Investigating the Interplay of Cyclic Nucleotide Gated Ion

Channel 2 and Auxin in Immune Signaling

By Alexander Fortuna

A thesis submitted in conformity with the requirements

for the degree of Master of Science Department of Cell and Systems Biology

University of Toronto

© Copyright by Alexander Fortuna 2015

ii

Investigating the Interplay of Cyclic Nucleotide Gated Ion Channel 2 and Auxin in

Immune Signaling

Alexander Fortuna

Master of Science

Cell and Systems Biology

University of Toronto

2015

Abstract

Cyclic nucleotide-gated ion channels (CNGCs) are non-selective, ligand-gated

cation channels present across eukaryotes. In Arabidopsis the CNGC family contains 20

members, which are believed to play important roles in biotic and abiotic stress

responses, ion homeostasis, and development through their Ca2+ conducting abilities.

Several CNGCs have been implicated in plant-pathogen interactions through genetic

studies. The defense, no death mutants dnd1 and hlm1/dnd2 are null mutants of the

closely related Arabidopsis CNGCs, CNGC2 and CNGC4, respectively, and have

distinct autoimmune phenotypes. Though these mutants have been well characterized

phenotypically, CNGC-mediated signal transduction is poorly understood. In order to

understand CNGC2-mediated defense signaling, I have investigated the first dnd1

suppressor mutant, repressor of defense no death 1 (rdd1-1D). In this thesis, I aimed to

understand the molecular mechanism by which rdd1-1D is able to suppress dnd1-

conferred phenotypes. Current data indicates that rdd1-1D is a loss-of-function mutation

in the auxin biosynthesis gene YUCCA6.

iii

Acknowledgments

I would like to sincerely thank my supervisor Dr. Keiko Yoshioka for the

opportunity to pursue my MSc degree in her lab and for inspiring me with her passion

and dedication for scientific research. I appreciated Dr. Yoshioka’s continuous support

and expertise throughout my graduate research studies and in the writing of this thesis.

I found Dr. Yoshioka’s patience, motivation, and immense knowledge truly invaluable.

Under her guidance, the Yoshioka lab provided me the opportunity to work

alongside some amazing people. Many thanks to current members Purva Karia,

Christine Cao and Dr. Wolfgang Moeder. Many thanks also to past members Kimberly

Chin and Huoi Ung. I would especially like to thank Tom DeFalco for his excellent

advice and encouragement over the last two years. I would also like to thank

undergraduate students, Jihyun Lee, Catherine Vo, Eugenia Daradur, Tim Xue, Megumi

Bachmann and Maxwell Olsen for their dedication and assistance with my research.

A sincere thank you also to the committee members, Dr. Darrell Desveaux and

Dr. Eiji Nambara for their guidance and helpful suggestions during my committee

meetings and to Dr. Thomas Berleth for his helpful perspective and discussions. I would

also like to thank Dr. Adriana Caragea and Dr. Wenzi Kurshumova for their assistance

in this study.

Finally, I would like to thank my friends and family members who have been

extremely supportive and encouraging during my graduate studies.

iv

Table of Contents

Abstract ........................................................................................................................................... ii

Acknowledgments .......................................................................................................................... iii

List of Publications ...................................................................................................................... viii

Table of Contents ........................................................................................................................... iv

List of Tables ............................................................................................................................... viii

List of Figures ................................................................................................................................. x

List of Appendices ........................................................................................................................ xii

List of Abbreviations ................................................................................................................... xiii

Chapter 1 Introduction ................................................................................................................. 1

1.1 Plant Immunity Signaling ................................................................................................... 2

1.2 Importance of ion flux in immune signaling ....................................................................... 3

1.3 Cyclic-Nucleotide Gated (CNGCs) .................................................................................... 4

1.4 Identification of repress of defense, no death 1 ................................................................ 11

1.5 Flavin-containing monooxygenases (FMOs) .................................................................... 13

1.6 Flavin containing monooxygenases in Arabidopsis ......................................................... 13

1.7 FMOs in auxin biosynthesis.............................................................................................. 14

1.8 Importance of localized auxin production and YUCCA gene expression ......................... 15

1.9 Auxin and Plant Immunity ................................................................................................ 17

1.10 Auxin and Ca2+ signaling .................................................................................................. 18

1.11 Thesis Aim and Overview ................................................................................................ 19

Chapter 2 Cyclic Nucleotide Gated Ion Channel 2: a new role in floral transition ................... 21

Abstract .................................................................................................................................... 22

2.1 Introduction ....................................................................................................................... 23

2.2 Materials and Methods ...................................................................................................... 25

v

2.2.1 Plant Growth Conditions ....................................................................................... 25

2.3 Results ............................................................................................................................... 25

2.3.1 The delayed floral transition of dnd1 is SA-independent ..................................... 25

2.3.2 The delayed flowering phenotype of cpr22 is SA-dependent .............................. 26

2.3.3 Additional CNGC knockout mutants do not have delayed floral transition ......... 27

2.4 Discussion ......................................................................................................................... 30

2.5 Acknowledgements ........................................................................................................... 30

Chapter 3 Identifying the causative mutation for repressor of defense no death 1 (rdd1-

1D), a novel suppressor of dnd1 .............................................................................................. 31

Abstract .................................................................................................................................... 32

3.1 Introduction ....................................................................................................................... 33

3.2 Materials and Methods ...................................................................................................... 35

3.2.1 Plant Growth Conditions ....................................................................................... 35

3.2.2 Double Mutant Analysis ....................................................................................... 35

3.2.3 Trypan Blue Staining ............................................................................................ 36

3.2.4 Measurement of endogenous salicylic acid .......................................................... 36

3.2.5 Plasmid Construction and Agrobacterium mediated transformation for complementation analysis ..................................................................................... 37

3.2.6 Pathogen infection assay ....................................................................................... 38

3.3 Results ............................................................................................................................... 39

3.3.1 Location of rdd1-1D mutation in YUCCA6 .......................................................... 39

3.3.2 yucca6-3k is able to suppress dnd1-conferred dwarf morphology ....................... 42

3.3.3 YUCCA6 over expressing line does not suppress dnd1-conferred dwarf morphology ........................................................................................................... 46

3.3.4 yucca6-3k is able to suppress dnd1-conferred spontaneous cell death

formation ............................................................................................................... 50

3.3.5 yucca6-3k is able to suppress dnd1-conferred delayed flowering phenotype ....... 50

3.3.6 yucca6-3k is able to suppress dnd1-conferred SA accumulation .......................... 53

vi

3.3.7 rdd1-1D is unable to suppress cpr22-conferred phenotypes ................................ 56

3.3.8 Complementation analysis .................................................................................... 59

3.3.9 Pathogen analysis .................................................................................................. 64

3.4 Discussion ......................................................................................................................... 64

Chapter 4 Investigating a Role for Cyclic Nucleotide Gated Ion Channel 2 in Auxin

Homeostasis ............................................................................................................................. 67

Abstract .................................................................................................................................... 68

4.1 Introduction ....................................................................................................................... 69

4.2 Materials and Methods ...................................................................................................... 72

4.2.1 Plant Growth Conditions ....................................................................................... 72

4.2.2 DR5::Green Fluorescence Protein (GFP) Visualization by Confocal Microscopy ........................................................................................................... 73

4.2.3 DR5::Glucuronidase (GUS) Visualization ........................................................... 73

4.3 Results ............................................................................................................................... 74

