
cranial ganglia
r--~~-lpigment cells
enterlc ganglla
smooth muscle
TRUNK pigment cells
dorsal root ganglia
sympathetic ganglla
adrenal medula
Figure 1.2 NCC migration throughout four different levels. Migration is the first step in NCCs adopting different lineages. In a process known as epithelial to mesenchymal transition (EMT), presumptive NCCs exit the neuroepithelium and take on different migration pathways each one leading to a change in gene expression and thereby a change in cell type (McKinney et al., 2013). Grey: NNE non-neural ectoderm; Yellow: NC neural crest; Blue: NP neural plate. Image retrieved from Huang and Saint-Jeannet (2004)
8
T bi 11 D' 't fN a e . 1vers1ry o eu ra IC tCIId. f res e enva 1ves . Neural crest cell derivatives
Mesenchymal Neuronal Cells Secretory Cells Pigmented Cells ce Us
Chondroblasts Sensory neurons Chromaffin cells Melanocytes
Osteoblasts Cholinergie Calcitonin-neurons producing cells
Fibroblasts Adrenergic Parafollicular cells neurons
Odontoblasts Satellite cells
Cardiac Schwann cells mesenchyme Myoblasts Glial cells
Adipocytes
Adapted from Simoes-Costa and Bronner (20 15).
1.2 Molecular mechanisms
From the transformation of ectoderm to neuroectodenn , to the changing expression
profiles of migrating NCCs, there are a multitude of morphogenie signais, eues and
gradients, both spatial and temporal that are responsible for these very physical and
behavioural changes. This complex network is built on effective communication from
one cell to another, relaying severa) specifying signais at once. This complex network
is known as the neural crest gene regulatory network (NC-GRN) (Meulemans and
Bronner-Fraser, 2002; Oliveri et al., 2003). Neurulation and NCC
migration/specification rely on the NC-GRN (Figure 1.3). The process begins with
the induction of the neural plate via the Wnt, FGF and BMP signal ling pathways
(Huang and Saint-Jeannet, 2004). It is the combination ofthese three signais and their
antagonists that create the gradients responsible for cell specification and
morphogenesis (Niehrs, 2004). The specifies as to how these three signalling
pathways work together in a concerted manner to induce the NPB is not fully
understood and is active! y being researched (Garnett et al., 20 12; Monsoro-Burq et
aL, 2005).
Signaling module
Neural pla border mo
te du le
est Neural cr specificat module
ion-
Neural crest migration module
Neural crest diversification circuits
WNTs BMPs
Notch FGFs
+ .. Zic Ms x Gbx2
Pax3/7 Tfap2 Dlx5/6
,, FoxD3
Pax3/7
Tfap2
Sox10
Pax3/7
Neurons and glia
• .. Snai1/2 Twist Sox5
Sox9 Ets1 Mye
Id Sox10 Myb
1
EMT t.
Sox5 FoxD3 Ebf1
Tfap2 RxrG
Id Snai1/2
Chondroblasts and osteoblasts
Mye
Myb
Melanocytes
Figure 1.3 Neural crest gene regulatory network.
9
-
-
Through different molecular signais, both temporal and spatial dependent, the NCGRN is responsible for the process ofNC development, from the induction of the NP to the diversification ofNC targets. Image retrieved from Simoes-Costa and Branner (2015).
10
1.2.1 Signais of the posterior neuroectoderm
The posterior neuroectoderm receives a different concentration, combination and
timing of signais than the anterior level. The outcome is a later onset of the NT and
NCCs in the posterior region as weil as different NCC lineages. In summary, anterior
structures are patterned first wh ile posterior on es occur later (Durs ton, 20 15).
The patterning of the anterior-posterior (AP) axis is regulated in part by Wnt and
FGF, as weil as Nodal and Retinoic acid morphogens and their antagonists (McGrew
et al., 1997). Higher Wnt and FGF concentrations specify posterior fates, and their
antagonism anterior fates (Figure 1.4) (Kiecker and Niehrs, 2001; Kudoh et al.,
2002). Meanwhile, BMP and its antagonists· pattern the dorsal-ventral (DY) axis
(Marchant et al., 1998; Patten and Placzek, 2002; Smith and Harland, 1992). Higher
BMP concentrations specify ventral fates and lower levels dorsal fates (Endo et al.,
2002; Little and Mullins, 2006). Again, ali three - Wnt, FGF and BMP
morphogens are needed to induce the formation of the NPB (home of the future
NCCs) but Wnt and FGF signais are mostlyt posteriorizing; BMP mostlyt
ventralizing (Hendrickx et al., 2009; Tuazon and Mullins, 2015).
• such as Noggin (Nog), Fo llistatin (Flst), and Chordin (Chd)
t Wnt and FGF s ignaling are crucial fo r AP patteming but they can be in vo lved in DY patterning as they can affect regulation of BMP act ivity
~ BMP is crucial for DY patterning but isoform BMP4 has posterioriz ing effects. ln ail , there is a trend where Wnt and FGF : posteriorizing and BMP : ventralizing, but overl ap may occur
Ante ri or
head
trunk
ta il
Posterior
Dkk sFRPs Cyp26
Lefty Sprouty
1 Wnt
Nodal< FGF
RA
Figure 1.4 AP axis patterning.
11
Higher concentrations of Wnt FGF Nodal and RA activity specify posterior fates; antagonists to these posteriorizing signais specify anterior fates. Image retrieved from Tuazon and Mullins (20 15)
A OFF B ON
LRP516 • Frizzled
~tenltl) TrCP,
p-catenin )
Figure 1.5 The canonical Wnt signalling pathway.
( lkatemn )
( P..catenln ) ( ~-catenin )
Wnt raspanslve-
(A) With no Wnt ligand present, the ~-catenin destmction complex is active and thus no ~-catenin is freely available to mediate Wnt responsive gene transcription. (B) Upon introduction of a Wnt ligand to the Frizzled receptor, the Dishevelled proteins are activated and impede the destruction of ~-catenin. This causes cytoplasmic levels of ~-catenin to rise to where they are then able to enter the nucleus and associate with Lefl /Tcf family of DNA-binding proteins and activate target gene transcription. Image retrieved from MacDonald et al. (2009)
12
1.2.1.1 Wnt and FGF signalling
The canonicat Wnt signalling pathway begins with the binding of Wnt ligands to
Frizzled receptors, which in turn stops the ~-catenin destruction complex. An
increasing concentration of ~-catenin in the cytoplasm favours its translocation into
the nucleus where it can then associate with transcription factors of the Lefl /Tcf
family and activate target gene expression (Figure 1.5) (Angers and Moon, 2009;
MacDonald et al., 2009; Petersen and Reddien, 2009). This modulation of gene
expression by the Wnt pathway is implicated in most stages of embryogenesis
throughout various phyla including deuterostomes, protostomes, and pre-bilaterians
(Petersen and Red dien, 2009; Stuhlmiller and Garcia-Castro, 20 12). In vertebrates,
the Wnt pathway is responsible for cell proliferation, cell fate, and body AP axis
patterning (Hi kas a and Sokol, 20 13). Particular to AP neural patterning is how W nt
specifies caudal CNS cell fa tes (Tuazon and Mullins, 20 15). lnversely, anterior levels
require antagonists of Wnt activity (Houart et al., 2002). For the most part, posterior
markers are tied to an increase in Wnt activity whereas a decrease in Wnt activity
increases an teri or markers (Kudoh et al., 2002; McGrew et al., 1997; Petersen and
Reddien, 2009).
Like the Wnt pathway, the FGF pathway is important in establishing the vertebrate
AP axis and specifying posterior cell fates (Figure 1.4; 1.6) (Tuazon and Mullins,
20 15). Suppressing FGF blacks posterior tissue development; overexpressing FGF
blacks anterior tissue development (Draper et al. , 2003; Isaacs et al., 1994;
Stuhlmiller and Garcia-Castro, 2012). As such, bath Wnt and FGF signais inhibit
anterior gene expression (Kudoh et al. , 2002). Concerning their origin, Wnt signais
come from the non-neural ectoderm and the paraxial mesoderm; FGF only cames
from the paraxial mesoderm (Huang and Saint-Jeannet, 2004) . NPB induction may
require bath Wnt and FGF signais, but it is still unclear ifFGF acts directly to induce
13
Extracellular
Plasma membrane
~ ! Cytoplum
PIP42' PIP3 IP3 • DAG / ! l f 1;) ----~~ PD
~ cal· K
! \ Cytoskeleton Targ ts
Figure 1.6 The FGF signalling pathway. FGF binding to FGF receptors leads to the recruitment of Grb2 and Ras among other effectors to activate the Erkl/MAP kinase pathway which in turn phosphorylates a diverse set of transcription factors initia ting their regulatory activity implicated in ce li growth, migration, and morphogenesis. In AP patterning, antagonists of the FGF pathway, such as Sprouty, work to inhibit posteriorizing fates , thus prompting anterior ones. Image retrieved from Mason (2007).
14
the NPB or indirect) y by activating the Wnt pathway (Garnett et al., 20 12). Severa)
studies by Hong et al. (2008) have suggested that Wnt originating from the paraxial
mesoderm is activated by FGF, indicating that FGF activates mesodermal Wnt which
in turn helps induce NCC formation.
Together, through overlap, direct or indirect methods, Wnt, FGF and BMP signalling
induce the NC precursors at the NPB. The first phase involves Wnt and FGF to
activate the expression of border specifiers su ch as Msxl 12 Pax3/7 and Zic. The
second phase involves Wnt, BMP as weil as BMP-antagonizing Notch signalling to
further specify the NCC, by stimulating the expression of Snai/2, FoxD3, and
Sox9/10 (NC-GRN Figure 1.3) (Huang and Saint-Jeannet, 2004; Tuazon and Mullins,
20 15).
1.2.1.2 Sox2
Important for the development of the neural primordia as well are the Sox-B l
transcription factors: Sox 1, Sox2 and Sox3 (Uchikawa et al., 2011 ). Sox2 expression
in particular is the most extensive, with full expression across the neural primordia
(Okuda et al., 2010). In fact, Sox2 is the most definitive marker of the early neural
plate (Papanayotou et al., 2008; Pevny and Ni colis, 201 0; Rex et al., 1997). From
embryo to adult, Sox2 plays a ro!e in 1naintaining neural progenitor populations
(Brazel et al., 2005; Ellis et al., 2004). Constitutively ex pressing Sox2 suppresses
neuronal differentiation; inhibiting Sox2 results in early neuronal differentiation
(Graham et al., 2003). Suppressing neuronal differentiation in the early
neuroectoderm is essential for maintaining the neural plate identity (Kishi et al.,
2000). In the posterior levels, during neural plate development, Wnt and FGF signais
work together to activate Sox2 expression (Takemoto et al., 2006).
15
1.2 .1.3 Cd'C genes
Cauda l Cdx genes are a lso impli cated in the development of the posterior NP and in
AP patterning (Marom et a l. , 1997), as we il as mediating the c losure of the NT
(Sava ry et a l. , 20 Il a) and players of the trun ca l NC-GRN (Sanchez-Ferras et al. ,
20 14; Sanchez-Ferras et a l. , 20 12). In addition, they play a significant role in gut
development (Beek and Stringer, 20 1 0 ; Si lberg et a l. , 2000).
Cdx genes a re related to the caudal Drosophila gene (cad) (Barad et a l. , 1988;
Mlodzik et al. , 1985). T here are three Cdx genes : Cdxl , Cdx2, and Cdx4. Du ring
embryogenesis, the three are expressed in the ca udal regions , occupy ing a li tissues
surrounding the primitive streak (Hou le et a l. , 2003a). T heir timing and location
a round this area varies; the Cdxl expression domain has the most rostra l reach,
fo ll owed by Cdx2 (F ig ure 1.7). Cdx2 is the earli est expressed, E3.5 at the
trophoectoderm. Cdx2-null mice die at this stage because they fa il to implant
(Strumpf et a l. , 2005). In embryon ic tissues , Cdxl /4 are expressed around E7.5 before
Cdx2 (E8.5). Expression of a li three attenuates as ti me goes on and recedes caudally
until Cdx4 ex pression is go ne at E10.5 and Cdxl /2 express ion rema in loca lized in the
gut where they continue throughout !ife (Lohnes, 2003; S ilberg et a l. , 2000). In AP
patterning, Cdx l /2 are functionally similar despite seq uence variabi lity (Savary et a l. ,
2009b; van den Akker et al. , 2002) . Knockout studies show that Cdxl is important for
AP patte rning (S ubramani an et a l. , 1995); Cdx2 is important for AP patterning and
gut development (Beek and Stringer, 20 1 0; Chawengsaksophak et a l. , 2004) and
Cdx4 has subtle importance in AP patterning (van Nes et al. , 2006).
T he three Cdx prote ins regulate Hox gene express ion, both directly (Beek et al. , 1995 ;
Marom et al. , 1997; Subramani an et a l. , 1995) and indirectly (Savary et a l. , 2009a) .
Tn vertebrates, Hox genes control ax ia l regionalization, as we il as the subdi vision of
the nascent vertebrae (Iimura et a l. , 2009). Stud ies suggest that Cd'C genes regulate
Hox genes by conveying the posteriorizing Wnt, FGF, and RA signais (Be l-Vialar et
16
al. , 2002; Hou le et a l. , 2003 b; Keenan et a l. , 2006; Lohnes, 2003 ; Pilon et a l. , 2006;
Pilon et al. , 2007; Shimizu et a l. , 2005). Wnt in particul ar seems to be important in
the process, w ith combin at ions of Wnt and FGF (Kee nan et a l. , 2006) and Wnt and
RA (P ilon et a l. , 2007) demonstrating the signiti cant contribution of Wnt signa is in
induc ing Cdx expression. Moreover, Cdx genes have been fo und to mediate Wnt
signa l ling in spec ify ing posterior morphogenesis in vertebrates (Sanchez-Ferras et a l. ,
20 12; Shimizu et a l. , 2005 ; Zhao et a l. , 20 14) . Cdx l can regulate its own promoter,
and signifi cantly more so with Lefl (a Wnt effector), suggesting the existence of
Wnt-Cdx joint regulatory complex (Beland et a l. , 2004). Cdx4 can also be regul ated
by Wnt s igna l ling (P il on et a l. , 2006) as we il as by Cdx2 (Savory et a l. , 20 Il b).
17
Cdxl
A
c --E7.5 8.5
Figure 1.7 The spatial and temporal differences of rn urine Cdx gene expression. (A) Cclxi expression begins at E7.5 at the ectoderm and mesoderm of the primiti ve streak. At E8.5, express ion extends to the neural plate, parax ial mesoderm, hindgut and ta ilbud; At E9.5, express ion regresses cauda lly, and lingers in somites and presumptive dermamyotome; at E l2.5 , express ion is limited to the gut where it persists (Houle et al. , 2003a; Meyer and Gruss, 1993 ; S ilberg et al. , 2000) . (B) Cdx2 embryoni c express ion begins at E8.5 in the neural plate, ne ura l tube, part of the notochord, the hindgut and a li ti ssues of the ta ilbud; at E9.5 , express ion regresses caudally in the NT, NP and notochord , hindgut and a li ti ssues of the tailbud more toward the tait bud; express ion is limited to the gut at E l2.5 where it pers ists (Beek et a l. , 1995; Silberg et al. , 2000). (C) Cdx4 express ion begins at E7.5 near the posterior end of the primitive streak; at E8.5-9.5, expression regresses caudally in the mesoderm a nd hindgut endoderm ; at E 1 0.5, express ion ends (Gamer and Wright, 1993 ; Lohnes, 2003). Image retrieved from Houle et a l. (2003a)
18
1.3 Gene regulati on and NC deve lopment
NC deve lopment re li es on a feed-forward system of regulatory eve nts known as the
gene regu latory netwo rk (F igure 1.3) (Me ulemans and Bronner-Fraser, 2004). Vital to
any GRN is proper gene regulation by c is-regul atory reg io ns. T hese regions are so
important that it has been suggested that mutations in pre-vertebrate c is-regulatory
regions were criti ca l for NC evo luti on, and by extens ion the evo luti on of the NC
GRN (Jandzik et al. , 20 15; Van Otterloo et al. , 20 13).
1.3. 1 Enhancers
En hancers are sma ll cis-regulatory e lements of a ro und 200-500bp that can be up to
1.5 Mb downstream, upstream or even intronic of their target gene; (F igure 1.8)
(Epste in , 2009; Rada-Iglesias et a l. , 20 1 3). T he ir function is to enhance target gene
expression by activating the promoter of the target gene (Pennacchi o et al., 20 13).
T ranscription factors bind to specifi e regulatory motifs on the enhancer and mediate
activation of the target promoter. In nature, an ac ti vated enhancer e 1 icits a
co nformat ion of the chromat in structure so that the en hancer ca n loop near the target
promo te r and mediate its act ivation (Kranz et al. , 20 Il ).
During development enhancers play a role in ce llul ar g rowth, differentiation and
mi grat ion by re iay ing the activator s ig nai s from growth facto rs and transcriptio n
factors (Howard and Davidson , 2004; Kranz et a l. , 20 1 1 ). In the context of the NC
GRN , for exampl e, enhancers of NPB specifï ers, such as Msx, Pax3/7, Z ic 1, Dlx3 /5
wou ld be at the receiving end of effectors of Wnt, FGF, BMP and otch signa lling.
T hen, these NPB spec ifier transcription factors , a long w ith other s igna is, wou ld be at
the delivery end of the enhancers of NC spec ifiers, such as Z ic3 , Sox9, Foxd3 ,
Sox 10, Snai2, and Twist, and so on until completion ofNCC differenti ation.
19
TSS
B R pp lE
Figure 1.8 Types of cis-regulatory elements. From left to right, BE: boundary element; E: enhancer; R: repressor (the counterpart of the enhancer-instead of promoting the activation of a target promoter they suppress activation) PP: proximal promoter; CP: core promoter; TSS: transcription start site; JE: intronic enhancer. E1/2 represent exons. Image retrieved from Epstein (2009)
Also of note regarding the current NC-GRN, as it relates to enhancer activity, is the
absence of known overlying epigenetic regulation needed for enhancer activity to
occur in the first place. In fact, several epigenetic modifiers regulating NC
development have been identified, and constitute a fairly recent area of active
research (Hu et al., 20 14; Liu and Xiao, 2011 ).
1.3 .1.1 Approaches for identifying and tes ting enhancers
According to Simoes-Costa and Branner (2013), there are three main methods used in
identifying enhancers (Figure 1.9). Ail three require: the construction of the putative
enhancer sequence with a minimal promoter to drive the expression of a reporter, and
the tes ting of the putative enhancer via activation through ce li transfections, or in vivo
stable or transient inducible animal models .
20
The first method requires inserting severa! non-coding sequences, one by one, into a
reporter construct in the hopes of identifying an enhancer. This method has the
potential advantage that random screening might find activity the other more focused
methods do not detect. However this method is out-dated, costly, more uncertain, and
laborious.
The second method screens only non-coding sequences with high vertebrate
homology. White homology does not necessarily equate to an enhancer, it is an
important indicator as highly conserved regions have been found to be enhancer
enriched (Bejerano et al. , 2004; Pennacchio et al. , 2006). Therefore, this method has a
significantly better chance of identifying an enhancer than the previous method. In
addition, a lot of sequencing data from numerous vertebrate species is freely and
readily available.
Lastly, the third method screens areas where ChiP-seq has identified areas of histone
modification and therefore possible sites with enhancer activity (Rada-Iglesias et al.,
20 12). This method relies on the temporal qualities of histone modification in a
specifie cell type, and can therefore target a specifie moment and axial leve! in
potential enhancer activity. If the Chip-seq datais not currently available, this would
add additional time and cost if it needs to be obtained.
APPROACHESFORENHANCER IDENTIFICATION
A. Screening of non-coding regions
'\ \ .\ ' ' \ ' \ \ \ .,
' \ . . \ \~ \, i
...
1 //,/ ! 1 .j. .. .j
B. Evolutionary conservation
C. Profillng of histone modifications
TEST ENHANCER FOR ACTIVITY
Cell transfection
Stable transgenesls
Figure 1.9 Techniques for identifying and testing enhancers.
21
First step: build a putative enhancer reporter construct, with the enhancer obtained from (A) Using a blind screen of non-coding areas; (B) Using a genome browser to identify conserved regions; (C) Using Chip-seq data to uncover areas of histone modification Second step: testing the enhancer by cel! transfections, transient and/or stable transgenic models. Image retrieved from Simoes-Costa and Branner (2013),
------· --------------------------------------------------------------------------------------
22
1 .3 .1.2 Characterization of binding motifs
While the previous methods can identify the existence of an enhancer, the location of
activity, and even test severa! potential TF activators, the characterization of binding
motifs requires different steps. Bioinformatic tools can be used to identify potential
binding motifs but characterization is not addressed until the binding of transcription
factor to DNA is demonstrated. Electrophoretic mobility shift assays (EMSA) are
often used in promoter/enhancer studies to reveal the binding of transcription factors
(Pares-Matos, 2013). A DNA pull-down assay followed by Mass-Spectrometry can
also be used to identify DNA-binding TFs (Drewett et al., 2001; Hubner et al., 2015) .
1.3.2 Cdx2NSE
The enhancer identification method, using evolutionary conserved regions, described
above and in Figure 1.9b, was used in the discovery of a neural specifie enhancer
(NSE) of Cdx2. Stable transgenesis showed reporter (LacZ) activity in the neural tube
(Figure 1.1 0) (Wang and Shashikant, 2007).
1.3 .2 .1 Cre-LoxP application of Cdx2NSE
As weil as providing evidence of regu latory interactions, enhancers can be used in a
variety of appiications, inciuàing fate mapping, time-iapse imaging and targei.ed ioss
of- and gain-of-function assays (S imoes-Costa and Bronner, 2013). One use of the
Cdx2NSE is to exploit its neural specifie activity and caudal localization in the
developing vertebrate. Cre-LoxP mouse systems in particular, are a great tool and
way of exp loi ting the temporal and spatial qualities of an enhancer. In this system, the
enhancer drives expression of Cre-recombinase in a transgenic line of mice. This
enzyme can recognize and excise DNA flanked by loxP sites. Conditional knockouts
are generated from the successful cross of a tissue specifie Cre-expressing line and a
line carrying a floxed (loxP-flanked) gene of interest (Lodish et al. , 2000) .
23
Cdx2
Figure 1.10 Identification and testing the Cdx2 neural specifie enhancer (NSE). Cdx2NSE was identified by its high vertebrate homology, and tested by a transgenic line canying the presumptive enhancer driving the activation of a LacZ reporter. Expression seems confined to the neural tube. Images retrieved from Wang and Shashikant (2007)
Sox2 YFP Sox2 YFP
B B' B"
r,· c C' C"
f', 1·'11
' . !
.,, D D' D" ,. 1
E E' E"
Figure 1.11 The enhancer is not entirely neural specifie. An e8.5 embryo obtained from a cross between Cdx2NSE-Cre driver line and R26RYFP had Cre and therefore YFP expressed mostly in the neuroectoderm with sorne present in mesodermal cells. Compare YFP to neural marker Sox2. Image retrieved from Coutaud and Pilon (2013)
24
Here, the gene of interest is removed in specifie Cre expressing tissues. Another use
of the system is to have loxP sites impeding the expression of a gene of interest. In
this example, the successful recombination leads to the tissue-specifie expression of
the target gene.
Seeing as how no such Cre-driver line targeting the posterior neuroectoderm existed,
Coutaud and Pilon (2013) decided to generate such a line using the Cdx2NSE.
However, once the Cre-driver line was created and tested it was revealed that the
enhancer had minor activity in the mesoderm as weil and was therefore not entirely
neural specifie (Figure 1.11).
1.4 Diseases ofNCC origin and neural tube defects
A better understanding of the NC-GRN and associated epigenetic mechanisms is
important because disruptions in these processes are believed to be the source of
severa! human congenital birth defects and cancers. Relatedly, mechanisms affecting
proper neurulation can lead to severa( neural tube defects (NTDs ). The result of these
regulatory disturbances vary, but may drive changes affecting cell-fate/programming
decisions, differentiation timing and/or migration patterns.
1.4.1 Neurocristopathies
Neurocristopathies are NC-derived developmental anomalies comprising over 700*
known syndromes and defects (Trainor, 201 0). Sorne of the best-known birth defects
due to NC malformation result in craniofacial defects, heart defects and agang1ionosis
of the colon. Many of these defects affect severa( known NC-GRN factors , such as
SoxlO, Pax3, Snail2 in Hirschsprung's disease and Waardenburg syndrome (Kim et
' For craniofacial defects a lone
25
al., 20 Il ; Para tore et al. , 2002), or Sox9 and Twist in CHARGE syndrome (Hu et a l. ,
20 14).
1.4. 1.1 Neura l crest cell-derived cancers
NCCs possess many of the same attr ibutes as metastatic ce ll s, such as migration and
invasion. T he assumption then is that ce ll s of NC origi n restatt many of the same
regulatory mechani sms needed for embryoni c development (S imoes-Costa and
Bronner, 20 13). Two of the most we ll-known cancers of NC lineage are
neurob lastoma and me lanoma.
Neuroblastoma arises from ce l! types of the sym patico-adrena l NCC lineage an d is
the most com mon so l id childhood tumour accounting for 7-l 0% of paediatric cases
(Brodeur, 2003; Jiang et a l. , 20 11 ; Schu lte et al. , 2013). The change from NCC to
mature sympathet ic gang li a, invo lves an epigenet ic switch needed to s il ence two
genes that promote cel! cycle progress ion and ce l! death avers ion- respect ive ly,
Mycn and Arid3b- the same switch left on is believed to lead to the oncogenes is of
neurob lastoma (Kobayas hi et a l. , 20 13). Mycn in particular has been shown to be
expressed in ear ly NCC during ventra l mig ration and in NCC undergo ing neuro nal
differentiation (Wakamatsu et a l. , 1997).
Studies have pointed o ut that the progress ion of metastat ic melanoma may invo lve
severa! genes of the NC spec ifi cat ion modul e (of the NC-GRN) such as Twist, SoxJO,
Slug and FoxD3 (Shakhova et a l. , 20 12; Shirley et a l. , 20 12) . S lug is needed by NCC
for EMT (Pe in ado et a l. , 2004), and Sox 10 for earl y me lanob last formation and
postnata l ma intenance (Harris et a l. , 2013 ; Potte rf et al. , 200 1 ). Twist 1 promotes
invas ion of melanoma cells by mediat ing the regulation of a matri x metalloproteinase
(MMP), an enzyme needed to breakdown the extracellular matrix (Weiss et al. ,
20 12). Twist in particular is overexpressed in many so l id tu mours ofNCC ori g in , not
just melanoma, but also g lioma and neuroblastoma, highlighting its imp01tance in the
metastatic process (Yang et a l. , 2004).
26
1.4.2 Neura l tube defects
Neura l tube defects a re congenita l deformi t ies brought about by an inadequate c losure
of the neura l tube. Prevalence of NTDs is f rom 0.75 to 1.1 2 per thousand, depend ing
on ethni e ori g in (Feuchtbaum et a l. , 1999). NTDs are compl ex in that a diet ri ch in
fo li e ac id may we il be an impotiant fac tor in prevention, as public health data seems
to indicate (Cze ize l et a l. , 20 1 1; Erickson, 2002; Pitkin, 2007). Interest ing ly, there is
a lso ev idence indicating that the NCC may a lso be protected by fo late
suppl ementat ion, suggesti ng the neuroepi the lium may be sens iti ve to fo late-re lated
pathways (A nto ny and Hansen, 2000; Li et a l. , 20 Il ). Other poss ible NTD ri sk
factors may inc lude environm enta l exposure to teratogenic agents such as heavy
metals, organic so lvents and agric ul tura l chemi ca ls (Seve r, 1995). Depending o n
where the anoma ly occurs the resul t is usua lly ::,pina bifida or anencepha ly, a lthoug h
there are severa! others. Improper c losure near the ca uda l neuro pore leads to spina
b{fida and anencepha ly in the rostra l neuropore . T he degree of affli ction vari es, as
does neuro logica l dysfunction. In spina bifida cogni tion, behav ior, ve rtebra l co lumn
and organ systems can be affected depending on severi ty (F letcher and Brei, 201 0).
Anencepha ly is usually more severe, because the neura l fo lds of the bra in rema in
open (Co pp and Greene, 20 13) leav ing a large port ion of the bra in undeve loped.
Surviva l rate is very low, i.e. 40% afte r 24 hours (Ba ird and Sadov ni ck, 1984). Proper
c losure of the c rania l and spinal neural tube depends a lot on the bending of the
neuroepi the lium at median (MHP) and dorsa l-late ra l hinge po ints (DLHPs). MHP
and DLHPs are both features of c rania l and intermediate sp inal neurul ati on. However,
MH P is a fea ture of upper spinal neurulat ion and DLHP of the lower spi ne (Copp et
a l. , 2003; Yamaguchi and Miura, 201 3; Ybot-Go nzalez et a l. , 2007). BMP antagoni st
Noggin is needed fo r crani a l as we il as lower and in te rmediate spinal DLHPs
(Stottmann et a l. , 2006). Co nverse ly , the upper spine MHP re lies on inhibi tion of
Noggin by Sonic-hedgehog, hence hi gher BMP leve ls (Ybot-Gonzalez et a l. , 2002;
Ybot-Gonza lez et a l. , 2007).
27
1.4.3 Too ls for studyi ng neurocristopathies and NTDs
Many of the molecular mechanisms so far described have been c larifi ed through the
use of mutant mo use models (Britsch et a l. , 2001; Dixon et a l. , 2006; Shakhova et a l. ,
20 12; Southard-Sm ith et a l. , 1 998; Weiss et a l. , 1997). Dom (Dom ina nt megacolon)
mice are mutant (chromosome 15) mode ls for Hirschsprung' s disease that arose
spo ntaneous ly, i.e. were not purposefully engineered, and they were in strumental in
demonstrating the importance of Sox 1 0 for development of the peripheral/enteric
nervous system (Puliti et a l. , 1995 ; Southard-Smith et a l. , 1998). Splotch mice a lso
arose spontaneo us ly (Russe l, 194 7), with a mutation in chromosome 1 (Epstein et a l. ,
199 la) and are often used as mode ls for NTDs (Moase and Trasler, 1992) . In
homozygous mice, the mutation is embryonic lethal by E14 (Li et a l. , 1999),
exhi biting NT and heart defects (Auerbach, 1954; Conway et a l. , 1997). T he
heterozygous Splotch line carries one nul! Pax3 a lle le, and is v iable havi ng minor
pigmentation abnormalities. ln the early 90 ' s when the Pax3 deletion was found to be
the cause of the mutation, it quickly indicated the importance of Pax3 for norma l
neural development (Epste in et a l. , 199 1 b; Gould ing et a l. , 1993).
Co nditi onal knockouts (KO) have also been used, many exploiting the NC-specific
Wntl-Cre transgenic line (Danie lian et al. , 1998) . Wnt1 is active in the embryonic NT
and migrating NCC, and is important for the formation of the midbrain- hindbrain
boundary (Eche lard et a l. , 1994; Lewis et a l. , 20 13). Us ing this NCINT spec ifie Cre
driver line, Dudas et a l. (2004) knocked-out a BMP receptor (A ik2) w hi ch then
produced crani ofacia l defects including cleft palate and mandible malformation and
suggested the importance of A lk2 in regulating the fo rmat ion of specifie crania l
features . Akiyama et a l. (2004) employed it to generate conditiona l Sox9 KO embryos
to examine cardiac NCC and the essential role of Sox9 in heart EMT. Degenhardt et
a l. (201 0) used it to rescue !ost Pax3 expression spec ifically in the NC, and
demonstrated the redundancy of Pax3 " neural crest enhancers" . Other Cre-dri ver
28
1 ines used include Nestin-Cre fo r the neuroectoderm (Dubo is et al. , 2006; Tronche et
a l. , 1999) and Pax3pro-Cre fo r the dorsa l N T and NC (Li et a l. , 2000).
O verexpress ion knock-in mutants are a lso instrum enta l mode ls fo r di sease . Gene
amplification in particular is known to be the di rect cause of many known human
di seases and ca ncers (Santarius et a l. , 201 0; Shastry , 1995). For exampl e, the c lass ic
ma instay mouse model for neuroblastoma is the T H-MYCN , where in Mycn
express ion is dri ven by the T H rat tyros ine hydroxy lase promoter (We iss et a l. , 1997).
As prev ious ly di scussed, Mycn is norma lly expressed in mi grating NCC switching to
neurona l lineage (Wakamatsu et a l. , 1997). At the tim e, MYCN was kn own to be a
proto-oncogene of hum an neurobl astoma but had not yet been tes ted in an animal
mode l. We iss et a l. ( 1997) prov ided direct in vivo ev idence that Mycn amplification
can contribute to the oncogenesis of neuroblastoma. T he TH-MYCN transgene is
integrated in chromosome 18 in a di sta l region, and the effect it has in di sturbing thi s
region is unknown. Due to thi s uncerta inty , A lthoff et a l. (2 0 15) created anoth er
Mycn-dri ven neuroblastoma mouse mode l targe ting the ROSA26 locus (chromosome
6) . Thi s locus has the advantage that it is ubiquitously expressed, di srupti ons in the
locus have no phys ical effect in mi ce, and can be ta rgeted w ith high effi c iency
(Sori ano, 1999; Zambrowicz et a l. , 1997). As of 2010, the re were a lready over 130
knock-in iines targeting the ROSA26 locus (Caso ia, 20 i 0) .
Ali these mutant models have the ir own limitations, and adva ntages, and yet they
share the fact that the di srupti ons they crea te he lp fue l our understanding of the
mo lecul a r mechani sms in vo lved in the deve lopment ofNC and NT anoma lies.
CH APTER TI
HYPOTHESES AN D OBJECTIV ES
2. 1 Hypotheses
Conditi ona l overexpress ion of Cdxl plays an oncogeni c ro le in the deve lopment of
neura l crest ce ll deri ved cancers in envisioned mo use mode l.
T he intronic neural spec ifi e enhancer (NSE) of Cdx2 co ntains identifiable non-neura l
spec ifie act ivato r binding sites .
2 .2 Objecti ves
Genera te a mo use mode! carry ing the Cre-inducible Cdx 1 overex press ion trans gene.
T he steps requi red to genera te the mode!: Construct ion of the targeting vector,
transfecti on of the vector in to ES ce ll s, select ion of the co loni es, verifica tion of
pos iti ve c lones, microinj ecti on and rearing.
ln vestigate regul ation of Cdx2NSE, and identify act ivato r-binding s ites w ith non
neura l spec ifie characteri stics . Test ing th e regul ati on would g ive us a bette r
understanding as to how the NSE architecture co uld be modified to drive truly neura l
spec ifi e express ion in a new Cre-dri ver line. Luc ife rase assays us ing neural Sox2
co uld identi fy w hich areas are potentia lly neural spec ifi e.
CHAPTER Ill
MATERJALS AND METHODS
3. 1 Construction of pROSA26-Cdx 1, ES targeting and verificati on
3. 1. 1 Vector construction
To make the pROSA26-Cdx 1 targeting vector, a PGKneoPA-Flag-Cdx 1-IRES
EGFP-BGH cassette was in serted into the PacT and Ascl s ites of the pROSA26-PA
targeting vector (Srini vas et al. , 2001). pROSA26-PA was a g ift from Frank
Costantini (Addgene plasmid # 2 L271) . lt is a vector used to target the ubiquitous
ROSA26 loc us, and contains the 3 ' and 5 ' ROSA26 arm s necessary for homologous
recombination (see Figure 4.2) as we il as a diphtheria toxin gene (PGK-DTA) for
negative selection in ES ce ll s. The final construct was digestion and sequencing
verified. ln total, the final construct contains a loxP-flanked PGKneoPA cassette
capable of kanamycin resistance expression in E.coli and eukaryot ic promoter PGK
for expression of neomycin resistance in the mammalian ES ce ll s. Next the N
terminal FLAG-tag is upstream of a full Cdxl ORF (807-bp) as previously described
(Be land et a l. , 2004), followed by an IRES-EGFP reporter seq uence.
3. 1.2 Target ing ES ce ll s
RI ES cells were cultured on mouse embryo fibroblast (MEF) feeder cells that were
mitotically inactivated by 1 Oflg/ml mitomyc in C treatment. The cul turing of the ES
ce ll s on MEF feeder ce ll s was done on ge lat in-coated plates to improve conditions.