4.3.1 rdd1-1D and yucca6-3k single mutant seedlings have similar response to auxin treatment in roots ........................................................................................ 74

4.3.2 rdd1-1D cngc2 and yucca6-3k cngc2 double mutant seedlings have similar response to auxin treatment in roots ..................................................................... 75

4.3.3 dnd1 seedlings do not display clear alterations in auxin sensitivity in roots or

vein pattering development ................................................................................... 75

4.3.4 No trend in root length observed between Col-wt, cngc2 and yucca6 mutants .... 76

4.3.5 Analysis of endogenous auxin levels using DR5::GFP and DR5::GUS systems . 84

4.4 Discussion ......................................................................................................................... 90

Chapter 5 General Discussion .................................................................................................. 94

5.1 Discussion ......................................................................................................................... 95

5.2 Future Directions ............................................................................................................ 103

References ................................................................................................................................... 107

Appendices .................................................................................................................................. 119

vii

viii

List of Publications

Fortuna A, Lee J, Ung H, Chin K, Moeder W, Yoshioka K. 2015. Crossroads of stress

responses, development and flowering regulation- the multiple roles of Cyclic

Nucleotide Gated Ion Channel 2. Plant Signal. Behav. 10: e989758.

ix

List of Tables

Table 3-1: Segregation analysis of yucca6-3k x cngc2

Table 3-2: Segregation analysis of yucca6-1D x cngc2

Table 3-3: Summary of complementation constructs

Table 3-4: Current status of complementation analysis

x

List of Figures

Figure 1-1: Structure of plant Cyclic Nucleotide Gated Ion Channels (CNGCs).

Figure 1-2: Phylogenetic tree of CNGCs in Arabidopsis

Figure 2-1: Flowering phenotypes in various CNGC related mutants

Figure 3-1: Analysis of rdd1-1D mutation in YUCCA6

Figure 3-2: Morphology and cell death analysis from yucca6-3k segregation analysis

Figure 3-3: Morphology of Col-wt and mutant plants from yucca6-1D segregation

analysis

Figure 3-4: Floral transition in Col-wt and mutant plants

Figure 3-5: Effect of yucca6-3k on salicylic acid accumulation in cngc2

Figure 3-6: Cell death and leaf morphology analysis of rdd1-1D and cpr22 related

mutants

Figure 3-7: Schematic diagram of complementation analysis

Figure 4-1: Auxin Sensitivity in Col-wt and mutant seedlings

Figure 4-2: Vein pattern development in Col-wt and CNGC mutants

Figure 4-3: DR5rev::GFP signal in Col-wt and mutant seedlings

Figure 4-4: DR5::GUS signal in Col-wt and mutant seedlings

xi

Figure 5-1: Proposed model for CNGC2-mediated signal transduction.

xii

List of Appendices

Table A2-1: Summary of mutants utilized

Table A3-1: Chapter 3 Primer Sequences

Table A4-1: Chapter 4 Primer Sequences

xiii

List of Abbreviations

χ2 Chi square

AtCNGC Arabidopsis thaliana cyclic nucleotide-gated channel

Ca2+ Calcium ion

CaM Calmodulin

CaMV35S Cauliflower mosaic virus CamV35S

Cl- Chloride ion

Col-wt Columbia wild type

cpr22 constitutive expresser of pathogenesis related genes 22

DAB 3,3′-Diaminobenzidine

dnd1 denfense, no death 1

dnd2 defense, no death 2

E. coli Escherichia coli

ETI Effector triggered immunity

F1 Filial generation 1

F2 Filial generation 2

xiv

GFP Green fluorescent protein

GLR Glutamate receptor-like

GUS Glucuronidase

hml1 HR-like lesion mimic 1

Hpa Hyalopernospora Arabidopsidis

HR Hypersensitive response

IAA Indole-3-acetic acid

IPA indole-3-pyruvate

L Leucine

PAMP Pathogen-associated molecular pattern

PR gene Pathogenesis related gene

PRR Pattern recognition receptors

P Proline

PTI Pattern (or PAMP) triggered immunity

PCD Programmed cell death

Psm Pseudomonas syringae pv. maculicola

Pst Pseudomonas syringae pv. tomato

xv

rdd1 represser of denfense no death 1

R gene Resistance gene

R protein Resistance protein

ROS Reactive oxygen species

SA Salicylic acid

SAUR Small auxin-up RNA

T-DNA Transfer DNA

UTR Untranslated region

Wt Wild type

qRT-PCR Quantitative real time polymerase chain reaction

1

Chapter 1

Introduction

2

1.1 Plant Immunity Signaling

As sessile organisms, which lack the mobility and adaptive immune system of

vertebrates, plants have evolved a complex signaling network which depends on

cell-autonomous events for survival. To defend against pathogen infection plants

have evolved an immune response encompassing two layers. The first layer of

defense is considered a basal immune response and is activated when conserved

pathogen-associated molecular patterns (PAMPs), are recognized by plasma

membrane localized receptors called pattern recognition receptors (PRRs). This

recognition leads to pattern-triggered immunity (PTI). The PTI response includes

changes in gene expression, rapid ion flux, the accumulation of hormones involved

in immunity signaling (Dodds and Rathjen 2010; Spoel and Dong 2012).

In an attempt to bypass PTI, several pathogens have evolved the ability to

deliver effector molecules into the host cell. Unlike PAMPs, which tend to be

conserved across pathogens, effectors are generally variable and dispensable

(Spoel and Dong 2012; J. Zhou et al. 2014). However, plants have co-evolved with

these pathogens, and as such have developed intracellular immune receptors, also

known as resistance (R) proteins, which can trigger a robust form of immunity called

effector-triggered immunity (ETI) upon effector recognition (Dodds and Rathjen

2010; Spoel and Dong 2012). ETI induction is generally stronger and faster than

PTI, and triggers a form of localized programmed cell death (PCD) called the

hypersensitive response (HR) through changes in gene expression and the

3

accumulation of reactive oxygen species (ROS) (Jones and Dangl 2006; Dodds and

Rathjen 2010; van Schie and Takken 2014).

1.2 Importance of ion flux in immune signaling

Ion flux across the plasma membrane have long been recognized as one of

the earliest responses to pathogen recognition (Hahlbrock et al. 1995; Gelli et al.

1997; Blume et al. 2000). An example of this is the rapid influx of Ca2+ and H+ and

effluxes of K+ and Cl- in parsley cells after treatment with the Phytophthora elicitor,

Pep-1 (Hahlbrock et al. 1995; Gelli et al. 1997). Additionally, electrophysiological

analyses have reported Ca2+ influxes at the plasma membrane of tomato

suspension cells in response to fungal effectors (Gelli et al. 1997). Many of the

downstream signaling pathways discussed previously have been shown to be

dependent on free Ca2+ available in the media, including the accumulation of ROS,

defense gene activation, and phytoalexin production (Scheel 1998). This has been

demonstrated through pharmacological analysis where the use of calcium chelators

has been shown to block downstream defense gene activation as well as the

accumulation of ROS (Gelli et al. 1997). Overall, Ca2+ is an important secondary

messenger relaying a variety of cues related to development, as well as biotic and

abiotic stress. It is hypothesized that a PAMP, discussed previously, when

recognized by its receptor results in the activation of plasma membrane localized ion

channels resulting in changes in cellular Ca2+ concentration. This ion flux can then

activate additional ion channels and pumps further stimulating a downstream

signaling cascade (Scheel 1998; Lee et al. 2014). Although the Ca2+ channels that

4

regulate this process are not completely clear, members of two large ion channel

families, the glutamate receptor-like (GLR) family and cyclic nucleotide-gated ion

channel (CNGC) family, have been suggested to be involved in this process

(Davenport 2002; Talke et al. 2003).