Also, the ES medium was supplemented with leukemia inhibitory factor (LIF) to
reduce possible differentiation of ES ce lls. The pROSA26-Cdx 1 vector was
32
linearized by Sacll and ( 12.5 ~t g DNA) electroporated (250Y, 500 ~F, in a 4mm
cuvette) into the R 1 E S ce ll s. Se lection of stable integrants was done w ith 200~g/ml
0418 for one week. Genomic DNA was extracted and iso lated . Screening of
successful homologo us recombinatio n was carri ed out first by PCR us ing an extern a l
forward ROSA26 primer and an interna i reverse primer to create a 1.1 kb fragment
(Table 3.1 ). Three c lones were identifi ed. Co nfirmat ion by So uth ern blot, albe it
inconc lu sive, was fo llowed soon after.
Table 3.1 Primers for PCR screen. PRIM ERN AM E 1 LEN GT H 1 SEQ UENCE
ROSA26TA RG-F1 1 23 I AAGAAGAGGCTGTGCTTTGGGGC ROSA26TARG-RJ 1 20 JAGGGCGGCTTGGTGCGTTTG
3.1.3 So uthern Blotting
Around 20 ~g of genom ic DNA was di gested w ith EcoRI and Kpnl and separated o n
a 0.8% agarose ge l. T he ge l was then depurinated, denatured and neutra lized .
Afterward s, the DNA was transferred over night onto a Hybond membrane
(A mersham) via a s implifi ed downwa rd capill a ry system, i.e. w itho ut added transfer
buffe r.
Once transferred, the membrane was rinsed with 2x SSC, dri ed, UV cross-iinked and
pre-hybridized . The pre-hybridizati on buffe r so luti on co nta ined salmo n sperm
ssDNA to block non-spec ifi e sites. After pre-hybridiza ti on, the membrane was
hybridized w ith probe ss DNA.
Two DNA probes were prepared to target the 5 ' o r 3 ' ROSA26 ends (Figure 4.2), at
25ng in 45 ~1 of TE buffe r. T he probes were [a.-32P] dCTP-labe ll ed using the
Rediprim e Il labe lling kit (Amersham) , and purifi ed by sepharose co lumns. The fin al
pro be concentrati on was approximate ly around 2.8ng/mL of hybridization buffer, as
per manufacturer' s in structi ons. Hybridiza tio n occurred over ni ght at 65°C. T he
33
membrane was la ter washed at least 2 times (2x SSC + 0. 1% SD S) before plac ing in
an exposure cassette fo r l-7 days .
3.2 Testing the neura l specifie regul ati on of Cdx2NSE
3.2 .1 Pl asmids
T he full Cdx2N SEforwa rd-LUC and Cdx2NSEreverse-LUC reporter constructs were
created in our lab (Co utaud , 20 13) and were readily ava il able . T hey conta in an 852-
bp Cdx2NSE sequence upstream of an Hsp68 minima l promoter in a luciferase gene
exp ress ion vector. Four Cdx2NSEreverse fragments were created by PCR w ith the
f01·ward primers conta ining a Hindlii s ite and reverse primers conta ining a Kpnl e ut
s ites (Table 3.3). PCR products were amplifi ed w ith Taq po lymerase (Feldan) and
c loned into a pG EMT-easy vector (Promega) and verifi ed by sequencing. The Hindlll
and Kpnl restri cti on s ites were then used to c lone into an Hsp68-LUC repo rter vector.
T he res ulting fo ur fragment NSE reporter constructs can be seen in Figure 5.4, and
include the N SE sequences conta ining putati ve Lef1 /Tcf sites : 4-3-2-1 (676-bp); 3-2-
1 (607-bp); 2-1 (43 7-bp); and 5 (1 93 -bp). T he construct w ith LEF/TCF s ite « 1 »
(2 19-bp) was made by d igesting Cdx2NSEreverse-LUC w ith BamHT to remove a
666bp fragment containing potenti a l Lefl /Tcf binding sites 5-4-3 -2. PGL3-0T served
as pos iti ve contro l in transfecti ons; it conta ins 3 copies of w ildtype Tcf-4 binding
sites and was a g ift from Bert Yogelstein (Addgene pl asmid # 16558) (Shih et a l. ,
2000) . Negati ve contro l was an empty vector conta ining just the promoter and
luciferase gene. Serving as the acti vators for the transfecti ons were Sox2, and Lefl
~catenin .
3.2.2 Luc iferase assay conditions
P 19 cells were cultured at 37°C (5% C0 2) in a -MEM supplemented w ith 7 .5% heat
inacti vated bov ine ca lf serum and 2.5% heat-inacti vated fe ta l bov ine serum (Wi sent) .
Approximate ly 2 hours prior to transfection, the P1 9 ce ll s were pl ated in 24 we il
pl ates with a ce li density of 3x 104 ce ll s/we ll and incubated at 37°C (5% C0 2) .
34
Co-transfection was carried out to test the effect of Sox2 and Lefl-~caten in
(indi vidua ll y and together) on the full Cdx2NSE forwa rd and reverse reporter
sequences. Cdx2NSEforward-Luc and Cdx2NSEreverse-Luc reporter DNA was
mai nta ined at 1 Oûng/we ll , as was the negative and positive contro l DNA. On the
other hand , Sox2 and Lefl-~catenin act ivator DNA ranged from Ong, 25ng, SOng or
1 Oûng/well . The total amou nt of DNA per weil was kept at 300ng. Tf needed, an
empty IRES-GFP express ion vector was used to compl ete this amount. GeneJu ice
was the transfection reagent (Novagen) and the ratio of reagent to DNA was 3~tl to
1 ~Lg DNA, as recommended, i.e . each weil had 0.9~tl of transfection reagent and
300ng of total DNA. This set of transfections was performed at !east three times in
triplicate.
Assay conditions were similar for the tests w ith the truncated Cdx2NSEreverse DNA
reporter seq uences. 1-!owever, thi s set underwent two key d ifferences in conditi ons:
the ce ll density used and activator DNA concentrat ion. Ce ll density was kept at
1.5x 104 ce lls/we ll and act iva tor DNA (Sox2 and/or Lefl- ~catenin) was kept at Ong or
1 Oûng/we ll. T hi s set oftransfect ions was performed at !east six times in triplicate.
A li transfections were incubated in the a -MEM FBS+CBS media for about 48 hours
at 37°C before performing post-transfection tests and analyses. After 48 hours,
transfection effic iency was visua lly assessed by m1 croscopy, 1.e. identi fy ing GFP
expressiOn. Next, tran sfection analys is (ge ne express ion) was quantified by
lumino metry. The steps fo r this included: med ia removal , rinsing we ll s w ith
phosphate buffered saline 1 X, and lysing the cell s. In a tube, 20~tl of the ce li lysate
was added to 1 00~1 of luc iferase assay buffer. Then , 50~1 of luc iferin was delivered
before placing tubes in the luminometer,
Luciferase activity was expressed as fo ld activat ion relative to the appropriate
reporter vector a tone. Each independent experiment was carried out in triplicate, at
!east three times .
35
3.2.3 Luciferase assay test and stat istica l ana lys is
A Il the results of the luc iferase assays are expressed as means+S.D. , and the
differences in luciferase activity are expressed as fo ld activat ion. The differences in
fo ld activat ion were examined by student's t-test for two-group comparisons. One
way ANOVA was performed to identify differences in multiple group comparisons
Graphs and a li statistica l ana lyses were done using GraphPad Prism version 6
software, and p-va lue ranges, wordi ng and aster isk representation follow the
Graph Pad stat ist ics gu ide (Tab le 3 .2).
T bi 3 2 P "d a e -va ue gUI e P VALUE DESCRIPTION REPRESENTATION
< 0.0001 Extremely significant **** 0.0001 to 0.001 Extreme ly significant *** 0.001 to 0.01 Very signifi cant ** 0.0 l to 0.05 Sign ifi cant * > 0.05 Not s ignificant NS
36
T bi 3 3 or a e . 1gos to construct Cl 2NSE {. X l f reverse stranc ragments. PRIMER LENGTH SEQUENCE (5 ' to 3') NAME Cdx2NSE- 31 AAGCTTACAGAATGCTGGCCAGGAACTGTTC antisense 4-3-2-1 LEFl!TCF-F Cdx2NSE- 35 AAGCTTTAGTGCTCAGTGCCTAGTTGAACAAC an ti sense CAG 3-2-1 LEFI/TCF-F Cdx2NSE- 29 AAGCTTGGGAACACAGTCGCAAACAATGC antisense 2-1 LEF I/TCF-F Cdx2NSE- 30 GGTACCGTGCTCTAAGAGCAGCATCCGTTC antisense-R Cdx2NSE- 35 AAGCTTCTATCCTGACCAAGTGACCTGTGATC an ti sense ATT 5 LEF I/TCF-F Cdx2NSE- 26 GGTACCGAACAGTTCCTGGCCAGCAT antisense-R
CHAPTER IV
TOWA RD THE GENERATION OF MlCE CARRY1NG T HE CRE-1N DUCIBLE CDXJTRANSGENE
4.1 Introduction
Disrupti ons in NCC deve lopment can cause severa! neurocristopathies and cancers,
inc luding crani ofac ia l defects, heart defects, co lon agang li onos is, neuroblastoma, and
me lanoma among others (Ha ll , 1999; T ra inor, 201 3). The aim of this proj ect was to
generate a knock-in mouse capab le of conditional Cdxl overexpress ion. Cons ider:ing
the key ro le Cdx prote ins pl ay in N C deve lopment, cond itional overexpress ion of this
key N C-GRN player co uld engender the pathogenes is of a NC deri ved cancer. NCC
deri ved tumours, such as melanoma and neuroblastoma, exhibit overexpress ion of
genes needed fo r NC development, in part icul ar those affecting EMT and mi gration
(Trainor, 2013) . It is the hope that thi s tool w ill he lp identi fy prev ious ly unknown
regul atory components linking the contributi ons of Cdx neural functions to
mali gnancy.
4.2 Results
4.2.1 Constructi on of the targeting vector
T he ROSA26 target ing vector was chosen to knock-in the transgene into the mouse
genome in the ROSA26 locus (mouse chromosome 6) (Sori ano, 1999; Srinivas et al. ,
200 1 ). Targeting th is locus en sures the transgene wi ll be stab ly integrated in one
place and will be ubiquitous ly expressed by an endogenous ROSA26 promoter. In
addition, di sruptions to the locus do not affect mouse viability or ce l! phenotype.
38
With this in mi nd, we used a vector containing the bicistronic FLAGCdx l -IRES-GFP
cassette th at had previously been c loned into a B igT vector. In this vector, a tloxed
PGK-neo-tpA cassette functions as a stop cassette for downstream FLAGCdx 1 -IRES
GFP. The fragment containing loxP-PGK-neo-tpA-I oxP-FLAGCdx 1-IRES-GFP-bpA
was then cloned into the PacT and Ascl restriction sites of pROSA26PA vector
(F igure 4.1) (Sr ini vas et al. , 200 1 ). The final 15-kb construct was digestion and
sequence ver ified. Once confirmed, 12.5 flg of Sacii- Iinearized targeting construct
DNA was e lectroporated into RI ES ce lls.
4.2.2 Target ing the ROSA26 locus
After electroporat ion, the ce ll s were expanded 1n medium contai ning the select io n
agent G418 fo r a week. Genom ic DNA from 96 co loni es res istant to G4 18 was
extracted and ana lyzed by PCR (F igure 4.3) and Southern blot (not shown). Figure
4.2 demonstrates the entire targeting process. The targeting vector pROSA26-Cdx 1
linearized and transfected targets the mouse ES ce ll ROSA26 locus usi ng the ROSA26
5' and 3' tlanking arms. Successful recombination was screened by F I and RI PCR
primers; three successful events were detected out of 96 colonies (Figure 4 .3). The
next step was to use Southern blot to further ve ri fy successful transgene targeting.
Here the genomic DNA of the 3 PCR-verified clones was digested by EcoRI and
Kpni , and the use of the 5' and 3 ' probes would detect a 4-kb an 8.8-kb fragment
w here the targeting event was successful , and in 11-kb fragment in the w i Id-type.
Fo ur success ive Southern blots were carri ed out yet failed to even detect the contro l
(WT).
IRES-EGFP
EcoRV
PGKp-Neo pA
Rosa26 3'
1 Kb plus DNA ladder
15Kb-----+
Figure 4.1 Sacll lioearization of pROSA26-Cdxl Linearization of the pROSA26-Cdxl targeting vector by Sacll
39
40
Eco RI Kpnl
Targeting construct
3'ROSA26 sequence
loxP loxP
WT EcoRI x Xba l x Kpr1 l
ROSA26 locus 8
Il K 0-----ŒJ
S' probe J' probe
Eco RI Kpr1 l ROSA26
r.-om-1--K=d<-lo~ allele 1 S'
loxP loxP <:=>
S'probe RI J' probe
-4 Kb ------1 8.8 Kb ~ !.
Figure 4.2 Depiction of gene targeting at the ROSA26 locus. At the top is the linearized targeting construct pROSA26-Cdx 1. In the middle is the wild-type ROSA26 locus, with exons 1, 2 and 3. At the bottom is the targeted knockin allele containing a floxed-PGK-neo-tpA cassette which functions as a stop cassette, interrupting the expression of a bicistronic cassette containing FLAG-tagged Cdxl and IRES-eGFP. PCR screening of successful knock-ins by the externat FI and internai R 1 prim ers give a 1.1-kb product. Southern blot screening uses externat ROSA 5' and 3 ' probes. ES cell genomic DNA digestion by EcoRI and Kpni permits the 5' probe to bi nd to a 4-kb fragment and the 3' to an 8.8-kb fragment where th targeting event is successfu l; in the WT allele the 2 probes would bind to an 11-kb fragment.
~-------------------------------------------------------------------------------------· -
Targeted
1 Kb plus DNA ladder
1.1 Kb )
Targeted 1.1 Kb >
Targeted 1.1 Kb >
PCR (F1-R1)
Figure 4.3 PCR screen of targeted ES clones.
41
*
*
*
PCR scree n of 96 clone DNA samples yie lded 3 correct ly targeted events (marked w ith an asterisk).
CHAPTER V
TESTJNG THE NEURAL SPEC LFIC REGULATION OF CDX2NSE
5.1 Introd uct ion
As previous ly discussed , Cdx genes integrate several of the posteriorizing signa is
important for development including Wnt, FGF, and RA s ignais (Be l-Via lar et a l. ,
2002; Houle et a l. , 2003b; Keenan et a l. , 2006; Lohnes, 2003 ; Pilon et a l. , 2006; Pilon
et al. , 2007; Shimizu et al. , 2005). Canonicat Wnt signa lling in particular is important
for inducing Pax3 expression in the posterior NPB, and with Cdx proteins as
intermed iar ies (Sanchez-Ferras et a l. , 20 14; Sanchez-Ferras et a l. , 20 12) . For this
reason, as weil as evidence demonstrated by others (P il on et al. , 2006; Shi miz u et a l. ,
2005; Zhao et a l. , 20 14), Cdx genes can be said to be prime targets of Wnt s igna l ling.
ln this project, the nuc lear effectors of the canonicat Wnt pathway, Lefl-~catenin ,
were used in the transfections aiming to si mulate the pathway and activate the neural
specifie enhancer (NSE) of Cdx2. Since the goal is to identify and distinguish the
s ites respons ive to neura l specifie activat ion (and in its absence the potentially non
neural ones), it was necessary to add Sox2-a neural spec ifi e transcription factor and
earl y marker of the neural plate (Papanayotou et a l. , 2008 ; Pevny and Nicoli s, 20 l 0;
Rex et a l. , 1997).
44
5.2 Results
5.2.1 Potential binding act ivators
Matlnspector is an online software used to predict the potential transcription facto r
binding s ite (TFBS) locat ion on a query DNA sequence, that is, a promoter or
enh ancer. The potential TFBSs detected when submitting the 852-bp NSE seq uence
to the genomati x servers was impress ive, at 256, w ith matrix s imil arity set to the
optimized >0.75 ; g reater than 0.80 is c lass ifï ed as a "good" match. Mindful of the
impor1ance of Wnt s ignalling, potenti al TFBSs of the nuclear effectors Lefl/Tcf we re
searched. T here were 6 potential s ites (F igure 5. 1 a, b). One s ite overlapped with
another and was therefore co unted as one, and wi Il hereafter be referred to as one; it
was a lso the on ly potential TFBS for Lefl /Tcf effectors found in the fo rward NSE
strand . Matrix simil ar ity ranged from 0.84-0.96, the hi ghest similarity scores came
from the reverse NSE strand.
Sox2 TFBS were fo und only as part of a composed binding s ite, wh ich includes Oct4,
Nanog, TcD (Tcf7 ll ) and Sa ll4b (F igure 5. 1 a,c), and class ifïed as STEM by the
software. There were on ly two of these composed binding s ites, and matrix simil arity
was marg ina l! y below the eut-off range for "good" ma tri x simil arity of >0.80. T hese
s ites were located on the reverse strand exc lusive ly.
A Exon 1
200 b~
Matri x families : • V$LEFF \/$ STEM
B LEF1/TCF
••vt• O.UI'H IIM!Qy -'•""''Y ln~•Uon
mm LEFifTC, !L1I.W..QJ.
liJI..eU LU Ina' I!J!JiWl
mm LEfttfCf ~
liJI..eU l.UIITC' lt1IWl.W
mm LEft/TCF ~
liJI..eU LEI'11TC, ~
c STEM
DMINitdMMI1a hm
"'"""'"""" TCI'ILI!' · t . lnvofr~~un .. wrct~gnaJ 103 ~PIItnwiY
TCP'ILEF• t . ~N'I lf'IIWnlilllgllal 3113 nneduaion Plllhwly
TCfU.F· t an HMG DOle UW1101pfX)n
020 -Td711 (TC13) TCFI\.CF· I (........,. ,.. DNA--) TCFI\.EF·1(....._
113 ONA_fi" ... OI>Ct)
...... OrtQ-.d '""''Y ....... O.WIItd MMrl.A
.... "Y - Wrtonn.adon
MOtif eotnpOMd Of ~-~-"' !!Wl:IJ!
binclngaitM fof liJlllilllll. OciA &o><2. Nonog. Tell
~Otl:am (Td11t) and SaiMb ln ... -. ... .__ Motif ootnQOMd Of ~-~"'"'
liUWo!
__ .,. liJlllilllll.
OciA toa2 Nonog Ttr.l
~01*'"" (ll:f111)MdS.IMbln
'""-. ... --
Cdx2 Exon 2 lntron 2 Exon 3
PoeiUon Sequence {red: cl-value> 60 ltrand
ID tn<I!A! O CAPITALS: core sequence)
ug 111 (-) 01r12 cgaatagCAAAgagctt
40il ..,, (-) ... , atcttatCAAAgaaaag ... ... (•) OlliiO aggagaTCTAaggggtg .,. ... (·) .... ctcacatCAAAacagat , .. n1 ( • ) 0847 agtaaaACAAtcaatcc ••• ,., (•) .... aaacaaTCAAtccaaga
'""Jon :.1 Sequence <red: cl-value> 60 .,.,.. '""" 10 - llm. CAPITALS: core sequence)
0
2tO 278 :Mg (-) 0710 cgcaggcGCATttgaatgg
, .. ... 1111 (-) 0 ,..a tgagtcaACATattaaatc
Figure 5.1 Potential Lefll!Tcf and STEM family binding sites.
45
2
(A) Forward and reverse strand locations of potential TFBS for (B) the canonicat Wnt pathway effectors of the Lefl /Tcf family (five total Binding-Sites, BS; shown in dark green), and (C) the combined binding sites of Oct4, Sox2, Nanog, Tcf3 (Tcf7Il) and Sall4b in pluripotent cells (two total STEM sites; shown in light green). The third STEM site seen in schematic A is an Oct3 and not a Sox2 binding site, and is thus excluded. BS6 is also excluded because it overlaps with BSS. Instead, for practicality, BSS combines both. Core sequences are the four most conserved consecutive nucleotides. Higher consensus index values indicate greater conservation. Images retrieved from Matlnspector.
46
5.2.2 NSE and potential TFBS sequence homology
The next step in analyzing the potential TFBSs is observing how they match up with
vertebrate conserved regions. A conserved TFBS may indicate evolutionary selection
and therefore a good candidate for further testing. The online University of California
Santa Cruz (UCSC) Genome Browser allows us to visually discern how conserved
the domains are. In Figure 5.2a, we can see the entire enhancer, in the forward
direction, and how sorne potential binding sites are more conserved than others. The
phyloP scores of 60 vertebrates, which include 40 placenta! mammals, determine
basewise conservation. The higher the positive score is, the greater the conservation.
Negative scores indicate faster evolution (Rhead et al., 2010). The range for the entire
enhancer was from -5.114 to 4.8243.
Figure 5.2b, is an expanded view of the binding sites found in Figure 5.2a. Potential
Lefl /Tcf binding-site-1 (BS 1) has the !east sequence homology, with a mean phyloP
score of -0.381687. BS2 has the highest conserved sequence, with a mean phyloP
score of 3.65525. Both conserved "STEM" binding sites are also highly conserved.
See Appendix Table A.1 for phyloP scores of each potential TFBS sequence as weil
as of their highest conserved core sequence.
"' !!i iii
:Il ::= m
!!! !ii :ij !i! i!! iii
!ii
~n
iii !!!
l!!
iii
::: ln :n iii :n :::
iii iii H! Il! iii n: q:
!:!
i!'i :.:
ii!
::: Hi H!
i t
48
B mmtO 10bases mm10 Obases Scale
1147.304.670 h 47,304.S60 I147.304.SSO 1 1147.304,520 h47,304,510 1147.304,500 11 47.304.490 :chr5 TCT T TAGTACTACAAGCTCTTTGCTATTCGGTCTGCACAC GGGCCGCAGCGCCCATTCAAATGCGCC TGCGGGGCTGTGT ->
UCSC Genes (Ae!Seq, GenBank. tRNAs & Comparative Genomics) UCSC Genes (RefSeq, GenBank. lANAs 8 Comparative Genomics) D
60 vertebrales Basewlse Consel'\'atlon by PhyloP 60vertebrat~S8 Conserva tion by PhyloP 60 r- - - - - - ---- .- - - - 1------ - - - -B s 1 Multlz Alignments ol60 Vertebrales
4 STEM l \'lultlz Alignments of 60 Ve~tebrates
2 TCTTTAGTACTA qAAGCTCTTTGCTATTCG~ TCTGCACAC GGGCCGCAGCGC~ CATTCA A ATGCGCCTGCGP GGCTGTGTM~• CCTTTAG 3 A T TACAAGCTCTTTGCTATTCG T TCTGCACA T GGGCCGCAGCGCCCA TT CAAA TGCGCCTGCG GGGCTGTGTAat CTC TT TGGG CT CCA GC > CC AT GC TAGA TGC GTGG C GGCCGCG G A GC G CATTC A AATGCGCC C GC C GGGC CGTGTA~~
CCT C TGTGG CT L CA GC C CTCG TGC C AT AC TA TCTGC GTGA GGGCCGCAGC CCCA TT CAAA TGCGCC TGC A GGG TC G TG T Human CCTTT GCGG CT CCA GC C CTCGG GC CG T • C ~ TCTG GGTGG GG CCG C AGCG CCC AT TC A AA TG CGCC T G G C G T G C C G TG T Tree_shrew
CT TCAGG CT CCCAGC CTGG TG TCGCG C AC CTGC GTGA GGGCCGC T GCGCCCATTCAAATGCGCCTGC CGGGC CGTGT~
CT T TGGGG CT GCGCCGCTCGGA CTAT GC TA TC TGC GGT C GGG CC TC TG CG CCCA TTC A AA TGCG CC TG CCG GG CCG TG T Shrew CCTTT GTGG CT CCAA CCT TCG TGC C ACA C TC TCTGC GTGA CG A CCG l A GCG CCC AT T CA A A TGC GCC TG CCC CGC CG TG 1 Elephan1 .... .. .... - . GGG TG G G G O. G T C T A TT CA C A G G A G G C T T C C.O. .0. A C C G TG T Opossum ....... . Platypus .. ...... . ... .. .. ..... • GGGCC C . C C G CA TT CAAA T -G GCC T T T GG A GC ~ G ..... G C Chicken ....... . ....... . . ......... .. . Fugu .. .... ........... ..... ................ ......... .. - ... • Zebrallsn
mm10 / 110bases mm10 10 bases Scale 1147.304.380 147,304,370 147,304,360 1 1147,304,200 1147.304,190 1147.304,180 1 :chrS
TTGAAAGCGCTT T TCTTTGA TAAGATCGCCTTGTCTTGCA CTGGGCAACTCCCACCCC T TAGATCTCCTT G TTCAACTAG -> UCSC Genes (ReiSeq, GenBank. lANAs & Comparative Genomics) UCSC Genes (AefSeq. GenBank. lANAs & Comparative Genomics)
m - 60 vertebrales Basewise Conservation b P~P -..- .... --- 60 vertebrales Basewise Conservation by PhyloP - - -~------ ---60
B s 2 MuiUz Allgnments ol60 Vertebrales B S 3 Muf! lz Allgnments ol60 Vertebrales
TTGAAAGC ~CTTTTCTTTGATAAGAT~GCCTTGTCTTGCA C TGGGCAAC TCC q ACCCCT T AGA TC TCC T tr G TT CAAC T AG Mouse T TG A AAG G GCTTTTCTTTGATAAGATC T CC TGTCTTGCA CTGGG T A ACTCCCACCCCTTAGATCTCCTTGTTCA A CTAG~I
TTGAAAG G GCTTTTCTTTGATAAGATCGCCTTGTC C TGCAC A GGC GG CTCCCACCC T T AGA TC T T C C TG T GGG AC T G G Rabbi! TTGA A AG G GCTTTTCTTTGATAAGATCGCCTTGTC C TGCAC A GGCAAC TCCCACCC T T C AGA TC T "TTC TG T GGG A T T G G Human TTGAAAG G GCTTTTCT T TGATAAGATCGCCTTGTC C TGCA C CAA GCAA G TCCCACCC T TTA A A CT T C C TG C GGG ACT G G Tree_shrew TTGAAAG G GCTTTTCTTTGATAAGATCGCCTTGTC C TGCA C CAA G AG ACTCCCACCC ~ TTAGATCT T C C TGT AGG ACT GG~
TTGAAAG G GCTTTTCTTTGATAAGATCGCCTTGT TC TGCA C C A D, GC G AC TCCCACCC T T C AGA TC T T C C TG T A AG AC CG GShrew TTGAAAG G GCTTTTCTTTGATAAGATCGCCTTGTCTTGCA CCA GGCAACTCCCACCC T C AGA T T T T C C TG ~, GG AC T G G Elephant ~ TGAAAG G GCTTTTCTTTGATAAGATCGCC C TGT TC TGCA ~C GGGCA T A TCCC G CC T CTTAGATCT ~ C C T CG C - AC T G G Opossum . ... . . ... . . . . . . . .. Platypus TTG A G A A.ü G TTTTCTTTGATA A GATC T CCTTGTC CC GCA .. Chic ken
··---- .. Fugu . . . . . . - ... - ... ~ . .... . ....... ... aua•••••• . . .. - ........ • • • • Zebrallsh mm10 10 bases mm10 10 bases Scale
l147,304, t60 1147,304,150 lt47,304,t40 1147.304.130 h 47.304,000 1147.303,990 h •7.303,980 1147,300,970 :chrS AGCACTAAAATCTATC T GTT T TGATG T GAGACC T CTCCAA CAGTAAAACAA T C A ATCCAAGATTTAA T ATGTTGACTCAA - >
UCSC Genes (Re!Seq, GenBank. lANAs & ComparaUve Genomics) UCSC Genes (ReiSeq. GenBank, lANAs & Comparative Genomics) D
60 vertebra les Basewise Conservation by PhyloP 60 vertebrales Basewlse Conservation by PhyloP 60 - ---- - - - ~ - - - -B 54 Mulllz Allgnments o! SO Vertebrales B S 5 Multiz Allgnments of 60 Vertebrales STEM2
AGC AC TAAAATCT~ TCTGTTTTGATGTGAG~CCTCTCCAA c;t AG T AAAAC A A TCAA TCCAA~ A IT TT AA TA TG T TGAC TCAjA~ouse A GC G C C AA G A C CTGTT C TG C TGTGAGA T CTCT CAA CAGTAAAACAATCAATCCAAGATTTAATATGTTGACTCAAA~ GGC GA T TG A A C CTGTT CA G C T TCA AGA 6GC C C CC G A CA G TA A A AC A AT CA AT CC A AG C T TT A AT AT G TT GA C 1 CA A AabbU G GC GA T TG A A C CTGTT CA G C TG CA AGA •GA C ~ C A AA CAGTAAAACAATCAATC TGG GATTTAATATGTTGACTCAA Hu~n
GGC GACTG A A C CTGTT CA G C TG r- AGA ~GA c r CC G A CA G TA A A AC A AT CA AT CC A AG AT T T A AT AT G TT GA CT CA A Tree_shrew r. oc r. r 'l A r C TGTT l1 G TG C1 GA GA C C'AGG A CAGTAAAACAATCAA C C T G AG G TTTAATATGTTGACTCAA~
GGC G T TG A Shrew GGC G C AGAGC C T A TT ~ TG GC G ~~ AGA A~A C C CC G A CAG T AAAACAA TCAA TCCAAGA TT T AA T ATG T TGAC TCAA Elephant A C~G CTG AA c TGTT CCTCAAG GA ~CGGATCTA AA A AG ù AAAACAATCA T TC TG AGATT A AT AT G TT GA CT CA G Opossum ...... Platypus . . . . . Chic ken . .. . ... Fugu . . " ............................... ................................. • Zebralfsh
Figure 5.2 Vertebrate homology ofthe neural specifie enhancer. Outlined in black are the potential TFBS for the Lefl /Tcf family of effectors (Binding-Sites 1-5), and in red, the composed binding sites for Sox2 and other pluripotent/stem cel! markers (STEM l/2). Panel B is an expanded view of Panel A. The basewise conservation of 60 vertebrates is plotted as positive ( conserved) or negative (acceleration) phyloP scores. Out of the 60 vertebrates, 40 are placenta! mammals and, of these, 8 are rodents, pika and rabbit. For clarity, 13 of the 60 vertebrates are represented in the sequence alignment. Taken from the UCSC genome browser.
49
5.2.3 Directional preference and synergy
Luciferase assays were carried out to test the potential activators Sox2 and Lefl
~catenin on the full Cdx2NSE forward and reverse reporter sequences. Transfections
were carried out on P 19 ce lis, which are pluripotent embryonic carcinoma that have
the ability to differentiate into muscle and neuron cell-types (McBumey, 1993) Lefl
~catenin are the nuclear effectors of posteriorizing Wnt signalling, and Sox2 is a
neural specifie transcription factor. The potential activators were either transfected
alone, or co-transfected together in doses that ranged from 0-100ng DNA each with
1 OOng of the Cdx2NSE luciferase reporter in either reverse or forward NSE strand.
The reverse strand displayed significant (2.2 times the sum of each) synergistic
activation with both Lefl-~catenin and Sox2. Also, the reverse strand experienced
very significant activation (p = 0.006) compared to the forward strand in co
transfection (Figure 5.3). This finding was surprising considering how enhancers do
not genera Il y display directional preference (Pennacchio et al., 20 13). lnterestingly,
ail potential TFBS except BS5 are 1ocated on the reverse strand. This test a1so
determined the dose (lOOng) needed for the disp1ayed synergy.
50
c: .Q -~ :0 <t "C 0
LL.
Cdx2 NSE forward- Luc
1 2 3 4 5 Il
Cdx2 NSE reverse- Luc
1 5 4 3 2 1 IÎucl Il
Lef1-j3CAT Sox2
Figure 5.3 Cdx2NSE synergy and directional preference.
P19 (n=3)
**
Lef1-j3CAT +Sox2
Luciferase assay tested the enhancer in the forward and reversed direction. Cotransfections utilized Lefl -~catenin to simulate the effectors of posteriorizing Wnt, and/or neural specifie Sox2. Doses ranged from 0-1 Oüng of activa tor DNA with 1 Oüng of Cdx2NSE. The reverse strand displayed significant synergy with both Leflpcatenin and Sox2 (**) indicates p = 0.006 (t-te t). Lucifeïase activity füï each condition was calculated as fo ld activation compared to the reporter vector alone. The results are represented as the average + standard deviation of triplicate samples from three experiments.
51
5.2.4 Testing the reverse NSE
Since the reverse NSE was identified as the preferred direction for Lefl-~catenin and
Sox2 activation, it merited additional transcriptional activation analysis . Fragmenting
the reverse NSE in smaller units permits us to test a narrower group of binding
demains. The reverse NSE was fragmented five times containing potential Lefl /Tcf
binding sites: (BS4,3,2,1), (BS4,3,2,1), (BS3,2,1), (BS2, 1), (BSl), and (BS5). The
location of the STEM (Sox2) composed binding sites was incidental, that is
fragmenting was focused on the potential binding sites for Lefl /Tcf. Removing BS5
in particular was important considering the binding motif is located on the forward
NSE strand. Surprisingly, compared to the full reverse NSE strand, the strand missing
BS5 had on average 0.7x the activation of the full sequence (Figure 5.4). Y et BS5 by
itself did not display any activation. There was a Iso a significant discrepancy between
the activation seen here for the full reverse NSE strand and what was seen previously
(Figure 5.3); possible reasons for this will be discussed in the following chapter.
Lefl-~catenin and Sox2 synergy was only apparent in the full reverse strand.
In summary, none of the fragments recapitulated the neural specifie activation or
synergy found in the full reverse strand as hoped. Thus it remains unknown which
areas of the NSE architecture are clearly sensitive to neural specifie activation and
which areas are not. In addition, it is unclear if the STEM (Sox2, Oct4, Nanog, Tcf3/
Tcf7Il and Sall4b) binding sites con tri bute to the evidenced activation. What is
known is that compared to the full forward NSE, the full reverse NSE is significantly
activated by the combined action of Sox2 and Wnt pathway effectors Lefl-~catenin.
52
13 2• 1 [wei !···~ ... ~ ...... 5;;..._t----
l 1= 1=
0 CTL 0 Lef1-{3CAT ~ Sox2 • Lef1-{3CA T + Sox2
1:i:1 b P19
~~ .. ! .... !.! .. =~----~----------~--------~-------(n-=6_) -~ 0 2 4 6 8
Fold Activation
Figure 5.4 Fragmentation results of Cdx-2NSEreverse. Dose was lOOng of activator DNA and lOOng of Cdx2NSEreverse fragments. Lefl~catenin and Sox2 together robustly activated the promoter, of the complete enhancer NSE reporter followed by the second most complete enhancer (BS4,3,2, 1). In ail there was a trend present, where the bigger the enhancer the bigger the activation. An asterisk (**)denotes a very significant p :S 0.005; and (*) a significant p :S 0.05. The Blue asterisk (*) marks the presence of the composed STEM binding site that includes Sox2.
CHAPTER VI
Dl SCUSSlON
6. 1 ln pursui t of a condi t iona l Cdxl overexpress ion mouse mode l
Homo logous recombinati on has long been used to in sert exogenous DNA co nstructs
into a specifi e genomi c locus in mi ce (Capecchi , 2005; Misra and Duncan, 2002).
T he ROSA26 locus in parti cul ar has been establi shed as the preferred genomi c locus
for gene targeting (Caso la, 201 0; Nyabi et a l. , 2009), because transgene integrati on is
stable, and express ion leve ls are constant and ubiquitous, not to mention that
di sturbing the locus produces no phenotype (Soriano, 1999; Zambrowicz et a l. ,
1997). New targeting strateg ies have emerged that successfully target the ROSA26
locus in mice. T hese inc lude z inc finger nuc leases (ZFNs), transcription activator-like
effector nucleases (TALEN s), and clustered regularly interspaced shot1 pa lindromic
repeats CRJSP R/Cas RNA-guided nuc leases (Cas9/gRNA) (Hermann et a l. , 20 12;
Hsu et a l. , 2014; Kasparek et a l. , 20 14 ; Meyer et a l. , 2010; Sato et a l. , 2015). These
nuc lease based technologies are capable of c leav ing genomi c DNA at spec ifi e sites,
causing doubl e-strand ed breaks to improve homologo us recombinati on as we il as
insertion of exogenous constructs medi ated by DNA repa ir. They are timesav ing and
cast-effi c ient methods that have as high as 4 .5% (ZFN s), 8.8% (TALENs), and 35%
(Cas9/gRNA) recombination efti c iency in gene targeting of the mouse ROSA26 locus
(Fujii et a l. , 201 3; Kaspa rek et a l. , 20 14; Meyer et a l. , 2010). However, there is still
some uncet1ainty regarding poss ible off-target DNA cleavages caused by these new
nuc lease-based strateg ies, as we il as th e cytotox ic ity that could be brought about by
the doubl e-stranded breaks . Targeting based on TAL EN s in parti cular seem most
54
promis ing, w ith K asparek et a l. (20 14) reporting no off-targe t acti vity found in the ir
approach. Neverthe less, concerns rema in s ince off-target DNA c leavages can still be
induced by these methods, resulting in unknown mutations that a re di ffi cult to detect
(Kong et a l. , 20 14; Koo et a l. , 20 15; Owens et a l. , 20 13). ln the end, the traditi onal
homologo us recombinati on strategy a imed at ta rgeting the ROSA26 locus is st ill the
esta bi ished so luti on, and our preferred cho ice, s in ce it does not induce off-target
mutati ons.