1.3 Cyclic-Nucleotide Gated (CNGCs)

CNGCs are nonselective cation channels that were first identified in animals

as key components of the vertebrate visual and olfactory systems (Zugotta and

Siegelbaum 1996). Mutations in mammalian CNGCs may result in losses to visual

and olfactory perception (Yau and Baylor 1989; Zugotta and Siegelbaum 1996). In

mammals six CNGC genes have been identified, encoding four α and two β

subunits. The typical mammalian CNGC channel is believed to be a heterotetramer,

with the specific subunit composition determining channel function. Mammalian

CNGCs share sequence homology to voltage-gated ion channels, and have been

shown to be ligand-gated through the binding of cyclic nucleotides, cAMP and cGMP

(Yau and Baylor 1989; Channels et al. 1994). cAMP and cGMP are important

secondary messengers involved in a wide variety of development processes and

have the capacity to bind to two protein domains, the GAF domains (cyclic GMP,

adenylyl cyclase, FhlA) and cyclic nucleotide binding domains (CNBDs) (Bridges et

al. 2005). CNBDs are found primarily in two groups in Arabidopsis, CNGCs and

Shaker-type K+ channels (Jammes et al. 2011).

In contrast to mammalian CNGCs, the Arabidopsis CNGC family has greatly

expanded to 20 members. This is a large portion of the predicted 56 coding

5

sequences identified as cation- conducting channels in Arabidopsis genome (Ward

et al. 2009; Zelman et al. 2012). As illustrated in Figure 1-1 CNGCs have six

transmembrane domains with a pore between the fifth and sixth domain.

Additionally, CNGCs have a calmodulin binding domain (CaMBD) which overlaps

with a cyclic nucleotide binding domain (CNBD). Genomic sequence alignment of

the 20 members categorizes them into four groups (group I to IV), with the fourth

group separated into subgroups (IVa and IVb) (Figure 1-2: Mäser et al. 2001).

Previous research suggests important roles for CNGCs in various environmental

stress responses, defense, development, and thermotolerance (Chan et al. 2008;

Chaiwongsar et al. 2009; Finka et al. 2012). Extensive work has been conducted

with several CNGCs to better understand their role in immunity signaling, specifically

AtCNGC2, AtCNGC4 (Group IVB), AtCNGC11, and AtCNGC12 (Group I) (Clough et

al. 2000; Jurkowski et al. 2004; Moeder et al. 2011; Chin et al. 2013).

Evidence for CNGC involvement in plant immunity has primarily been

gathered using genetic analyses. Null mutants of CNGC2 and CNGC4, known as

defense, no death 1 (dnd1) and HR-like lesion mimic 1/defense no death 2

(hlm1/dnd2), respectively, have distinct defense related phenotypes (Clough et al.

2000; Jurkowski et al. 2004). The dnd mutants were isolated in a screen for reduced

HR upon infection with the avirulent bacterial pathogen Pseudomonas syringae pv.

glycinea (Psg). Additionally, the dnd mutants exhibit autoimmune phenotypes,

including constitutive expression of Pathogenesis related (PR) genes, elevated

accumulation of salicylic acid (SA), and enhanced resistance to a broad spectrum of

pathogens. Further, dnd mutants exhibit a dwarf morphology, Ca2+ hypersensitivity,

6

and a delayed floral transition, a novel phenotype recently discovered in our lab.

(Clough et al. 2000; Jurkowski et al. 2004; Chan et al. 2008; Chin et al. 2013). As

CNGC2 and 4 null mutations induce autoimmunity, one can hypothesize that they

are negative regulators of defense. However, since they have reduced HR upon

pathogen infection, some researchers consider them as positive regulators. Thus

there is on-going debate about their role in defense (Moeder et al. 2011).

Another CNGC mutant that shows alterations in defense response is

constitutive expresser of PR genes 22 (cpr22), which results from the expression of

the chimeric CNGC11/12 channel (Yoshioka et al. 2006). cpr22 plants exhibit a

stunted morphology with curly leaves, HR-like spontaneous lesion formation in

leaves, elevated endogenous levels of SA, and enhanced disease resistance

(Yoshioka et al. 2001). Additionally, T-DNA knockout lines for CNGC11 and

CNGC12 exhibit a partial breakdown in pathogen resistance, together suggesting

that CNGC11 and CNGC12 are positive regulators of pathogen resistance

(Yoshioka et al. 2006).

7

8

Figure 1-1: Phylogenetic tree of Arabidopsis CNGCs. There are 20 CNGC genes in

Arabidopsis which have been subdivided into 4 groups, with group 4 being further

divided into A and B. CNGC11 and CNGC12 from group I and CNGC2 and CNGC4

from group IV B have been implicated in immune signaling. Copyright 2001 by

Mäser et al. Reprinted with permission.

9

10

Figure 1-2: Predicted structure of plant Cyclic Nucleotide Gated Ion Channels

(CNGCs). CNGCs are characterized by having six transmembrane domains (S) with

a pore between the fifth and the sixth transmembrane domains (P). CNGCs have a

calmodulin binding domain (CaMBD) which overlaps with a cyclic nucleotide binding

domain (CNBD). Copyright 2009 by Chin et al. 2009. Reprinted with permission.

11

1.4 Identification of repress of defense, no death 1

To identify novel components of CNGC2-mediated signaling, a suppressor

screen of a T-DNA knockout line for CNGC2 (cngc2-3, hereafter referred to as

cngc2) was undertaken, identifying the first suppressor of dnd1, repressor of

defense, no death 1 (rdd1, as the mutation is dominant, we call this mutant rdd1-1D

cngc2). For this screen approximately 10,000 cngc2 T-DNA mutant seeds were

mutagenized with EMS, M2 seeds were collected and screened for mutants

suppressing the dwarf phenotype conferred by the cngc2 mutation. rdd1-1D partially

suppresses dnd1 (cngc2) mediated stunted growth and enhanced resistance, and

completely suppresses delayed flowering (Chin et al. 2013). Furthermore, rdd1-1D

suppresses hlm1/dnd2 (cngc4)-mediated phenotypes, indicating that CNGC2 and

CNGC4 likely have converging or overlapping signaling pathways. rdd1-1D was

determined to be a dominant mutation as all backcrossed first generation plants (F1

progenies of rdd1-1D cngc2 x cngc2) exhibited rdd1-1D cngc2 morphology and

subsequent self-pollination resulted in 3:1 segregation (Chin et al. 2013).

A combination of conventional map-based cloning and whole genome

sequencing identified 4 potential causative mutations for rdd1-1D at the loci

At5G24680, At5G25590, At5G25620, and At5G26050. In At5G25620 a non-

synonymous amino acid substitution, proline to leucine, was identified at residue 289

within the coding sequence. At5G25620 encodes a flavin-containing

monooxygenase-like protein YUCCA6. YUCCAs are an 11 member family in

Arabidopsis which have been reported to be involved in auxin biosynthesis (Zhao et

12

al. 2001). Auxin is an important plant hormone with a range of roles in growth and

development, including cell division, elongation, differentiation, flowering, fruit

ripening and gravitropism (Zhao 2014). Research to date, which will be detailed in

this thesis, suggests the point mutation in YUCCA6 is the causative mutation of

rdd1-1D cngc2. This provides a new perspective about the biological function of

CNGC2 and likely CNGC4 in auxin signaling.