We ta rgeted the mouse ROSA26 loc us w ith a vecto r co nta ining a fl oxed-PGK-neo
tpA stop cassette upstream of FLAG-Cdxl and IRES-eGFP. After transfecti on and
G418 se lecti on, se lected co lo ni es were screened fo r the knock-in a ll e le. Out of 96
c lones screened, 3 (3 .1 %) were PC R-pos itive fo r integration of the knock-in constru ct
corresponding to the F 1-RI primer screen parameters. Southern blot was
inconclusive, w ith none of the contro ls working, indicating that technica l iss ues most
li ke ly pl ayed a ro le. lt wo uld be very interesting to find out how often PCR-pos iti ve
ROSA26 loc us targeted c lones are a lso So uthern blot-pos iti ve. Soriano (1999)
reported that 8 o ut of 8 (1 00%) were. Of note, o ur 1.1-kb PCR screen method was
nearly identi cal in locati on as described by ( 1.2-kb) So ri ano ( 1999) . Despite the
more-than 103 ROSA26 locus knoc k-ins described to date (Caso la, 201 0), thi s
in furmali un i now li kely seen as tri ial and no longer divul ged. Moreover, if any
verifi ca tion deta il s are g iven they usua ll y extend to Southern blot confirmat io n.
So uthern blot not only va lidates 5 ' and 3' transgene in sertion but a lso di stingui shes
the w ild-type from the ta rgeted a lle le. T he advantage of PCR is that it a llows us to
examine a large number of c lones in a high-thro ughput manner, sifting and
identify ing a sma ll e r number of poss ibly pos itive c lones. In the end, even if our three
c lones were validated by southern blot, that is just 3 .1 % of a li c lones, whi ch is be low
the reported average ta rgeting effic iency for the ROSA26 1ocus, a round 25% for G4 1 8
se lected c lones (Hohenste in et a l. , 2008) . T he lower number of pos iti ve c lones co uld
be to due to transfecti on optimization issues.
55
In thi s proj ect, 96 d rug-se lected clones from a 96-we ll feeder plate were part
transferred onto four 24-we ll ti ssue culture plates, and the c lones remaining in the 96-
well plate were then stored at -80°C. Cell s in the 24-we ll pl ates were a ll owed to
expand until confl uent. T he advantage of transferring 96 drug-se lected c lones onto
four 24-well pl ates versus another 96-we ll plate is that the ce l! lysate from a 24-we ll
plate permits fo r severa! Southern blots whereas a 96-we ll plate would onl y produce
enough DNA for one Southern blot (Limaye et a l. , 2009). In our case, there was
enough DNA fo r two southern blots targeting the 5 ' and 3 ' ROSA26 knock-in locus.
However, since there was a need fo r more DNA, the entire 96-we ll plate used fo r
storing the clones at -80°C was defro sted and the three PCR-pos iti ve clones and one
WT clone was passaged fo r subsequent storing and DNA harvesting. Once aga in, the
harvested DNA was exha usted in two more Southern blots that fa iled to target the 5'
and 3 ' knock-in locus, and WT ROSA26 locus. For futu re work thi s suggests that it
may be best to start back from the targeting vector, instead of defrosting and
passaging more c lones. For one, passaging ES ce ll s more and more increases the
potential for differentiation. Two, storing the ES c lones- now c ryogeni cally frozen in
liquid nitrogen- for long peri ods of time dimini shes the ir capacity to be successfull y
micro injected. And three, literature indicates that targeting the ROSA26 locus is
highly effi cient. Thi s means that improvements in transfecti on conditions could like ly
y ield more pos iti ve clones, closer to the reported 25% average (Hohenste in et a l. ,
2008) . In additi on, no contro l probes were tested. lt may be worth whil e to target the
GFP sequence, for example .
If homologo us recombinati on and integration of our Cdxl transgene were confirmed,
we wo uld be we il on our way to generate a knock-in mouse capable of conditiona l
Cdxl overexpress ion. Knowing that Cdx prote ins pl ay a s ignifi cant ro le in the initia l
steps of NC deve lopment and that overexpress ion of genes needed fo r NC
development is manifested in NC deri ved tumours (Trainor, 20 13), we had
hypothes ized that a mouse line overexpress ing Cdxl would gene rate NCC deri ved
56
tu mours su ch as neuroblastoma. Without the mou se 1 ine, we a re sti Il le ft w ith thi s
question. M oreover, still to be discovered are the regul atory mechani sms implicating
the neura l fun cti ons of Cd x 1 prote ins to the supposed N C-deri ved mal ignancy. Of
note, this mo use li ne capable of conditi ona l Cdx.l overexpress ion wo uld be the first to
demonstrate such a link, w ith overexpress ion particul arl y a imed at neura l crest
deve lopment. There ex ists one reported transgenic Cdx l overexpress ion mouse mode!
w hose express ion is not induc ible in spec ifi e ti ssues/context and is under the co ntro l
of a Cclxi promoter instead of the stable and ubiquito us express ion confe rred by the
ROSA26 promoter; overexpress ion in thi s case had homeotic and axia l patterning
defects (Gaun t et a l. , 2008) .
6.2 lnsig hts into Cdx2NSE neura l regul ati on and potenti a l applica ti on
Cdxl genes a re just one of three Cdx genes that have been mainta ined by vertebrates
for over hundreds of millions of years, name ly Cdxl /2/l in mamma ls, birds,
amphibians and certain fi sh (Faas and lsaacs, 2009; Marletaz et a l. , 201 5; Mulley and
Holland, 201 0). Cdx prote in s share some spatial overl ap as we il as core pathways,
particula rly those necessa ry for AP vertebra l patterning (Beek and Stringer, 201 0).
T here is a lso ev idence ind icating that Cdx prote ins medi ate Pax3 express ion- Pax]
be ing a key transcripti on fac tor in the NC-GRN (Sanchez-Ferras et al. , 20 14;
Sanchez-Ferras et a l. , 20 12) . See Appendi x B fo r our study on Pax3 regulatio n by
Cdx2 prote in s. Prior to the emergence of the NC-G RN , pre-vertebrate chordates
like ly o nly had one Cdx gene, as is the case for the cephalochordate amphioxus
(A mphiCdx gene) (Broo ke et a l. , 1 998). Cd x l/2/4 prote in s may share si mi lar targets,
and some redundancy may be an evo luti onary safeguard advantage, but Marletaz et
a l. (20 15) argues th at wh ile corroboration and overl ap ex ist between Cd x pro te ins,
' O rtho logous gene nomenc lature may d iffe r between spec ies
57
the ir indi v idual functi ona l roles are li ke ly di stinct. T his may indi cate that prio r to the
emergence of Cdx 1/2/4, the so le Cdx gene precursor underwent ge ne dupli cati on and
sequentia l modifi cations that lead to th ree neo-funct ionalized Cdx prote in-coding
genes. lt is no secret that sequentia l geneti c modifi cations guide the evo lut ion of
novel features (H olland et a l. , 1994; Penni si, 2008). However, w hether the c riti ca l
dri v ing fo rce underly ing ve rtebrate evo lution was due to prote in neo-functi ona li zati on
or changes in c is-regulatory sequences is sti Il contested (Carro ll , 2008 ; Lev i ne, 20 l 0 ;
Lynch and Wagner, 2008) . Concerning the evo lution of vetiebrates and the NC-G RN ,
both Van Otterl oo et a l. (20 13) and Jandz ik et a l. (20 15) make the case that mutations
in pre-vertebrate c is-regulatory reg io ns were c ri tica l fo r NC evo lu tion. Van Otter loo
et a l. (2 01 3) go so fa r as to suggest that there is litt le ev idence supporting the c la im
that prote in neo-fun ctiona lizati on gui ded the evo lu tion of the NC-G RN. T hi s line of
thin king is the most accepted by evo lutionary deve lopmenta l bi o log ists (I-loekstra and
Coyne, 2007; Wray , 2007) . Pati of the reasoning is that many prote in-coding genes
have highly conserved functi ons among diffe rent species (such as non-vertebrate Cdx
gene homologs AmphiCdx and cad) but vari ati ons in express ion leve ls abound
(Wittkopp and Ka lay, 20 12). W hether the dri v ing fo rce fo r NC-GRN evo lu t ion was
due to prote in-codi ng or c is-regul atory changes, both evo luti onary novelt ies have
ul t imate ly contributed to its deve lopment.
ln the case of Cdx2NSE, the enha ncer is undoubtedl y a vertebrate-exc lus ive feature,
invo lv ing a c is-regulato ry sequence that enhances the express ion of a vertebrate " neo
functi onalized" prote in. C learly it is di ffic ult to di scern , w ithout go ing back in time,
which nove l feature was more cruc ia l fo r vertebra te deve lo pment, the (NSE)
enhancer for Cdx2 o r Cdx2 itse lf. Of course th e main contenti on is not w hether o ne
feature came first or not, but whi ch nove lty was c ri tica l fo r vertebrate evo luti on (and
poss ibl y fo r evo lution of Cdx neural fun cti ons) . If we fo llow conventio n, most
evo lutionary bio log ists would probabl y agree that divergences in the NSE sequence
are more like ly to guide important phenotypic changes. However, the ro le and impact
58
of the NSE is not yet known. A vertebrate homo logy comparison of Cdx2NSE carried
out by Wang and Shashikant (2007) indicated a 65-98% NSE sequence simil ar ity
between mouse and Xenopus , chi ck, oposs um , dog, rat and human. There was no
homology w ith zebrafish and fugu, which is not surpri s ing g iven the fact that teleost
fish, like zebrafish and fugu, do not have Cdx2 genes, but instead have two cop ies of
Cdxl and a s ing le Cdx4 gene (M ull ey et al. , 2006). 8oth Cdx2 and its own NSE seem
to be highl y conserved together, w hi ch indi cates a se lect ive pressure to ma intain both
features.
Whil e our take 1n the study ing the enhance r does not come from an evo luti onary
sta ndpo int but a pract ica l one, it is worth noting that enhancers as a who le are often
hi ghly conserved, yet subtl e var iat io ns a re li kely between spec ies and thus we ca nnat
rely on homo logy a lone when ana lyzi ng putative TFBSs. Our own homology
com parison of the 852-bp NSE, described by Wang and Shashikant (2007), with 100-
vertebrate spec ies from the UCSC Genome browser a lso va lidated the hi g h
conservation of the NSE. However, there was di vergence in seq uence conservat io n
for putative Lefl /Tcf bindings s ites found by computationa l ana lys is, with onl y
putative s ites BS2, BS3 , and BS5 appear ing to have co nservation w ithin the 100-
vertebrates. A more se lective search, inc luding just mouse, human, rhesus and
e lephant does ïeveal conseïvation of putative BS4 but not BS l , which is the nwst
divergent of a li five potential Lefl /Tcf binding s ites. Luc ife rase assay tests ofBS l by
itse lf, w ith cano ni ca l Wnt pathway effectors Lefl -~catenin revealed no act ivation ,
even wh en co-transfected with Sox2. Granted, the synergy of both Lefl - ~caten in and
Sox2 was only evident in the full Cdx2NSEreverse- luc reporter, thus no rea l
infe rences can be made on indi vidua l or groups of potential binding sites. What can
be said is that a trend is present between the enhancer fragments (F igure 5.4), where
the larger the fragment the greater the activat ion. A lso, it is unc lear if the two hi ghly
conserved potentia l STEM (Sox2) binding sites contributed to any act ivation.
59
T he activ ity of Cdx2NSE is not entire ly located in the neu ra l spec ifie regions of the
developing posterior neuroectoderm and NT, but also slig htly in posterior
mesodermal tissue (Coutaud and Pilon, 2013). The goa l here was to test the NSE for
neural specifie activity, and identify the regions most receptive to neural spec ifi e
activation so that a truly neural spec ifie enhancer co uld be constructed and used for
the generati o n of a Cre mouse line driven by the new enhancer. T hi s new modified
enhancer would hopefully be both neura l spec ifie and innocuous in vivo. Even if
lucife rase assays had identified a promtsmg NSE, its true utili ty would not be
apparent until tested in the transgenic mouse line. T hi s risk may be worth taking
considering how such a tool , a imed at posterior neural development, does not yet
exist.
Si nce no one ptece of the NSE architecture stood out more in neural spec ifie
activation, it remains to be identified. If, for example, one small segment had been
ab le to nearly recapitulate the neural specifi e activation of the full seq uence, then by
exclusion the mi ss ing segments could be categorized as non-neura l spec ifie. While
stilljust a c lue this could indi cate that the smaller segment stro ngly sens iti ve to neura l
spec ifie activat ion could be used in regenerat ing a truly CcL"~:2NSE Cre mouse line.
Unable to distinguish these s ites through act ivatio n by Sox2 and Lefl -Pcatenin, there
are two suggestions for future research. One is to repeat, and two is to test NSE
activation by other activator proteins needed for posterior neurogenesis.
Admitted ly, concerning the discrepancy between the act ivation displayed by the full
Cd.x2NSEreverse in tests from Figure 5.4 and 5.3 , it wou ld have been better to
maintain the same cel! density as before, for co ntinuity , and reproducibility. As
mentioned in materia ls and methods, the first tests demonstrated in F igure 5.3 used
3x 104 ce ll s/we ll , and I.Sx 104 ce ll s/we ll in the second tests shown in Figure 5.4.
Somewhere a long the way 3x 104 ce ll s/well mistakenly became 3x 104 cell s/mL and
thus 1.5x l04 ce ll s/well (i.e. SOO~d/we ll). Still , 3x l04 cells/m L falls somewhat c lose to
60
the recommended (Novagen) dens ity range of 4-8x l 04 ce li s/mL for ad herent cells
rang ing from three-quarters to about one-third the amount. Even if not the
recommended amount, one (n= J) 6xl04 ce lls/mL was compared to one (n= l ) 3x l04
cells/mL, us ing the same transfection materi a ls and conditions (done on the sam e
day). The results did not suggest more cells were respons ibl e for the discrepancy, in
fact sli ghtly more activation was observed in the test w ith less cell density. T hi s could
suggest the inconsistency is due in large part to Jower qua lity of transfectio n
material s, as the substantia l time lapse from the first test and the second test meant
the orig in al mate rials had been exhausted and new ones were made. The comparison
be ing an n=l is of course not s ignifi cant and a rework is still warranted, yet it is c lea r
that materials and conditions need to be as similar as possible, and the best strategy is
to test consecutively in as short a peri od of ti me w ith enough of the same reso urces
for the whole duration. Desp ite the overs ight, conditions were kept the same for the
second part of the tests, and so a li things being equa l, inferences can st ill be made,
though aga in st i Il warrants retesting.
Testing the NSE with other poss ible key activators, and their combinations, could
a lso help ident ify regions more sens itive to ne ura l specifie activat ion. Matlnspector
ana lysis identifi ed four Ets motifs invo lved in FGF signalling (Appendi x C) . T he
nïüst conseïved of the fou ï motifs weïe the two most-inneï ones. Like wi th the
putative Lefl /Tcf motifs invo lved in Wnt signa lling, the Ets motifs were mostly
detected on the reverse NSE strand (3 out of 4). It could be the case that act ivat ion by
both Wnt and FGF pathways is synergistic.
Cdx2 has also bee n shown to regulate its own expression, through response elements,
sensitive to Cdx2 activation, located proximal to the promoter (X u et al. , 1999).
However, Matinspector did not identify any motifs for Cdx2 (o r Cdxl/4).
Wh ile the Sox-B l transcription family of transcription factors Soxl, Sox2 and Sox3
share 80% sequence s imil arity and are functionally redundant (Zhang, 2014), Sox2
61
rema ins the most defini te pan-neural spec ifi e marker of the earl y neural plate (Okuda
et al. , 201 0). ln addition, Sox2 pla ys a role in maintaining neural progenitor
populations throughout early development and beyond (Elli s et a l. , 2004). Sox2 along
w ith Oct4 and Nanog play a role in maintaining cell pluripotency as weil as
specify ing lineages. Sox2 maintains neuroectodermal identity while Na nog and Oct4
direct the differentiation of the mesoendoderm (Lo h and Lim, 20 11 ; Thomson et al. ,
20 Il ; Zhang, 20 14). Sox2, Nanog and Oct4 are a li part of the two putative composed
STEM binding sites fo und on the reve rse Cdx2NSE strand. lt co uld be possible that
these sites are equa ll y responsible for the neural specifie and mesoderm activity of the
enhancer observed by Coutaud and Pilon (20 13). If true, separat ing two fate -guiding
traits from one shared binding s ite is likely to be unmanageab le.
Enhancers are robust and complex regulatory features. Unlike mutations in protein
coding sequences, changes to enhancers are not as deleterious, nor like ly to be more
pleiotropic (Wittkopp and Kalay, 2012). Yet, their contr ibution to evo lution is
believed to be major. Attemptin g to uncover and refashion the complexity and
robustness acquired through hundreds of milli ons of years of evo lution is a
formidab le task. The identification of neural specifie regions rests to be discovered ,
and the hope is that, despite the c hallenges, retesting and adding the action of FGF
signa lling will help uncover these regions.