13

1.5 Flavin-containing monooxygenases (FMOs)

Flavin containing monooxygenases (FMOs) are a type of flavoprotein found in

both eukaryotes and prokaryotes, having first been discovered in the 1960s in liver

extracts as novel enzymes able to catalyze the N-oxidation of N, N-dimethylaniline

(Ziegler 1990). FMOs have been extensively studied in humans where they have

been shown to play a role in xenobiotic detoxification in the liver. FMOs require

NADPH and oxygen to catalyze the oxygenation of their substrate. In mammals 5

FMO genes have been identified (FMO1-FMO5), and FMO1 has been shown to

have a wide substrate specificity (Ziegler 1990). Known substrates for FMOs are soft

nucleophiles, and usually contain sulfur or nitrogen (Krueger and Williams 2005).

However, many nucleophiles are essential for proper cell function, such as

glutathione, and must be excluded from FMO oxidation, though the mechanism for

exclusion is not known. Further, those substrates which are oxidized will either result

in a detoxification or bioactivation and this depends on the properties of the

substrate (Cashman 2002). While the primary role of mammalian FMOs appears to

be detoxification, relatively little is known about the role of FMOs in plants.

1.6 Flavin containing monooxygenases in Arabidopsis

The Arabidopsis FMO family is separated into 3 clades, with FMO1 and a

pseudogene in clade 1, 11 YUCCA genes in clade II, and 16 additional FMOs in

clade III (See Chapter 3; Schlaich 2007). The first FMO-like enzymes were identified

in Arabidopsis through activating-tagging in search of auxin biosynthesis enzymes

14

(Zhao et al. 2001). Forward genetic screens for auxin deficient mutants have been

mostly unsuccessful, likely due to redundancy and complexity of the auxin

biosynthetic pathway (Wu et al. 2015). Through activation tagging Zhao et al. (2001)

identified two independent yucca mutants which had characteristic auxin

overproduction phenotypes, including long hypocotyls, epinastic cotyledons,

elongated petioles and enhanced apical dominance. Through measuring

endogenous auxin levels it was discovered that yucca mutants contained >50 %

more free IAA than Col-wt. Indole-3-acetic acid (IAA) is the major natural auxin in

plants, and YUCCAs are believed to catalyze the rate limiting step in its biosynthesis

(Mashiguchi et al. 2013). Further, to investigate whether the auxin over-

accumulating phenotypes observed were in fact caused by IAA accumulation the

authors crossed yucca with iaaL, an auxin conjugating enzyme, and observed a loss

of the previously observed phenotypes (Zhao et al. 2001).

1.7 FMOs in auxin biosynthesis

YUCCA family enzymes are hypothesized to be responsible for the rate

limiting step in the indole-3-pyruvic acid (IPA) intermediate biosynthesis pathway,

one of four proposed pathways, and the only characterized at the genetic and

biochemical level to date (Mashiguchi et al. 2011; Dai et al. 2013; Wu et al. 2015).

IAA biosynthesis from tryptophan is a two-step process. First the TRYPTOPHAN

AMINOTRANSFERASE OF ARABIDOPSIS (TAA) family of transaminases convert

tryptophan to indole-3-pyruvate (IPA). Second, oxidative decarboxylation of IPA

through the activity of YUCCA family enzymes produces IAA (Mashiguchi et al.

15

2011). This biosynthetic process is highly conserved and has been identified in

several species throughout the plant kingdom, both in monocots and dicots (Cheng

et al. 2006; Zhao 2014). The other three proposed pathways are the indole-3-

acetamide (IAM), tryptamine (TAM) and indole-3-acetaldoxime (IAOx) pathways

(Zhao 2014).

1.8 Importance of localized auxin production and YUCCA

gene expression

The tightly controlled spatiotemporal expression of YUCCA genes is necessary

for proper organ initiation and tissue patterning (Cheng et al. 2006). It has long been

hypothesized that auxin is primarily synthesized in young, fast growing areas such

as leaves, flowers and root tips where it is then transported to different regions as

required. However, mounting evidence has suggested a role for local auxin

biosynthesis (Zhao et al. 2008). The YUCCA gene family in Arabidopsis contains 11

members, which share 44 to 64 % amino acid sequence homology (Zhao et al.

2001). The YUCCA gene family is believed to have overlapping function, as no

single knockout has been reported to have an obvious developmental phenotype,

and only double knockouts of closely related YUCCA genes, such as yuc1 yucca4 or

yucca2 yucca6 have observable developmental abnormalities. These developmental

abnormalities may include defects in floral development and vein patterning (Cheng

et al. 2006). Additionally, though not linked to a developmental abnormality, a T-

DNA knockout line of YUCCA6 does display a broader leaf compared to Col-wt (Kim

et al. 2007). The developmental defects of yucca1 yucca4 or yucca2 yucca6 can be

16

rescued by the transformation of iaaM, a bacterial gene which produces auxin,

driven by the YUCCA6 promoter. However, the mutant phenotypes were not

rescued by exogenous auxin supplementation. This indicates that the developmental

defects observed were in fact due to aberrant auxin levels, and that specific

expression of YUCCA genes is necessary for rescuing the phenotype (Cheng et al.

2006).

Although auxin has been linked to plant-pathogen defense, as will be

discussed, the role of the YUCCA gene family in plant-pathogen defense is

uncharacterized. One Arabidopsis FMO, FMO1 has been implicated in pathogen

defense, though the mechanism is uncharacterized and has not been connected to

auxin signaling. FMO1 has been extensively studied as a marker for plant cell death

and as a component of systemic acquired resistance (SAR) (Mishina and Zeier

2006; Olszak et al. 2006).

Several members of the YUCCA gene family have been shown to be

negatively regulated at the transcript level by the presence of free auxin (Suzuki et

al. 2015). Similarly, the expression of YUCCA enzymes is induced upon application

of the auxin biosynthetic inhibitor kynurenine (Suzuki et al. 2015). Suzuki et al.

(2015) investigated YUCCA gene expression levels in the auxin deficient mutant

wei8-1 tar2-1, and it was observed that YUCCA1, YUCCA2, and YUCCA4

transcripts were upregulated while YUCCA6 levels remained unchanged.

Additionally, in YUCCA1 over-expression lines, which had 2 times greater IAA than

wild-type, TAA1, TAR2, YUCCA2, YUCCA4, and YUCCA6 transcript levels were all

lower than wild type. Taken together, these results indicate that the YUCCA genes

17

necessary for growth and development share a negative feedback regulation

mechanism that does not always overlap (Suzuki et al. 2015). As detailed in this

thesis, rdd1-1D is likely a mutant allele of the auxin biosynthesis gene YUCCA6.

1.9 Auxin and Plant Immunity

In addition to its important role in growth and development, the plant hormone

auxin has also been studied as a regulator of plant defense, though the molecular

mechanisms connecting auxin to defense are largely unknown. As a growth

promoting hormone, auxin has been hypothesized to antagonize defense activation,

and several studies have investigated the relationship between auxin and the

important defense hormone SA (Cui et al. 2013; Mutka et al. 2013; Wang et al.

2007). The evolutionary basis for such antagonism may be that defense and

development are both energy consuming processes, and plants, with limited

resources, must find a balance. Current research suggests auxin affects defense

activation in both an SA-dependent and independent manner (Kazan and Manners

2009; Mutka et al. 2013).

Several auxin responsive genes, which are important for growth and

development, may play an additional role in immune signaling, including Aux/IAA,

GH3 and small auxin-up RNA (SAUR) gene families (Woodward and Bartel 2005;

Robert-Seilaniantz et al. 2007; Kazan and Manners 2009; Mutka et al. 2013). The

GH3 family is responsible for auxin conjugation and several studies have linked the

18

overexpression of certain GH3 genes to enhanced resistance, including GH3.8 in

rice and GH3.5 in Arabidopsis (Zhang et al. 2007; Bari and Jones 2009).

Several biotrophic and hemibiotrophic pathogens which manipulate auxin

signaling to increase their fitness/ pathogenicity have been identified (Cui et al.

2013; Mutka et al. 2013). Additionally, the exogenous application of auxin to plants

has been shown to promote susceptibility in several instances, such as the co-

inoculation of auxin with Pseudomonas syringae pv. maculicola (Psm), which results

in increased pathogen growth compared to Psm alone (Wang et al. 2007).

Additionally, the bacterial type III effector avrRpt2, a cysteine protease, has been

shown to alter the auxin physiology of its host to promote pathogen growth (Chen et

al. 2007). Further, global genome-wide expression analysis after virulent

Pseudomonas syringae pv. tomato DC3000 (Pst) infection by Thilmony et al. (2006)

identified that auxin biosynthetic genes appeared to be upregulated and the

expression of several Aux/IAA genes, responsible for repression of auxin inducible

genes, were repressed. All of these instances discussed suggest that auxin may

promote disease susceptibility and subsequent repression of auxin signaling

potentially results in enhanced resistance in plants (Bari and Jones 2009).

1.10 Auxin and Ca2+ signaling

Ca2+ is an important secondary messenger in a wide variety of physiological

processes, and one such role may be an involvement in auxin signal transduction

(Vanneste and Friml 2013; Di et al. 2015). Ca2+ has been shown to induce auxin

related gene expression in wheat, and this expression is reversed by the Ca2+

19

chelator EGTA (Singla et al. 2006). Additionally, auxin has been shown to induce the

expression of calmodulin (CaM), a Ca2+ binding protein, and increase intracellular

Ca2+ levels (Gehring et al. 1990). The source of the Ca2+ signal remains unknown,

but it is possible that the apoplast, vacuole, endoplasmic reticulum or other small

organelle are responsible for auxin-induced Ca2+ release (Vanneste and Friml

2013).

1.11 Thesis Aim and Overview

The main objectives of my thesis was to better understand the molecular

mechanisms underlying the pleiotropic phenotypes of CNGC2, identify the causative

mutation of the novel dnd1 suppressor rdd1-1D, and elucidate the molecular

mechanism of suppression. As discussed, CNGC genes are a 20 member family in

Arabidopsis involved in a diverse array of physiology processes. Specifically,

CNGC2 is involved growth and development, ion homeostasis, thermotolerance,

and floral transition. For my first objective I aim to understand SA dependency of the

delayed flowering phenotype in the CNGC2 null mutant dnd1, recently discovered in

our lab (Chapter 2). Several dnd1 phenotypes have been shown to be SA-

dependent. For this analysis I conducted epistatic analysis with SA biosynthesis and

signaling mutants. Here, I present data that the delayed flowering phenotype of dnd1

is SA-independent, and the role of CNGC2 (and likely CNGC4) in floral transition is

likely unique among the CNGC family.

For my second objective I aimed to identify the causative mutation of rdd1-1D

(Chapter 3). Previous work by Dr. Kimberley Chin identified four potential causative

20

mutations for rdd1-1D in At5G24680, At5G25590, At5G25620, and At5G26050,

none of which are known components of CNGC2-mediated signaling. Using double

mutant analysis, work presented in Chapter 3 indicates that rdd1-1D is likely a loss-

of-function mutation in At5G25620, an auxin biosynthesis gene. Building from this

observation we explore a possible connection between CNGC2 and auxin in

development (Chapter 4).

21

Chapter 2

Cyclic Nucleotide Gated Ion Channel 2: a new role in floral

transition

Modified from: Crossroads of stress responses, development and flowering

regulation- the many roles of Cyclic Nucleotide Gated Ion Channel 2

A. Fortuna, J. Lee, H. Ung, K. Chin, W. Moeder and K. Yoshioka (2015)

Plant Signaling and Behavior 10 (3)

Reprinted with permission of The Society of Plant Signaling and Behavior

(http://plantbehavior.org)

22

Abstract

Cyclic nucleotide-gated ion channels (CNGCs) are non-selective cation

channels that form a 20 member family in Arabidopsis thaliana. CNGCs have been

implemented in a wide variety of physiological processes including growth and

development, and responses to environmental stresses and pathogens. It has been

suggested that CNGC2 plays a role in the defense response and other physiological

processes through its role as a Ca2+ conducting channel. The null mutant of CNGC2,

“defense, no death” (dnd1), exhibits smaller stature and an autoimmune phenotype,

including constitutive expression of pathogenesis-related (PR) genes and elevated

levels of salicylic acid (SA). In addition, the recently novel phenotype of dnd1 in

flowering transition has been reported. It exhibits significantly late floral transition,

indicating the involvement of CNGC2 in regulation of flowering timing. SA, an

important signaling molecule for pathogen defense responses, is also known to be

involved in flowering transition regulation, as reported that elevated levels of SA

promote early flowering. However, dnd1, despite its high accumulation of SA,

displays a late flowering phenotype. In this work, we have investigated 1) whether

the late flowering phenotype in dnd1 is SA dependent, and 2) if other CNGCs are

also involved in flowering transition regulation. Through double mutant analysis

using SA biosynthesis and signaling mutants, it was discovered that the dnd1 late

flowering phenotype is SA-independent. Furthermore, no other CNGC null mutants

that have been analyzed so far exhibits a late flowering phenotype like dnd1. This

data indicates a unique role for CNGC2 in SA-independent floral transition.

23

2.1 Introduction

Cyclic nucleotide-gated ion channels (CNGCs) are non-selective cation

channels that were first identified in animals, where they play key roles in light and

olfactory signaling. In mammals, there are six genes that encode CNGCs and the

typical mammalian CNGC consists of four CNGC subunits. The predicted structures

of plant CNGCs are similar to their animal counterparts; however, in plants an

expansion of the CNGC family occurred. The Arabidopsis thaliana genome has 20

members in the CNGC family. This expansion may indicate diverse biological roles

of CNGCs in plants. In Arabidopsis, 20 CNGCs are classified into four groups (group

I-IV), where group IV is further divided into subgroup IVA and IVB (Chin et al. 2009).

They have been implicated in a diverse range of biological phenomena such as

defense responses, pollen tube growth, ion homeostasis and thermo-tolerance (Chin

et al. 2009; Finka et al. 2012). In addition, recent electrophysiological studies

showed that plant CNGCs are likely Ca2+ permeable channels that are involved in a

variety of physiological phenomena (Finka et al. 2012; L. Zhou et al. 2014).