APPENDJ X A
E VOLUTlONARY CONSERVATION OF CDX2NSE POTENTI AL TFBS
Table A.l Potential TFBS phyloP plot scores. Core sequences are the most conserved four consecuti ve nucleotides. Pos it ive scores ind icate conservation, whereas negati ve scores predict acce lerati on. Data retri eved from the UCSC aenome browse r o ·
Query Sequence Bases Mi nim um Maximum Mean Standard Deviat io n
Cdx2NSE 842 -5 .114 4.8243 1.02275 1.71358
BS l 17 -1.70244 0 .439701 -0.381687 0 .608727 BSl CORE 4 -0.433024 0 .122346 -0.195008 0.282368
STE Ml 19 0.122346 4.16861 1.90538 1.31179
STEM l CORE 4 0.836394 2.42317 1.82813 0 .690139
BS2 17 1.94713 4.80332 3.65525 0.9985
BS2 CORE 4 3.29589 4.80332 3.83143 0.71368
BS3 16 -0.414598 3.91757 1.20367 1.17115
BS3 CORE 4 -0.112354 1.90261 0.970687 0 .852398
BS4 17 -2 .12732 2.3056 0.261006 1.03406
BS4 CO RE 4 -0.515346 0.0891417 -0.112354 0.284958 ·-
BS5 21 -0.716843 3.01083 1.49482 1.09648
BS5 COR El 4 1.39887 3.01083 2.40635 0.698003
BS5 CORE2 4 1.29812 3.01083 1.85223 0.791155
STEM2 19 -0.213102 3.11158 1.66399 0.976333
STE M2 CO RE 4 0.995874 3.11158 1.70111 0.95578
APPENDIX B
REGULATION OF PAX3 NCE3 BY CDX2 PROTEINS
The paired box transcription facto r Pax3 is an important regul ator of the NC-GRN
circuit, involved in the spec ifi cat ion of the neura l crest. T here are three neura l crest
e lement (NCE) enhancers for Pax3 : NCEI , NCE2 and NCE3. Recent enh ancer
studi es have demonstrated that Wnt-mediated signalling activates NCE2 through
intermed iary Cdx proteins, and that Z ic2 transcription factors a lso regulate NCE2
(Sanchez-Ferras et a l. , 20 14; Sanchez-Ferras et a l. , 20 12). NCEl/2 were the first
neural crest enhancers to be described , and as most enhancers go, they are highly
conserved; they are also similar in function (L i et a l. , 1999; Mi lewski et a l. , 2004;
Pruitt et al. , 2004). The third enhancer known in the lab, as NCE3 , was discovered by
Degenhardt et a l. (20 1 0) and is an introni c enhancer w ith seem ingly redundant
functions to NCE 1/2. T herefore, the hypothesis is that Cdx proteins regul ate the
NCE3 of Pax3 just as other en hancers of Pax3. Presented here are the luciferase tests
ana lyz ing the effect of Cdx2 as an act ivator on NCE3 forward and reverse sequences
that we inserted into a lucife rase expression vector with an 800-bp Pax3 promoter.
There seems to be dose-dependent act ivation of the Pax3 promoter by Cdx2 proteins
with maxim um activat ion achieved at Cdx2 dose 25 ng (F igure A. l).
66
- NCE3forward-Luc
NCE3reverse~Luc *
* ns
ns
c 0
+=1 C'Q > ·-ü < "C 10 ~
0
Cdx2 (ng) 0 2.5 5 10 25
F igure B.l Effects on NCE3 reporter activation by Cdx2 proteins. (*) represents p :S 0.05
N2A (n=3)
50
APPENDIXC
CDX2NSE AND FGF SIGNALLING
ç:llts:ll; bMÇdi!:~D flflt21: c.> l!!ttdl!li flmlb'...lll!Dnmtm PotJtlon Addldor\111~ :hquenca ...... - - & ( rwd : ~>&O & ...... O.tai'-<1 F'•mltv ...... o.talled MatrtX tnfomlaUon - .. lll<IJo< &
m & .......... CAPITAL&~) F~lly ........... ..,,..,.,. SPI-t~. ~ murineETS1 ~ ~~lllc::tof 217 237 227 1·) ... , ---'"""" P\JI ..,,.., ... ~ muriM ETS1 ~ GA8P GA blndlng pn:u.n lOI 321 '" (•) .. .,
~-!Odon
"""""'""" ....,...__,,... ~ murWMIETS1 ~ 430 <50 ... (-) 092.&
-~ '"""" tiPU 1 ,....,,
"""""""" Eta • fAmly nwnber ELF-2 ~ muriMETS1 ~ ... 075 ... l·l 0900 - ClGMeiQ<Uxoe
""""' {NERF1a)
, .. ~-...... ..,_ ... _ .... _ .. ""'- ..... -
.... " . .... ..,_..- ... --..
68
Figure C.2 Placement of putative Ets motifs and homology. Ets motifs involve FGF signalling. GGAA characterize the core sequences of most Ets motifs, which makes their importance more problematic to address. At the top are the results of Matlnspector; there are four putative Ets motifs, with high matrix similarity particularly for the three reverse NSE stands. At the bottom are the results of 100-vertebrate homology comparison (UCSC genome browser) with respect to putative binding sites . A-C are the putative Ets motifs, in red are the putative motifs for the STEM binding sites, and light black the putative Lefl/Tcf binding sites. Ets sites BIC are the most conserved.
REFERENCES
Akiyama, H. , Chaboiss ier, M.C. , Behringer, R.R. , Rowitch, D.H., Schedl , A. , Epste in, J. A., de Crombrugghe, B., 2004. Essential ro le of Sox9 in the pathway that contra is formati on of cardiac va lves and septa. Proc Natl Acad Se i U S A 101 , 6502-6507.
A lthoff, K. , Beckers, A. , Be ll , E. , No rtmeyer, M., Thor, T. , Sprussel, A. , L indner, S. , De Preter, K., Florin , A. , Heukamp, L.C., Kl e in-Hi tpass, L. , Astrahantseff, K ., Kumps, C. , Spe leman, F., Eggert, A. , Westermann , F ., Schramm, A ., Schulte, J.H ., 201 5. A Cre-conditi ona l MYCN-driven neuroblastoma mouse made l as an improved too l fo r preclinica l studi es. Oncogene 34, 3357-3368 .
Amore, G., Yavroui an, R.G ., Pete rson, K.J. , Ransick, A., McClay, D .R., Dav idson, E. H., 2003. Spdeadringer, a sea urchin embryo gene required separate ly in skeletogenic and oral ectoderm gene regul atory networks. Dev Bio l 26 1, 55-81.
Angers, S. , M oo n, R.T. , 2009. Proximal eve nts in Wnt signa l transduction. Nat Rev Mo l Ce ll Bio l 10, 468-477.
Antony, A.C. , Hansen, D .K ., 2000 . Hypothes is: fo late-responsive neura l tube defects and neurocri stopathi es. Terato logy 62, 42-50.
Auerbach, R., 1954. Analys is of the deve lopmenta l effects of a letha l mutati on in the house mouse . Journa l ofExperim ental Zoo logy 127, 305-329 .
Baggiolini , A. , Varum, S. , M ateos, J.M., Bettosini , D., John, N ., Bona lli , M. , Z iegler, U. , Dimou, L. , C levers, H ., P utTe r, R ., Sommer, L. , 2015 . Premigratory and migratory neura l crest ce ll s are multipotent in vivo. Ce ll Stem Ce ll 16, 3 14-322.
Baird, P. A., Sadovnick, A. D., 1984 . Surv ival 111 in fants with anencepha ly . C lin Pediatr (Phil a) 23 , 268-27 1.
Barad, M. , Jack, T. , Chadw ick, R. , McG inni s, W., 1988. A nove l, ti ssue-specifi e, Droso phila homeobox gene. EMBO J 7, 2 15 1-2 16 1.
Beek, F. , Erler, T. , Ru sse ll , A., James, R. , 1995 . Express ion of Cdx-2 in the mouse embryo and placenta : poss ibl e ro le in patterning of the extra-embryo ni c membranes. Dev Dyn 204 , 21 9-227.
Beek, F. , Stringer, E .J ., 2010. T he ro le of Cdx genes in the gut and in ax ia l deve lopment. Biochem Soc Trans 38, 353-357.
70
Bejerano, G. , Pheasant, M ., Makunin , I. , Stephen, S. , Kent, W.J. , Mattick, J.S. , Haussier, D. , 2004. Ultraconserved elements in the human genome. Sc ience 304, 132 1-1 325.
Bel-Vialar, S. , Itasaki, N. , Krumlauf, R. , 2002 . Initiat ing Hox gene expression: in the early chick neural tube differentiai sensitivity to FGF and RA signaling subdi vides the HoxB genes in two distinct groups. Development 129, 5103-5115 .
Beland, M., Pilon, N., Houle, M. , Oh, K. , Sy lvestre, J.R. , Prinos, P. , Lohnes, D. , 2004. Cdx 1 autoregulation is governed by a novel Cdx 1-LEF 1 transcription complex. Mol Ce ll Biol24, 5028-5038.
Brazel , C.Y., Limke, T.L. , Osborne, J.K. , Miura, T., Cai , J., Pevny, L. , Rao, M.S. , 2005. Sox2 expression defines a heterogeneous population of neurosphere-forming cells in the adult murine brain. Aging Ce ll 4, 197-207.
Britsch, S. , Goerich, D.E. , Riethmacher, D. , Peirano, R.l. , Rossner, M. , Nave, K.-A. , Birchmeier, C., Wegner, M., 2001. The transcription factor Sox lü is a key regulator of peripheral gli al development. Genes & development 15, 66-78.
Brodeur, G.M ., 2003. Neurob lastoma: biological in sights into a clinica l enigma. Nat Rev Cancer 3, 203-216.
Brooke, N.M., Garcia-Fernandez, J. , Holland , P.W. , 1998. The ParaHox gene cluster is an evolutionary sister of the Hox gene cluster. Nature 392, 920-922.
Capecchi , M.R., 2005. Gene targeting in mice: functiona l ana lysis of the mammalian genome for the twenty-first century. Nat Rev Genet 6, 507-512.
Carro ll , S.B. , 2008. Evo-devo and an expand ing evo lutionary synthesis : a genetic theory of morphological evo lution. Ce li 134, 25-36.
Caso la, S., 20 1 O. Mouse model s for miRNA expression: the ROSA26 locus. Methods Mol Biol 667, i45- i 63 .
Chawengsaksophak, K., de Graaff, W., Rossant, J., Deschamps, J. , Beek, F ., 2004. Cdx2 is essential for ax ial elongation in mouse development. Proc Natl Acad Sei U S A 101 , 7641-7645.
Co las, J.F. , Schoen wo lf, G.C. , 200 1. Towards a ce llular and molecular understanding ofneuru lation. Dev Dyn 22 1, 117-145 .
Conway, S.J. , Henderson , D.J. , Kirby, M.L., Anderson , R.H. , Copp, A.J. , 1997. Development of a letha l congenital heart defect in the sp lotch (Pax3) mutant mouse. Cardiovascular research 36, 163 -173.
Copp, A.J. , Greene, N.D. , 2013 . Neural tube defects--disorders of neurulation and related embryonic processes. Wiley [nterdiscip Rev Dev Biol 2, 213-227.
,---------------------------- ·-- --
71
Copp, A.J ., Greene, N.D. , Murdoch, J.N. , 2003. T he genetic basis of mammalian neurulation. Nat Rev Genet 4, 784-793.
Co utaud, B ., 20 13 . Génération d'un nouvel out il pour l'étude in v ivo de la neurogenèse et caractérisation de la régulation et des fo nctions neura les des protéines CDX (Mémoire de maîtrise) . Un iversité du Québec à Montréal. Récupéré d'Archipel, l'archive de publications é lectroniques de I' UQAM httQ ://www .archi pel.ugam.ca/5914L
Coutaud , B. , Pilon , N. , 20 13. C haracte rization of a novel transgenic mouse line expressing Cre recombinase under the control of the Cdx2 neural spec ifi e enhancer. Genesis 51 , 777-784.
Cze ize l, A.E. , Bartfai , Z., Banhidy, F ., 20 11. Primary prevention of neura l-tube defects and sorne other congenital abnorma liti es by folie acid and multivitamins: hi story, mi ssed opportu ni ty and tasks. Th er Ad v Drug Saf 2, 173-188.
Danielian , P.S. , Muccino, D. , Rowitch, D.H. , Michael , S.K. , McMahon , A .P., 1998. Modification of gene act ivity in mouse embryos in utero by a tamoxifen-inducible form of Cre recombinase. Curr Biol8, 1323-1326.
Degenhardt, K.R. , Milewski , R.C. , Padmanabhan, A. , Mi ller, M., Singh, M.K. , Lang, D., Eng leka, K.A. , Wu, M., Li, J. , Zhou, D. , Antonucci , N. L i, L. , Epstein , J.A. , 201 O. Distinct en hancers at the Pax3 locus can function redundantly to regulate neural tube and neura l crest expressions. Dev Biol 339, 519-527.
Dixon, J. , Jones, N.C. , Sande ll , L.L. , Jayasi nghe, S.M. , Crane, J ., Rey, J.P. , Dixon, M.J. , Trainor, P.A. , 2006. Tcofl /Treac le is required for neura l crest cel ! formation a nd proliferation deficiencies that cause craniofacial abnormaliti es. Proc Natl Acad Sei U SA 103, 13403- 13408.
Draper, B .W. , Stock, D .W., Kimmel, C.B. , 2003. Zebrafish fgf24 functions with fgf8 to promote posterior mesodermal development. Development 130, 4639-4654.
Drewett, V. , Molina, H. , Mi ll ar, A. , Muller, S. , von Hesler, F. , Shaw, P.E. , 2001. DNA-bound transcription factor complexes ana lysed by mass-spectrometry : binding of nove l proteins to the human c-fos SRE and re lated seq uences. Nucleic Acids Res 29, 479-487.
Du band , J.L. , 20 1 O. Diversity in the mo lecu lar and ce llul ar strategies of ep ithe liumta-mesenchyme transitions: lnsights from the neural crest. Ce ll Adh Migr 4 , 458-482.
Dubois, N.C. , Hofmann, D. , Kaloulis , K. , Bishop, J.M ., Trumpp, A. , 2006. Nesti nCre transgenic mouse line Nes-Crel mediates highly efficient Cre/ loxP mediated recombination in the nervous system, kidney, and somite-derived tissues. Genesis 44, 355-360.
72
Dudas, M. , Sridurongri t, S., Nagy, A., Okazaki , K., Kaartinen, Y., 2004. Craniofacia l defects in mice lacking B MP type I receptor A lk2 in neura l crest ce ll s. Mech Dev 12 1' 173-182.
Dupin , E., Ca lloni , G. W., Le Douarin , N.M. , 201 O. T he cephalic neura l crest of amni ote vertebrates is co mposed of a large majority of precurso rs endowed with neura l, me lanocyti c, chond rogeni c and osteogenic potenti a lities . Cel! Cycle 9, 23 8-249.
Dupin, E., Somm er, L. , 201 2. Neura l c rest progenitors and stem ce ll s : from early development to adulthood. Dev Bio l 366, 83-95.
Durston, A.J., 201 5. T i me, space and the vertebrate body axis. Sem in Ce li Dev Bio l.
Echelard, Y., Yass il eva, G ., McMahon, A. P., 1994 . C is-acting regul atory sequences govern ing Wn t-1 express ion in the developing mouse CNS . Deve lopment 120, 22 13-2224.
E lli s, P., Fagan, B .M., Magness , S.T. , Hutton, S. , Taranova, 0. , Hayashi , S., McMahon, A. , Rao, M ., Pev ny, L. , 2004. SOX2, a persistent marker fo r multipotenti a l neura l stem cell s deri ved from embryonic stem cell s, the embryo or the adul t. Dev Neurosc i 26, 148-1 65.
Endo, Y., Osumi , N ., Wakarn atsu, Y. , 2002 . Bimodal funct ions of Notch-rnediated s igna li ng are invo lved in neural crest formati on during av ian ectoderrn deve loprnent. Deve lopment 129, 863-873.
Epste in, D.J ., 2009. C is-regul atory mutati ons in hurnan disease . Brief Funct Genorni c Proteomi c 8, 3 10-3 16.
Epste in , D.J ., Ma lo, D ., Vekemans, M., G ros, P. , 199 1a. Mo lecul ar characterizat ion of a de letion enco mpass ing the splotch mutation on mouse chromoso me 1. Genomi cs 10, 89-93.
Epste in , D.J ., Yekemans, M. , Gros, P ., 199 1 b. Splotch (Sp2H), a mutati on affecting deve lopment of the mouse neura l tube, shows a de leti on w ithin the paired homeodoma in of Pax-3. Ce ll 67, 767-774.
E ri ckson, J.O., 2002. Fo lie ac id and prevention of spina bifi da and anencepha ly. 10 years after the U.S. Publi c Hea lth Se rvice recommendat ion. MM WR Recomm Rep 51' 1-3.
Faas, L. , Isaacs, H. V., 2009. Overlapping functions of Cdx 1, Cdx2 , and Cdx4 in t he deve lopment of the amphibian Xenopus trop icali s. Dev Dyn 238, 835-852.
Feuchtbaum, L. B., Currier, R.J. , R.igg le, S., Roberson, M. , Lorey, F.W ., Cunn ingham, G .C., 1999. Neura l tube defect prevalence in Ca li fo rnia (1990-1 994): e li c iting patterns by type of defect and materna i race/ethnicity. Genet Test 3, 265-272.
73
F letcher, J.M. , Bre i, T .J ., 20 10. Introduct ion: Spina bifi da--a mult idi scipl inary perspective. Dev Disabil Res Rev 16, 1-5.
Fujii , W. , Kawasaki, K. , Sug iura, K. , Naito, K ., 201 3. Effic ient ge neration of largescale genome-modified mi ce using gRN A and CAS9 endonuc lease. N uc le ic Ac ids Res 4 1 , e 1 87 .
Gamer, L.W., W ri ght, C.V ., 1993 . M urine Cdx-4 bears stri king sim il arit ies to the Drosophil a caudal gene in its ho meodo main sequence and earl y express ion pattern. Mech Dev 43 , 71-8 1.
Garnett, A.T., Square, T.A., Mede iros, D .M. , 201 2. BMP, Wnt and FGF signa is are integrated through evo lu t ionarily conserved enhancers to achieve robust express ion of Pax3 and Z ic genes at the zebrafi sh neura l plate border. Deve lopment 139, 4220-423 1.
Gaunt, S.J. , Drage, D ., T rubshaw, R .C., 2008. lncreased Cdx prote in dose effects upon axia l patte rning in transgeni c li nes of mi ce. Deve lopment 135, 25 11-2520.
G hysen, A., 2003. The orig in and evo lution of the nervous system. lnt J Dev Biol 47, 555-562.
Gomez, C ., Ozbudak, E.M., Wunderli ch, J ., Baumann , D ., Lewis, J., Pourqui e, 0 .1
2008 . Contro l of segment num ber in vertebrate embryos. Nature 454, 33 5-339.
Go ulding, M ., Sten·er, S ., Fleming , J. , Balling, R. , Nadea u, J. , M oo re, K. , Brown, S ., Stee l, K. , Gruss, P. , 1993. Analys is of the Pax-3 gene in the mo use mutant sp lotch. Geno mi cs 17, 355-363.
Goulding, M .D. , Cha lepakis, G. , Deutsch, U. , Erse lius, J.R., G russ, P., 199 1. Pax-3, a nove lmurine DNA binding prote in expressed during early neurogenes is. EMBO J 10, 11 35-11 47.
Graham, V. , Khudyakov , J., E lli s, P., Pevny, L. , 2003. SOX2 funct ions to main ta in neura l progenitor identity. Neuron 39, 749-765.
G uth rie, S., L umsden, A. , 199 1. Form ation and regenerat ion of rhombomere boundari es in the deve loping chi ck hindbra in. Deve lopment 11 2, 22 1-229.
Hall , B.K. 1999. T he neura l c rest in deve lopment and evo lutio n. Springer Science & Business Medi a.
Harris, M.L. , Buac, K. , Shakhova, 0 ., Hakami , R.M ., Wegner, M. , Sommer, L. , Pavan, W.J. , 201 3. A dua l ro le for SOX l O in the ma in tenance of the postnata l me lanocyte lineage and the di ffe rentiati on of me lanocyte stem ce ll progenitors . PLoS Genet 9, e 1003644.
Hendrickx, M. , Van, X.H. , Ley ns, L. , 2009. Ante rior-poste rior patte rning of neura l di ffe rentiated embryonic stem cell s by canonicat Wnts , Fgfs, Bmp4 and the ir respect ive antago ni sts . Dev Growth Diffe r 51, 687-698.
74
Hermann, M. , Maeder, M.L. , Rector, K. , Ruiz, J. , Becher, B. , Burki , K ., Khayter, C ., Aguzzi, A. , Joung, J .K., Buch, T ., Pe lczar, P. , 20 12. Evaluat ion of OPEN z inc finger nucleases fo r direct gene targeting of the ROSA26 locus in mouse embryos. PLoS One 7 , e41796.
Hikasa, H. , Soko l, S.Y. , 20 13. Wnt signaling in vertebrate ax is spec ificat ion. Co ld Spri ng Harb Perspect Biol 5, a007955.
Hinman, V.F. , Nguyen, A.T. , Cameron, R.A. , Davidson , E.H. , 2003. Developmental gene regulatory network arch itecture across 500 million years of ech inoderm evo lution. Proc Nat! Acad Sei US A 100, 13356-1 336 1.
Hoekstra, H.E., Coyne, J .A. , 2007. T he locus of evo lution: evo devo and the genet ics of adaptat ion. Evo lution; international journal of organi c evo lution 6 1, 995-1016 .
Hohenstein, P. , Slight, J ., Ozdemir, D.D. , Burn , S.F., Berry, R. , Hastie, N.D. , 2008. High-efficiency Rosa26 knock-in vector construction for Cre-re gu lated overexpressio n and RN Ai. PathoGenetics 1, 3.
Holland, P.W. , Garcia-Fernàndez, J. , Wi lli ams, N .A. , Sidow, A. , 1994. Gene duplications and the ori g in s ofvertebrate development. Development 1994, 125 -1 33.
Hong, C.S. , Park, B.Y., Saint-Jeannet, J.P. , 2008. Fgf8a induces neural crest indirect ly through the activat ion of Wnt8 in the paraxial mesoderm. Development 135, 3903-39 1 O.
Houatt, C., Caneparo, L. , Heisenberg, C., Batth, K. , Take-Uchi , M. , Wilson, S. , 2002. Estab lishment of the telencephalon during gastrul ation by loca l antago ni sm of Wnt signa ling. Neuron 35, 255 -265.
Houle, M., A llan, 0., Lohnes, 0 ., 2003a. Cdx homeodomain proteins in vertebral patterning, Advances in Developmental Biology and Biochemistry. E lsevier, pp. 69-105.
Hou le, M., Sy lvestre, J.R. , Lohnes, 0., 2003b. Retinoic acid regulates a subset of Cdx 1 function in vivo . Oeve lopment 130, 6555-6567 .
Howard , M.L. , Davidson , E.H., 2004. c is-Regulatory control circu its in development. Dev Biol 27 1, 109- Il 8.
Hsu, P.O. , Lander, E.S., Zhang, F. , 20 14. Development and app li cations ofCRISPRCas9 for genome engineering. Ce ll 157, 1262-1278.
Hu , N. , Strob l-Mazzu ll a, P.H. , Bronner, M .E., 20 14. E pigenetic regulation in neural c rest development. Dev Biol 396, 159-168.
Huang, X. , Sa in t-Jeannet, J.P. , 2004. Induction of the neural crest and the opportunities of !ife on the edge . Dev Biol 275 , 1-11.
75
Hubner, N.C ., Nguyen, L.N. , Hornig, N.C., Stunnenberg, H.G. , 20 15 . A quantitative proteomics too l to identi fy DNA-protein interacti ons in primary cells or blood. J Proteome Res 14, 13 15-1 329.
Ii mura, T., Denans, N. , Pourquie, 0. , 2009. Establi shment of Hox vertebral identities in the embryo nic spine precursors. Curr Top Dev Bi ol 88, 201-234.
lsaacs, H.V. , Pownall , M.E. , S lack, J .M. , 1994. eFGF regul ates Xbra express ion during Xenopus gastrulation. EMBO J 13, 4469-4481.
Jandz ik, 0., Garnett, A.T. , Square, T.A. , Catte ll , M.V. , Yu, J.K. , Medeiros, D.M. , 20 15 . Evo lution of the new vertebrate head by co-option of an anc ient chordate ske letal tissue. Nature 5 18, 534-537.
Jessen, K.R., Mirsky, R. , 2005 . The o rig in and development of g lial ce ll s in peripheral ne rves. Nat Rev Ne urosci 6, 671-682.
Jiang, M. , Stanke, J., Lahti , J.M., 20 Il. The connections between neural crest development and neuroblasto ma. C urT Top Dev Bio l 94, 77-127.
Kasparek, P. , Krausova, M., Haneckova, R., Kriz, Y., Z bodakova, 0 ., Korinek, V. , Sedl acek, R., 2014. Effici ent gene targeting of the Rosa26 locus in mouse zygotes using TALE nucleases. FEBS Lett 588, 3982-3988.
Keenan, T.D. , Sharrard, R.M ., Tsaacs, H.V. , 2006. FGF s igna l transduction and the regulation of Cdx gene express ion. Dev Biol 299, 478-488.
Kiecker, C. , N iehrs, C. , 200 l. A morphogen grad ient of Wnt/beta-catenin s igna lling regulates anteroposterior neura l patterning in Xenopus. Deve lopment 128, 4189-4201.
Kim, H. , Kang, K. , Ekram, M.B. , Roh, T.Y. , Kim , J., 2011. Aebp2 as an epigeneti c regu lator for neural crest ce ll s. PLoS One 6, e25 174.
Ki shi , M. , Mizuseki , K. , Sasa i, N. , Yamazaki , H. , Shiota, K. , Nakani shi , S., Sasa i, Y. , 2000. Requirement of Sox2-med iated s ignaling for di ffe rentiation of early Xenopus neuroectodenn . Development 127, 791-800.
Kobayashi , K. , Jakt, L.M. , Nishikawa, S.J ., 20 13. Epigenetic regul at ion of the neurobl astoma genes, Arid3b and Mycn. Oncogene 32, 2640-2648.
Kong, Q. , Hai , T., Ma, J., Huang, T. , Jiang, 0., X ie, 8., Wu, M., Wang, J. , Song, Y., Wang, Y., 20 14. Rosa26 locus supports tissue-specifie promoter dri ving transgene express ion spec ifically in pig.
Koo, T. , Lee, J ., Kim , J. S., 2015. Measuring and Reducing Off-Target Activities of Programmable N ucleases lncluding C RJ SPR-Cas9. Molecules and cell s 38, 475-481.
Kranz, A.L. , Ei ls, R., Konig, R. , 20 11. Enhancers regulate progress ion of deve lopment in mamma li an ce ll s. N uc le ic Acids Res 39, 8689-8702.
76
Krispin , S., N itzan, E. , Kassem, Y. , Ka lche im, C. , 201 O. Ev idence for a dynami c spatiotempora l fa te map and earl y fa te restri cti o ns of premi grato ry a vian neura l crest. Deve lopment 137, 585-595 .
K udoh, T. , W il so n, S. W. , Daw id, LB ., 2002. Dist inct ro tes fo r Fgf, Wnt and retino ic ac id in posteriori z ing the neural ectoderm . Deve lopment 129, 4335 -4346.
Lev ine, M., 201 O. Transcriptiona l enhancers in animal deve lopment and evo luti o n. Curr Bio l 20, R754-763.
Lewis, A.E. , Vasudevan, H .N. , O'Ne ill , A. K., S01· iano, P., Bush, J .O., 201 3. T he w ide ly used Wntl-Cre transgene ca uses deve lopmenta l phenoty pes by ectopie acti vation of Wnt signa ling. Dev Bio l 379, 229-234.
Li, J ., Chen, F. , Epste in, J.A. , 2000. Neura l crest express ion of Cre recombi nase directed by the prox ima l Pax3 pro mote r in transgenic mice. Genes is 26, 162-1 64 .
Li, J ., Liu , K.C ., Jin , F. , Lu, M.M. , E pste in , J .A., 1999. T ransgenic resc ue of co ngenita l heart di sease and spina bifida in Splotch mi ce. Deve lopment 126, 2495-2503.
Li , J ., Shi , Y., Sun , J., Z hang, Y., Mao, B. , 20 Il. Xenopus reduced fo late carri e r regul ates neura l crest development epigeneti ca lly . PLoS One 6, e271 98.
Limaye, A. , Ha ll , B., Kulkarni , A. B., 2009. Manipul ation of mouse embryoni c ste m ce ll s for knocko ut mouse producti on. CurTent protoco ls in ce ll bio logy 1 editori a l board, Juan S. Boni fac ino .. . [et a l.] Chapter 19, Unit 19 13 19 13 11-24.
Little, S.C., Mullin s, M.C., 2006. Extrace llul ar modul atio n of BMP activity in patte rning the dorsoventra l ax is. Birth Defects Res C Embryo Today 78, 224-242.
Liu, Y. , X iao, A. , 20 li . Epigeneti c regul ati on in neural c rest deve lopment. Birth Defects Res A C lin Mol Terato l 9 1, 788-796.
Lodi sh, H., Berk, A. , Z ipursky, S.L. , a l. , e ., 2000. Recombinati on between Homologo us DNA Sites. Mo lecula r Ce ll Bio logy. 4th edi tion., Secti on 12. 15 Available f rom: http://www.ncb i.n lm .nih.gov/books/N BK2 1725/.
Loh, K. M., Lim, B., 20 Il . A precari ous balance: pluripotency facto rs as lineage spec ifiers. Ce ll stem ce l! 8, 363 -369.
Lohnes, D ., 2003 . T he Cdx 1 homeodomain prote in: an in tegrato r of posteri or s igna ling in the mouse . Bioessays 25 , 97 1-980.
Lynch, V.J ., Wagner, G. P., 2008 . Resurrecting the ro te of transcript ion factor change in deve lopmenta l evo luti on. Evo luti on; intern ationa l j ourna l of o rganic evo luti on 62, 2 131-2 154.
MacD ona ld, B.T. , Tama i, K., He, X. , 2009. W nt/beta-catenin sig na ling: components, mechani sms, and di seases. Dev Ce ll 17, 9-26 .
77
Magli , A. , Schnettler, E. , Rinaldi , F. , Bremer, P. , Perlingeiro, R.C. , 2013 . Functional dissection of Pax3 in paraxial mesoderm development and myogenesis. Stem Cell s 3 1, 59-70.
Marchant, L. , Linker, C ., Ruiz, P. , GuetTera, N. , Mayor, R. , 1998. The inductive properties of mesoderm suggest that the neura l crest cells are spec ified by a BMP grad ient. Dev Biol 198 , 3 19-329.
Marieb, E.N. , Hoehn, K. , 2007. Human anatomy & physiology, 7th ed. Pearson Benjamin Cummings, San Francisco.
Marletaz, F., Maeso, I. , Faas, L. , lsaacs, H .V. , Holland, P.W., 20 15 . Cdx ParaHox genes acqu ired distinct developmental roles after gene duplication in vertebrate evo lution. BMC biology 13 , 56.
Marom, K. , Shapira, E. , Fainsod, A. , 1997. The chicken caudal genes estab li sh an anterior-posterior gradient by partially overlapp ing temporal and spat ial patterns of express ion. Mech Dev 64, 41-52.
Mason, 1., 2007. Initi at ion to end point: the multiple roles of fibroblast growth factors in neural development. Nat Rev Neurosci 8, 583-596.
Mayor, R. , Theveneau , E. , 2013. T he neural crest. Development 140, 2247-225 1.
McBurney, M.W., 1993. Pl9 embryonal carc inoma ce ll s. TntJ Dev Biol37, 135-140.
McGrew, L.L. , Hoppler, S., Moon, R.T. , 1997. Wnt and FGF pathways cooperatively pattern anteroposterior neural ectoderm in Xenopus. Mech Dev 69, 105-114.
McKinney, M.C. , Fukatsu , K., Morrison, J. , McLennan, R. , Branner, M.E. , Kulesa, P.M. , 2013 . Ev idence for dynamic rearrangements but lack of fa te or position restri ctions in premigratory av ian trunk neural crest. Development 140, 820-830.
Meulemans, D. , Branner-Fraser, M. , 2002. Amphioxus and lamprey AP-2 genes: implications for neura l crest evo lution and migration patterns. Development 129, 4953-4962.
Meulemans, D., Branner-Fraser, M., 2004. Gene-regu latory interactions in neural crest evo lution and deve lopment. Dev Ce l! 7, 29 1-299.
Meyer, B.T. , Gruss, P. , 1993. Mouse Cdx-1 expressio n during gastrulation. Development 117, 191-203.
Meyer, M. , de Angelis , M.H. , Wurst, W. , Kuhn, R. , 20 1 O. Gene targeting by homologous recombination in mouse zygotes mediated by z inc-finger nucleases. Proc Nat! Acad Sei US A 107, 15022-15026.
Milewski , R.C. , Chi , N.C. , Li, J. , Brown, C. , Lu, M.M. , Epste in , J.A., 2004. Identification of minimal enhancer elements sufficient for Pax3 express ion in neura l crest and implication ofTead2 as a regulator of Pax3. Development 131, 829-837.
78
M isra, R.P., Duncan, S.A. , 2002. Gene ta rgeting in the mouse: advances in introduct ion of transgenes in to the genome by homo logous recombination. Endocrine 19, 229-238 .
M lodz ik, M ., Fjose, A. , Gehring, W.J ., 1985. Iso lati on of cauda l, a Drosophila homeo box-conta ining gene w ith maternai ex press ion, whose transc ripts fo rm a co ncentration gradi ent at the pre-blastoderm stage. EMBO J 4, 296 1-2969.
Moase, C.E., Tras ler, D .G. , 1992. Sp lotch locus mouse mutants : mode ls fo r neura l tu be defects and Waardenburg syndrome ty pe l in hu mans. J Med Genet 29, 145-151.
Monsoro-Bu rq, A .H ., Wang, E. , Harland, R. , 2005. Msx l and Pax3 cooperate to medi ate FGF8 and WNT s ignais during Xenopus neural crest induction. Dev Cell 8, 167-178.
Mulley, J .F., Chiu, C. H., Ho ll and, P. W., 2006. B reakup of a homeobox c luster after genome dupli cat ion in te leosts. Proc Natl Acad Se i US A 103 , 10369- 10372.
M ull ey, J.F ., Holland, P. W., 20 1 O. Para i le i retention of Pdx2 genes in cart il ag inous fish and coelacanths. Mo lecular bio logy and evo lution 27, 2386-239 1.
N iehrs, C., 2004. Regionall y spec ifi e induction by the Spemann-Mango ld organizer. N at Rev Genet 5, 425 -434.
Nyabi , 0., Naessens, M ., Ha igh, K., Gembarska, A. , Goossens, S., Maetens, M. , De Clercq, S., D rogat, B. , Haeneba lcke, L., Bartun kova, S., De Vos, 1. , De Craene, B. , Karimi , M., Berx, G., Nagy, A. , Hil so n, P., Marine, J.C. , Ha igh, J.J. , 2009. Effi c ie nt mouse transgenes is us ing Gateway-co mpat ibl e ROSA26 locus targeting vectors and F l hybrid ES cell s. N ucle ic Acids Res 37, e55 .
Okuda, Y., Ogura, E ., Ko ndoh, I-l. , Kamachi , Y., 201 O. B 1 SOX coo rdinate ce ll spec ifi cat ion w ith patte rning and morphogenes is in the early zebrafi sh embryo. PLoS Genet 6, e 1000936.
O liveri , P ., Dav idson, E.H. , McC iay, D.R., 2003. Activation of pmar l co ntra is spec ifi cation of mi cromeres in the sea urchin embryo . Dev Bio l 258 , 32-43.
Owens, J .B ., Mauro, D ., Stoytchev, 1. , Bhakta, M.S. , Kim , M.S ., Sega l, D.J ., Mo isyadi , S., 201 3. T ranscription acti vator like effector (TALE)-directed piggyB ac transpos ition in hu man ce li s. N uc leic Ac ids Res 4 I, 9 197-9207.
Papanayotou, C., Mey, A., Birot, A.M., Saka, Y. , Boast, S., Smi th, J.C. , Samaru t, J ., Stern , C.D ., 2008. A mechani sm regul at ing the onset of Sox2 express ion in the embryoni c neura l plate . PLoS Bio l 6, e2.
Paratore, C. , Eichenberger, C. , Suter, U ., Sommer, L., 2002. Sox 10 haplo insuffi c iency affects main tenance of progenitor ce lls in a mouse model of Hirschsprung di sease. Hum Mol Genet Il , 3075-3085.
79
Pares-Matos, E . !. , 20 13. E lectrophoret ic mo bi 1 ity-shift and super-sh ift assays for studies and characterizat ion of protein-DNA complexes. Methods Mol Biol 977, 159-167.
Patten , l. , Placzek, M. , 2002. Opponent activit ies of Shh and BMP signa ling during floor plate induction in vivo . Curr Biol 12, 47-52.
Peinado, H., Portillo, F. , Ca no, A ., 2004. Transcriptiona l regulation of cad heri ns during deve1opment and carc inogenes is. Int J Dev Biol48, 365 -375.