Group IVB comprises only two members, CNGC2 and CNGC4. They are the

most divergent members of the CNGC family and both are reported to be involved in

pathogen defense responses as loss-of-function mutants of CNGC2 or CNGC4

show remarkably similar autoimmune phenotypes. The null mutant of CNGC2,

“defense, no death” (dnd1), has been extensively characterized and is known as a

rare autoimmune mutant with impaired hypersensitive responses (HR). The HR is a

characteristic defense response which is a type of programmed cell death around

24

the sites of pathogen entry. Despite the impairment of HR upon pathogen infection,

the dnd1 mutant displays constitutive defense responses, such as elevated

expression of Pathogenesis-Related (PR) genes, high levels of salicylic acid (SA) -

an important signaling molecule for resistance against biotrophic pathogens, and

conditional HR-like spontaneous lesions without pathogen infection. Consequently,

dnd1 plants show enhanced broad spectrum resistance against several

taxonomically unrelated pathogens. In addition, it exhibits characteristic

morphological phenotypes, such as small stature and senescence-like chlorosis at

the tips of the leaves, indicating roles of CNGC2 in both defense and development.

Recently, we discovered a novel phenotype in dnd1, which is delayed

flowering transition and this phenotype was observed under both long and short day

conditions, although enhanced in the latter condition (Chin et al., 2013, Figure 2-1A).

Flowering transition is tightly regulated by endogenous and external cues. In

addition, it is known that various stresses, such as ultraviolet-C radiation, pathogen

infection and extreme temperatures can promote flowering. Interestingly, it has been

reported that SA positively regulates flowering timing in Arabidopsis (Martínez et al.

2004). SA-deficient mutants, such as nahG, sid2 and eds5/sid1, exhibit late

flowering phenotypes, while SA hyper-accumulating mutants, such as acd6 show

early flowering transition, supporting this notion (Martínez et al. 2004; Wang et al.

2011). However, contrary to the positive role of SA, HOPW1-INTERACTING3

(WIN3), which positively regulates broad-spectrum disease resistance through SA

signaling, suppresses flowering transition (Wang et al. 2011). Thus, the relationship

of SA, defense activation and flowering timing is complex. This raises several

25

questions regarding the delayed flowering phenotype in dnd1 in spite of the high

levels of SA accumulation: 1) does SA play a role in the delayed flowering transition

phenotype in dnd1?, 2) is the flowering transition phenotype in dnd1 a by-product of

hyperactivation of defense signaling?, and 3) are other CNGCs also involved in the

regulation of flowering?

2.2 Materials and Methods

2.2.1 Plant Growth Conditions

Arabidopsis thaliana seeds were sown on Sunshine Mix #1 (Sun Gro

Horticulture Canada Ltd) and stratified at 4oC for 4 days. Plants were grown in a

growth chamber with a 16 hour photoperiod (16 hour light/ 8 hour dark) at 22oC.

Floral transition was scored every second day from first bolt. Summary of all mutants

utilized in this chapter is provided in Table A2-1.

2.3 Results

2.3.1 The delayed floral transition of dnd1 is SA-independent

To address the above mentioned first and second questions, we monitored

the timing of flowering transition in double mutants of dnd1 with SA-deficient and

defense signaling mutants. SID2 (ICS1) is a major SA biosynthesis gene for defense

responses; thus, dnd1 sid2 exhibits reduced levels of SA compared to the dnd1

single mutant (Genger et al. 2008). NPR1 is a major component of SA signaling and

npr1 mutants show a deficiency in SA-induced defense responses. It has been

26

reported that dnd1 npr1 exhibits similar susceptibility to wild type plants against

pathogens; thus, the enhanced pathogen resistance of dnd1 is NPR1-dependent

(Genger et al. 2008). As shown in Figure 2-1A and B, both double mutants, dnd1

sid2 and dnd1 npr1, exhibited no significant difference in flowering transition from

the dnd1 single mutant, indicating that the delayed flowering transition phenotype in

dnd1 is independent from SA accumulation or NPR1-mediated defense activation.

2.3.2 The delayed flowering phenotype of cpr22 is SA-dependent

To test whether other CNGC mutants that are related to SA accumulation and

defense response activation also show similar delayed flowering transition

phenotypes, we monitored constitutive expresser of PR genes22 (cpr22). cpr22

displays autoimmune phenotypes with increased SA accumulation and constitutive

PR gene expression, similar to dnd1 (Yoshioka et al. 2006). It is a gain-of-function

mutant and its phenotype is due to the expression of the chimeric CNGC11/12 gene

(Yoshioka et al. 2006). As show in Figure 2-1C and D, cpr22 does not show delayed

flowering transition. Rather we observed a consistent early flowering phenotype in

cpr22 compared to its wild type Wassilewskija (Ws) ecotype. This indicates that

elevated SA levels in cpr22 promote flowering transition, as expected by the positive

role of SA in flowering transition. To further address this question, we monitored

flowering transition in the double mutant of cpr22 and sid2. Since cpr22 has a Ws

background and sid2 has a Columbia ecotype background, we used mixed

background lines from a cpr22 x sid2 cross for this analysis. As expected, cpr22

SID2 showed earlier flowering transition than CPR22 SID2 wild type by a few days

27

(Figure 2-1C and D). Also, CPR22 sid2 showed delayed flowering, as expected.

Interestingly, the double mutant cpr22 sid2 showed almost the same flowering

transition as CPR22 sid2, indicating that the earlier flowering phenotype in cpr22 is

due to its SA accumulation. This agrees well with the reported positive role of SA in

flowering transition, unlike what we observe in dnd1.

2.3.3 Additional CNGC knockout mutants do not have delayed

floral transition

In addition to the altered floral transition phenotypes discussed, it is possible

that some other CNGC members share a common role in flowering transition and

the loss-of-function of any CNGCs (loss of their channel function) might cause a

similar late flowering phenotype that is not related to SA. To address this point, we

have monitored various CNGC loss-of-function mutants including cngc11 and

cngc12. However, as shown in Figure 2-1D, knockout mutants for CNGC3, 11, 12,

19 and 20 did not exhibit any significant delay in flowering transition, suggesting that

it is not a common feature in CNGC knockout mutants. Recently, we showed that

null mutants of CNGC4 also have delayed flowering phenotypes, like dnd1, and that

CNGC2 and 4 likely form a channel complex together (Chin et al. 2013). In other

words, these data suggested that the two group IVB CNGCs have a unique role in

flowering transition.

28

29

Figure 2-1. Flowering phenotypes in various CNGC-related mutants. A) The

delayed flowering phenotype in dnd1 is SA/NPR1 independent. rdd1-1D cngc2, the

suppressor of dnd1, served as a control. n = 12 - 25. dnd1 and cngc2 are loss-of-

function mutants in CNGC2. SID2 is an auxin biosynthesis gene and sid2 plants are

SA-deficient. NPR1 is an SA signaling molecule and npr1 plants are deficient in SA-

induced defense responses. B) Flowering phenotype of about 5 week old Col-wt and

mutant plants. C) The early flowering phenotype of cpr22 is SA-dependent. cpr22

has a Ws background and sid2 has Col ecotype background. n = 10 - 35. D)

Flowering phenotype of various CNGC T-DNA Insertion lines, n = 21 - 31. Flowering

time was measured as described in Chin et al. (2013). Error bars = SE, Bars marked

with the same letter indicate no significant difference (Student’s t-test, p < 0.05).

Plants were grown in Sunshine Mix #1 with a photoperiod of 16 h light and 8 h dark.

All experiments have been repeated at least three times with similar results.