Pennacchio, L.A. , Ah itu v, N. , Moses, A.M., Prabhakar, S. , Nobrega, M.A., Shoukry, M., Minovitsky, S. , Dubchak, 1. , Holt, A., Lewis, K.D. , Plajzer-Frick, 1. , Akiyama, J. , De Val, S. , Afza l, V. , Black, B.L. , Cou ronne, 0., Eisen, M.B., Vise! , A. , Rubin, E.M. , 2006. In vivo enhancer analysis of human conserved non-coding sequences. Nature 444, 499-502.
Pennacchio, L.A. , Bickmore, W. , Dean , A. , Nobrega, M.A. , Bejerano, G. , 2013. E nhancers: five essentia l questions. Nat Rev Genet 14, 288-295.
Pennisi , E. , 2008. Evo lutionary biology. Deciphering the genet ics of evolution. Science 32 1, 760-763.
Pen·is, R. , 1997. The extracellu lar ma tri x in neura l crest-ce l! migration. Trends Neurosci 20, 23 -31.
Petersen, C.P., Reddien, P. W. , 2009. Wnt signa ling and the polarity of the primary body axis. Ce l! 139, 1056-1 068.
Pevny, L.H., Nico li s, S.K. , 201 O. Sox2 rotes in neural stem cel ls. lnt J Biochem Ce l! Biol42, 421-424.
Pilon, N. , Oh, K. , Sylvestre, J.R. , Bouchard, N. , Savory, J. , Lohnes, 0. , 2006. Cdx4 is a direct target of the canonica t Wnt pathway. Dev Biol 289, 55-63.
Pilon, N. , Oh, K. , Sy lvestre, J.R. , Savory, J.G. , Lohnes, D. , 2007. Wnt signaling is a key mediator of Cdx1 express ion in vivo. Development 134, 23 15-2323.
Pitkin, R.M. , 2007. Folate and neural tube defects. Am J C lin Nutr 85 , 285S-288S.
Potterf, S.B. , Mollaaghababa, R. , Hou , L. , Southard-Sm ith, E.M ., Hornyak, T.J ., Arn heiter, H. , Pa van, W .J ., 200 1. Ana lysis of SOX L 0 function in neura l crest-derived melanocyte development: SOX 1 0-dependent transcriptional control of dopachrome tautomerase. Developmental bio logy 237, 245-257.
Prinos, P. , Joseph, S., Oh, K. , Meyer, B.!. , Gruss, P. , Lo hnes, D. , 200 L. Multiple pathways governing Cdxl expression during murine development. Dev Biol 239 , 257-269.
80
Pruitt, S.C., Bussman, A. , Mas lov , A. Y. , N ato li , T. A. , 1-Ie inaman, R ., 2004. Hox/Pbx and Brn binding s ites medi ate Pax3 express ion in v itro and in vivo . Gene Expr Pattern s 4 , 671-685 .
Puli t i, A., Prehu, M.O., S imon-Chazottes, D. , Ferkdadji , L., Peuchmaur, M., Goossens, M. , Guenet, J. L. , 1995 . A high-reso lution genetic map of mouse chromoso me 15 encompass ing the Dominant megaco lon (Dom) locus. Mamm Genome 6, 763-768.
Purves, D., Willi ams, S.M., 200 l. N eurosc ience, 2nd ed. S inauer Associates, Sunderland, Mass.
Rada-Ig les ias, A. , Bajpa i, R. , Prescott, S. , Brugmann , S.A ., Swigut, T. , Wysocka, J. , 201 2 . Epigenomi c annota ti on of enhancers predi cts transcriptiona l regulators of hu man neura l crest. Cel! Stem Ce l! 11 , 633-648 .
Rada-I g les ias, A., Prescott, S .L., Wysocka, J., 201 3. Human geneti c va ri ation within neural cres t enhancers: mo lecular and phenotypic impli cati ons. Philos T ran s R Soc Lond B Bio l Se i 368 , 201 20360.
Rex, M. , Orme, A. , Uwanogho, D., To inton, K., Wigmore, P .M., Sharpe, P.T. , Scotting, P.J ., 1997. Dynami c expression of chi cken Sox2 and Sox3 genes in ectoderm induced to fo rm neural ti ssue. Dev Dyn 209, 323-332.
Rhead, B. , Karo lchik, D. , Kuhn, R.M., Hinrichs, A.S. , Zweig, A.S., Fujita, P.A. , Diekhans, M. , Smith, K. E., Rosenbloom, K.R. , Raney, B.J. , Pohl , A., Pheasant, M. , Meyer, L.R. Learned, K., Hsu, F. , Hillman-Jackson, J. , Harte, R.A. , Giardine, B ., Dreszer, T. R. , Clawson, H. , Barber, G.P. , Hauss ier, D. , Kent, W.J., 201 O. The UCSC Genome Browser database: update 201 O. N uc leic Ac ids Res 38, 0 613-6 19.
Russe l, W., 1947. Splotch, a new mutation in the house mouse. Mus musculus.
Sanche:t:-Ferras , 0 ., Berna,, G., Laberge-Perraul t, E., Pilon, N. , 20 14 . Induction and dorsa l restri ction of Pa ired-box 3 (Pax3) gene expression in the cauda l neuroectoderm is medi ated by integration of multipl e pathways on a short neural c rest enhancer. Biochim Biophys Acta 1839, 546-55 8.
Sanchez-Ferras, 0 ., Coutaud, B., Djavanbakht Samani , T. , T remblay, l. , So uchkova, 0 ., Pilon, N ., 201 2. Caudal-re lated homeobox (Cdx) prote in-dependent integrati on of canoni ca t Wnt signaling on paired-box 3 (Pax3) neura l crest enhancer. J Bio l Chem 287, 16623- 16635.
Santa rius, T. , Shipl ey, J., Brewe r, D., Stratto n, M .R., Cooper, C.S. , 201 O. A census of amplifi ed and overexpressed hu man cancer genes. a ture Rev iews Cancer 10, 59-64 .
Sato, T ., Sakuma, T ., Yokoni shi , T. , Katagiri , K. , Kamimura, S., Ogonuki , N. , Ogura, A. , Yamamoto, T. , Ogawa, T. , 201 5. Genome Editing in Mouse Spermatogo ni al Stem Ce l! L ines Us ing TALEN and Doubl e-Nicking CRJ SPR/Cas9. Stem ce l! repot1s 5, 75-82 .
81
Savory, J .G. , Bouchard, N. , Pierre, V. , Rijli , F.M. , De Repentigny, Y. , Kothary, R. , Lohnes, D. , 2009a. Cdx2 regulation of posterior development through non-Hox targets. Development 136, 4099-41 1 O.
Savory, J.G. , Mansfield, M. , Rijli , F.M. , Lohnes, D. , 20 ll a . Cdx mediates neura l tube closure through transcriptional regulation of the planar ce ll polarity gene Ptk7 . Deve lopment 138, 136 1-1 370.
Savory, J.G. , Mansfield , M., St Lou is, C. , Lohnes, D., 20 11b. Cdx4 is a Cdx2 target gene. Mech Dev 128, 41-48.
Savo ry, J.G. , Pilon, N. , Gra inger, S., Sylvestre, J .R., Beland, M. , Houle, M. , Oh, K., Lohnes, D., 2009b. Cdx 1 and Cdx2 are functionally eq uiv alent in vertebral patterning. Dev Biol330, 114-1 22.
Schu lte, J.H. , L indner, S. , Bohrer, A. , Maurer, J. , De Preter, K. , Lefever, S., Heukamp, L. , Schulte, S. , Molenaar, J. , Versteeg, R ., Thor, T. , Kunkele, A. , Vandesompe le, J. , Spe leman, F. , Schorle, H. , Eggert, A ., Schram m, A. , 20 13. MYCN and ALKF 1174L are suffic ient to drive neuroblastoma development from neural crest progenitor cells. Oncogene 32, 1059-1065.
Sever, L.E. , 1995 . Looking fo r causes of neural tube defects: w here does the env ironment fit in? Environ Health Perspect 103 Suppl 6, 165 -1 71.
Shakhova, 0., Z ingg, D. , Schaefer, S.M ., Hari, L., C ivenni , G. , Blunschi , J., C laudinot, S. , Okoniewski , M. , Beermann, F ., Mihic-Probst, D. , Moch, H. , Wegner, M. , Dummer, R. , Ban·andon, Y. , C ine lli , P. , Sommer, L., 20 12. Sox lü promotes the format ion and mai ntenance of g ia nt congenita l naevi and melanoma. Nat Ce ll Biol 14, 882-890.
Shastry, B. , 1995. Overexpression of genes in health and sickness. A bird's eye v iew. Co mparat ive Biochemistry and Physiology Part B: Biochemistry and Molecular Biology 11 2, 1-13 .
Shimizu, T. , Bae, Y.K. , Muraoka, 0., Hibi , M. , 2005. Interaction of Wnt and caudalrelated genes in zebrafish posterior body formation . Dev 8 iol 279, 125- 141.
Shirley, S.H. , Greene, V.R. , Duncan , L.M. , Torres Cabala, C.A., Grimm, E.A. , Kusewitt, D.F., 20 12. S lug expression during me lanoma progression. A m J Pathol 180, 24 79-2489 .
S ilberg, D.G. , Swain, G.P. , Suh, E.R. , Traber, P.G. , 2000. Cdx 1 and cdx2 express ion during intestinal development. Gastroenterology 11 9, 96 1-97 1.
S imoes-Costa, M. , Branner, M.E. , 20 13. lnsights into neura l c rest development and evo lution fro m genomic ana lysis . Genome Res 23 , 1069- 1 080.
Si moes-Costa, M. , Branner, M.E., 20 15. Estab li shing neural crest identity: a gene regulatory recipe . Development 142 , 242-257.
82
Smith, J. L. , Schoenwo lf, G.C. , 1997. Neurul ati on: com1ng to c losure. T rend s Neurosc i 20, 51 0-51 7 .
Smith, W. C. , Harl and, R.M ., 1992. Express ion c lo ning of nogg in, a new dorsa liz ing factor localized to the Spemann organizer in Xenopus embryos . Ce ll 70, 829-840.
Soriano, P., 1999. Genera lized lacZ express ion w ith the RO SA26 Cre reporter stra in . N atGenet 2 l , 70-71.
Southard- Smith , E.M. , Kos, L., Pavan, W.J., 1998. Sox iO mutati on di srupts neural crest development in Dom Hirschsprung mouse mode l. N at Genet 18, 60-64.
Srinivas, S ., Watanabe, T ., Lin , C.-S ., Willi am, C.M. , Tanabe, Y., Jesse ll , T.M ., Costantini , F ., 200 l. Cre reporter stra ins p roduced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Deve lopmenta l Bio logy 1.
Stottmann , R.W., Berrong, M. , Matta, K. , C ho i, M. , Klin gensmith, J. , 2006. T he BMP antagoni st Noggin promotes crani a l and spinal neurulation by di stinct mechani sms. Dev Bio l 295 , 647-663.
Strumpf, 0 ., Mao, C.A. , Yamanaka, Y., Ralston, A ., C hawengsa ksop·hak, K., Beek, F. , Rossant, J ., 2005. Cdx2 is required for correc t ce ll fa te spec ificati on and differenti ati on of trophectoderm in the mouse bl astocyst. Deve lopment 132, 2093-2 102.
Stuhlmiller, T.J. , Garc ia-Castro, M.l. , 2012. Current perspectives of the signaling pathways directin g neural crest inducti on. Cell Mol Life Se i 69, 3715-3737.
Subramanian, V ., M eyer, B.l. , Gruss, P., 1995. Di srupti on of the murine homeobox gene Cdx l affects ax ia l ske leta l identities by a lte ring the mesoderm a l express io n domains of Hox genes. Ce ll 83 , 64 1-653.
Takemoto, T ., Uchikawa, M., Kamachi , Y., Kondoh, H. , 2006. Convergence of Wnt and FGF s igna is in the genes is of posterior neura i pi ate through acti vat ion of the Sox2 enhancer N -1 . Development 133 , 297-306.
Tam, P.P., Behringer, R .R ., 1997. Mouse gastrul at ion: the formation of a mamma li an body pl an. Mech Dev 68, 3-2 5.
T homson, M., Liu, S.J ., Zou, L.-N. , Smith, Z ., Me issner, A. , Ramanath an, S. , 2011 . Pluripotency factors in embryo nic stem ce ll s regul ate di ffe renti ation into germ layers. Ce ll 145 , 875-889.
Tra inor, P ., 201 3. Neura l c rest ce ll s : evo lution, deve lopment and di sease. Academie Press.
T ra ino r, P. A. , 201 O. C rani ofac ia l birth defects: The ro le of neura l crest ce ll s in the etio logy and pathogenes is of T reacher Co llins syndro me and the potenti a l fo r prevention. Am J Med Genet A 152A, 2984-2994.
83
Tronche, F. , Kellendonk, C. , Kretz, 0., Gass, P. , Anl ag, K. , Orban, P.C., Bock, R ., Klein , R. , Schutz, G. , 1999. Disruption of the g lucoco rtico id receptor gene in the nervo us system results in reduced anx iety. Nat Genet 23 , 99-103.
T uazo n, F.B ., Mullins, M.C. , 20 15. Temporally coordinated s igna is progressively patte rn the ante roposter ior and dorsoventra l body axes. Sem in Ce ll Dev Bio l.
Uchi kawa, M. , Yoshida, M. , lwafuchi-Do i, M. , Matsuda, K. , lshida, Y. , Takemoto, T. , Kondoh , H. , 20 11. Bl and 82 Sox gene expression during neura l plate development in chicken and mouse embryos: uni versa l versus species-dependent features. Dev Growth Diffe r 53 , 761-771.
van den Akker, E. , Forlani , S. , Chawengsaksophak, K. , de Graaff, W. , Beek, F. , Meyer, B.I. , Deschamps, J. , 2002. Cdx1 and Cdx2 have overl app ing functions in anteroposter ior patterning and posterior axis e longation. Development 129, 2 181-2 193.
van Nes, J. , de Graaff, W. , Lebrin , F. , Gerhard , M. , Beek, F. , Deschamps, J. , 2006. T he Cdx4 mutat ion affects axia l development and reveals an essentia l ro le of Cdx genes in the ontogenesis of the placenta! labyrinth in mi ce. Development 133, 419-428.
Van Otterloo, E., Corne ll , R. A., Medeiros, D.M. Garnett, A.T. , 20 13. · Gene regulatory evo lution and the ori g in of macroevo luti onary novelties: in sights from the neural crest. Genesis 51 , 457-470.
Wakamatsu, Y. , Watanabe, Y. , Nakamura, H. , Kondoh , H. , 1997 . Regulat ion of the neural crest ce l! fate by N-myc: promotion of ventra l migration and neurona l differentiation. Development 124, 1953-1 962.
Wang, W.C. , Shashikant, C.S. , 2007. Evidence for posit ive and negat ive regul ation of the mouse Cdx2 gene. J Exp Zoo! B Mol Dev Evo l 308, 308-32 1.
Weiss, M.B., Abel, E.V. , Mayberry, M.M. , Basile, K.J. , Berger, A.C ., Apl in, A.E. , 20 12. TW IST ! is an ERK I/2 effector that promotes invas io n and regulates MMP-1 express ion in human melanoma ce ll s. Cancer Res 72, 6382-6392.
Weiss, W.A. , Aldape, K. , Mohapatra, G. , Feuerstein, B.G., Bishop, J.M. , 1997. Targeted expression of MYCN causes neuro blastoma in transgenic mice. EMBO J 16, 2985-2995.
Weston, J .A. , Butler, S.L. , 1966. Tempora l factors affecting loca l ization of neura l crest ce ll s in tbe chicken embryo. Dev Biol14, 246-266 .
Wittkopp, P.J. , Kalay, G ., 20 12. Cis-regu la tory e lements : mo l ecu lar mec han isms and evo lutionary processes underlyi ng divergence. Nature Reviews Genet ics 13, 59-69.
Wray, G.A. , 2007. The evo lutionary sign ifi cance of cis-regulatory mutations. Nat Rev Genet 8, 206-2 16.
84
X u, F., Li, H ., Jin , T. , 1999. Ce l! ty pe-spec ifie autoregulati on of the Caudal-re lated homeobox gene Cdx-2/3 . J Biol C hem 274, 34310-343 16.
Yamaguchi , Y., Miura, M ., 20 13. How to form and c lose the brain : insight into the mechani sm of crania l neura l tube c los ure in mammal s. Ce l! Mo l Life Sei 70, 317 1-3 186.
Yang, J ., Mani , S.A., Donaher, J .L. , Ramaswamy, S. , Itzykso n, R.A., Come, C. , Savagner, P. , Gite lman, f. , Richardson , A. , Wei nberg, R.A ., 2004. Twist, a maste r regulator of morphogenes is, plays an essent ia l ro le in tumor metastas is. Ce ll 117, 927-939.
Ybot-Gonza lez, P ., Cogram, P., Gerre lli , 0., Copp, A.J. , 2002. Sonic hedge hog and the mo lecul ar regulation of mo use neura l tube c los ure. Oeve l op ment 129, 2507-2517.
Ybot-Gonza lez, P., Gaston -Massuet, C ., G ird ler, G. , Klingensmith, J ., Arke ll , R. , G reene, N .D. , Copp, A .J., 2007. Neura l plate morphogenes is during mouse neurulation is regul ated by antagonism of Bmp s igna lling . Development 134, 3203-32 11 0
Zambrowicz, B.P. , Imamoto, A. , Fiering, S., Herzenberg, L.A ., Kerr, W.G. , Sorian o, P. , 1997. Di sruption of overl apping transcripts in the ROSA beta geo 26 gene trap stra in leads to w idespread express ion of beta-ga lactos idase in mouse em bryos and hematopo ieti c ce ll s. Proc Nat! Acad Se i U SA 94, 3789-3794.
Z hang, S. , 20 14. Sox2, a key factor in the regu lation of pluripotency and neural diffe rentiation. World Jo urna l of Stem Ce ll s 6, 305.
Z hao, T. , Gan , Q. , Stokes, A., Lass iter, R.N. , Wang, Y., Chan , J. , Han, J .X. Pl easure, D. E., Epstein , J. A. , Z hou, C.J ., 20 14. be ta-catenin regu lates Pax3 and Cdx2 for caudal neura l tube c losure and e longat io n. Deve lopment 141 , 148-1 57 .