30

2.4 Discussion

Although they share similar autoimmune phenotypes, cpr22 (CNGC11/12) is

a gain-of-function and dnd1 (cngc2) is a loss-of-function mutant of CNGCs. In

addition, the loss-of-function mutants for CNGC11 and 12 (cngc11 and cngc12)

show a partial breakdown of pathogen resistance (Yoshioka et al. 2006). These data

indicate a striking difference in the molecular mechanisms that govern defense

signaling mediated by CNGC11 and 12 from that of CNGC2 (Yoshioka et al. 2006;

Moeder et al. 2011). Thus, the flowering phenotype difference between cpr22 and

dnd1 is not surprising.

Through extensive analyses of the dnd1 mutant, CNGC2 is the best-

characterized CNGC member and it has been suggested that CNGC2 transduces

the Ca2+ signal after pathogen infection upon recognition of Pathogen Associated

Molecular Patterns (PAMPs) (Ali et al. 2007). In this work, we demonstrated that the

novel delayed flowering phenotype in dnd1 is not a by-product of SA accumulation.

It is likely another authentic biological role of CNGC2 (and CNGC4) and is likely

unique among CNGCs. Further analysis of the molecular mechanism of the delayed

flowering transition in dnd1 and cngc4 will shed light on this novel biological function

of CNGCs in flowering.

2.5 Acknowledgements

I would like to acknowledge Dr. Andrew Bent for providing the seeds of

various double mutants of dnd1.

31

Chapter 3

Identifying the causative mutation for repressor of defense no

death 1 (rdd1-1D), a novel suppressor of dnd1

32

Abstract

In Arabidopsis CNGCs are believed to play important roles in biotic and

abiotic stress responses, ion homeostasis, and development through their Ca2+

conducting capabilities. Several CNGCs have been implicated in plant-pathogen

interactions through genetic studies, including the defense, no death mutants dnd1

and hlm1/dnd2, which are null mutants of the closely related Arabidopsis CNGCs,

CNGC2 and CNGC4. The dnd mutants show distinct autoimmune phenotypes such

as dwarf morphology, elevated levels of salicylic acid and constitutive expression of

PR genes. Though these mutants have been well characterized phenotypically,

CNGC-mediated defense signaling is poorly understood.

In order to better understand CNGC2-mediated defense signaling, the first

dnd1 suppressor mutant, repressor of defense no death 1 (rdd1-1D), was previously

identified in the Yoshioka lab (Chin et al 2013). Map-based cloning and whole-

genome sequencing narrowed the causative mutation of rdd1-1D to four candidate

genes (Chin and Yoshioka, unpublished data). In this chapter, I aimed to specify one

of these candidate mutations as the causal mutation of rdd1-1D.

Current data indicates that rdd1-1D is a loss-of-function mutation in the auxin

biosynthesis gene YUCCA6. I explore a possible connection between CNGC2 and

auxin signaling in Chapter 4.

33

3.1 Introduction

Ion homeostasis is an important component of many plant signaling

pathways. Plant ion channels are unique compared to animal ion channels as they

primarily use Ca2+ and Cl- in the generation of action potentials (Ward et al., 2009).

This is due in part to Na+ toxicity in the majority of plant cells and the lack of a Na+

gradient necessary for action potential generation (Álvarez-Aragón et al. 2015).

Additional roles for plant ion channels include the release of ions from cellular

compartments or the influx of ions across the plasma membrane (Ward et al. 2009).

Specifically, ion flux across the plasma membrane has been shown to be an

important early component of plant immunity signaling (Gelli et al. 1997). Although

the Ca2+ channels that regulate this process are poorly characterized, members of

two large ion channel families, the glutamate receptor-like (GLR) family and the

cyclic nucleotide-gated ion channel (CNGC) family may be involved in this process

(Davenport 2002; Talke et al. 2003). CNGC2 has been extensively studied for its

role in immunity signaling, though there are no known components of the

downstream signaling pathway.

To identify novel components on CNGC2-mediated signaling a suppressor

screen was undertaken using a T-DNA knockout line of CNGC2. rdd1-1D cngc2 was

identified by its intermediate morphology between Col-wt and cngc2 plants (Chin et

al. 2013). Characterization of rdd1-1D cngc2 discovered that rdd1-1D suppresses

the majority of dnd1-mediated traits, including dwarf morphology, elevated levels of

34

SA and PR gene expression (Chin et al. 2013). Additionally rdd1-1D suppressed the

dnd1-mediated floral transition phenotype that was discussed in Chapter 2.

Map-based cloning mapped the location of the rdd1-1D mutation to a 800kb

region in the upper arm of chromosome 5 which contains 193 coding sequences

(Chin and Yoshioka, unpublished). Whole genome sequencing identified four

candidate genes, one of which contained a mutation in the coding region of

At5G25620. This gene is YUCCA6, an auxin biosynthesis gene, and one of 11

YUCCAs in Arabidopsis (Chin and Yoshioka, unpublished data). YUCCAs encode

flavin-containing monooxygenases (FMOs), a 29 gene family in Arabidopsis

(Schlaich 2007).

The Arabidopsis FMO family is separated into 3 clades, with FMO1 and a

pseudogene in clade 1, the 11 YUCCA genes in clade II, and 16 FMOs in clade III,

including one that has been shown to S-oxygenate glucosinolates (Schlaich 2007).

There are four motifs common to N-hydroxylating enzymes, an FAD-binding motif

(GXGXXG) towards the N-terminus, a NADP-binding motif (GXGXXG) and an FMO-

identifying motif (FxGxxxHxxxY/F) towards the center, and a ‘L/FATGY’ motif

towards the C-terminus (Stehr et al. 1998). Little is known about the role of the

FATGY motif, however it has been postulated that the L/FATGY is part of the

substrate binding site providing a hydrophobic pocket (Schlaich 2007; Stehr et al.

1998).

In this chapter I investigate the causative mutation of rdd1-1D and explore if

the rdd1-1D is a gain or a loss-of-function mutation.

35

3.2 Materials and Methods

3.2.1 Plant Growth Conditions

For segregation analysis Arabidopsis thaliana seeds were sown on Sunshine

Mix #1 (Sun Gro Horticulture Canada Ltd) and stratified at 4oC for 4 days. Plants

were grown in a growth chamber with a 9 hour photoperiod (9 hour light/ 15 hour

dark) at 22oC.

For the monitoring of floral transition Arabidopsis thaliana seeds were sown

on Sunshine Mix #1 (Sun Gro Horticulture Canada Ltd) and stratified at 4oC for 4

days. Plants were grown in a growth chamber with a 16 hour photoperiod (16 hour

light/ 8 hour dark) at 22oC.

For the monitoring of the measurement of endogenous SA levels Arabidopsis

thaliana seeds were sown on Sunshine Mix #1 (Sun Gro Horticulture Canada Ltd)

and stratified at 4 oC for 4 days. Plants were grown in a growth chamber with a 9

hour photoperiod (9 hour light/ 15 hour dark) at 22oC.

Summary of all mutants utilized in this chapter is provided in Table A2-1

3.2.2 Double Mutant Analysis

To investigate the nature of the rdd1-1D mutation epistatic analysis was

conducted using two T‐DNA insertion lines for YUCCA6. These mutants are yucca6‐

3k (salk_093708C) and yucca6‐1D (CS67234). yucca6‐3k has a T-DNA insertion in

36

its first intron resulting in loss-of-function knockout mutation. Contrarily, yucca6‐1D

has a T-DNA insertion approximately 10,000 bp upstream of the start codon for

YUCCA6, resulting in its overexpression. For yucca6-3k homozygosity of the

insertion was confirmed by PCR analysis using gene specific primers for YUCCA6

(Salk_093708C-LP, Salk_093708C-RP; Table A3-1). For yucca6-1D homozygosity

of the insertion was confirmed by PCR analysis using insertion position specific

primers for YUCCA6 (Yucca6-1D-F, Yucca6-1D-R; Table A3-1).

Each of these mutants was crossed with the CNGC2 T-DNA knockout

mutant cngc2-3 (Salk_066908). Previous work confirmed the knockout status of

cngc2-3 and yucca6-3k by RT-PCR (Kim et al. 2007; Chin et al. 2013). For each

cross the F1 was allowed to self-pollinate and F2 seeds were collected and were

sown for segregation analysis as described.

3.2.3 Trypan Blue Staining

Leaf samples were taken from 3–4 week old plants grown on soil, and

vacuum-infiltrated with trypan blue staining solution (10ml of 1 mg/ml trypan blue

dye, 10ml acetic acid, 10ml glycerol, 9.6ml phenol). Leaves were then boiled for 4

minutes and incubated overnight at room temperature. Leaves were then de-stained

with chloral hydrate (2.5g/ml) and mounted on 80% glycerol for light microscopy.

3.2.4 Measurement of endogenous salicylic acid

An SA biosensor, Acinetobacter sp. ADPWH_lux, was utilized to measure

endogenous SA levels in Col-wt and the various mutants we are investigating. This

37

biosensor was originally developed by Huang et al. (2005), and further modified by

DeFraia et al. (2008). Acinetobacter sp. ADPWH_lux. is a genetically engineered

Acinetobacter strain which contains a SA-inducible luxCDABE operon that produces

a luminescent signal with intensity relative to the quantity of SA present.

For this analysis plant extracts were prepared from 50-100 mg of leaf tissue

collected from 5-6 week old plants. Tissue was ground in 300 μl of acetate buffer

(pH 5.6), and after centrifugation (10 min at 14,000 rpm), the supernatant was

collected and divided into two tubes. One half of each sample was hydrolyzed with

β-glucosidase at 37°C for 90 minutes. Standard samples of various concentration of

SA were made by adding know amounts of SA into mixture of ethanol, acetate buffer

and Col-wt plant extract. Overnight culture of Acinetobacter sp. ADPWH_lux. was

adjusted to OD600=0.4, and 60μl of the culture, 50μl of LB and 20μl of either

standard mixture or plant extract were placed into wells of black 96-well plates, then

incubated at 37° C for 60 minutes before measuring the luminescence in a

microplate reader (Tecan, San Jose). This analysis was completed by the Yoshioka

lab undergraduate student Ms. Megumi Bachmann.

3.2.5 Plasmid Construction and Agrobacterium mediated

transformation for complementation analysis

For complementation I have constructed cDNA and genomic clones for

YUCCA6, both of which contain the rdd1 mutation. The genomic clone begins

approximately 2000bp upstream of the coding sequence and was cloned from rdd1-

1D cngc2 genomic DNA into the plant expression vector PBIN19 in the XmaI and

38

SpeI restriction enzyme sites and the insertion was sequenced for fidelity

(proYUCCA6::YUCCA6rdd1). The expression construction was transformed into

Agrobacterium tumefaciens strain GV3301 and transformed into dnd1 plants using

the floral dip method (Clough and Bent 1998).

For the overexpression construct, the YUCCA6 was cloned from rdd1-1D

cngc2 cDNA into a modified PBI121 plant expression vector using the XbaI and StuI

restriction enzyme sites (termed CaM35S::YUC6rdd1). In the modified PBI121 vector

the GUS gene has been removed. In this plasmid the expressed gene is controlled

by a strong constitutive CaMV35S promoter (Wang and Bai 1990). The insertion

was sequenced for fidelity. The plasmid was transformed into the Agrobacterium

tumefaciens strain GV3301 and transformed into dnd1 and Col-0 plants using the

floral dip method (Clough and Bent 1998).

3.2.6 Pathogen infection assay

Infection with Hyaloperonospora arabidopsidis (Hpa) isolate Noco2, was

performed as described previously with 2 x105 spores/ml (Yoshioka et al., 2006).

Infections have been performed using 3-4 week old plants as well as 7 day old

seedlings. Additional experiments are being completed using a higher spore titer of

8 x105 spores/ml.

39

3.3 Results

3.3.1 Location of rdd1-1D mutation in YUCCA6

As discussed previously, rdd1-1D contains a non-synonymous amino acid

substitution of a proline to leucine at residue 289 in YUCCA6. YUCCA6 contains

four motifs common to N-hydroxylating enzymes, an FAD-binding motif (GXGXXG)

toward N-terminal, a NADP-binding motif (GXGXXG) and FMO-identifying motif

(FxGxxxHxxxY/F) toward center, and a ‘L/FATGY’ motif toward C-terminal. As

shown in Figure 3-1 residue 289 is not contained in any of these regions. However,

the region surrounding the P-289-L mutation is conserved among the 11 YUCCA

family members, and though uncharacterized may be necessary for proper enzyme

function.

40

41

Figure 3-1: Alignment of rdd1-1D mutation in YUCCA6. A) Location of rdd1-1D

proline to leucine mutation (P289L) relative to conserved domains. YUCCA enzymes

contain two conserved GXGXXG motifs: NADPH (blue) and FAD (green) binding

motif. Additionally a FMO-identifying motif FxGxxxHxxxY/F’ (grey) and a L/FATGY

motif (brown). B) Alignment of 278-337 amino acid region in YUCCA6. The proline at

residue 289 is conserved in YUCCA1-10, as indicated by red arrow.

42

3.3.2 yucca6-3k is able to suppress dnd1-conferred dwarf

morphology

To test if the YUCCA6 knockout mutant yucca6-3k could suppress cngc2-

mediated dwarf morphology the F2 progenies of a yucca6‐3k x cngc2 cross (148 F2

plants) were analyzed. If yucca6-3k is able to suppress cngc2-mediated

morphological phenotypes, then one would expect a segregation ratio of 9:3:3:1

(Col‐0: yucca6‐3k: cngc2: rdd1 cngc2). However, yucca6‐3k single mutant does not

have significant morphological difference from Col‐0, and thus I combined the two

categories “Col‐0” and “yucca6‐3k” into a single category “Col-0 wild type like”,

resulting in a prediction of 12:3:1 (Col‐0 wild type like: cngc2: rdd1 cngc2). The

morphological segregation result is shown in Table 3-1. The actual segregation was

12.8:2.6:0.6. χ2 analysis accepts the hypothesis that yucca6‐3k suppresses cngc2

morphological phenotypes (P=0.22). The phenotype of the double mutant is shown

in Figure 3-2.

43

44

45

Figure 3-2: Morphological and spontaneous cell death analysis in Col-0 and cngc2

single and double mutant plants. (Left panel) Morphology of 5 week old Col-0 and

mutant plants grown in short day conditions (12L:12D). yucca6-3k cngc2 is able to

partially suppress cngc2-conferred dwarf morphology. yucca6-3k, the YUCCA6

knockout line, has shorter and wider rosette leaves compared to Col-0. Scale bar =

1 cm. (Right panel) Trypan blue staining reveals a reduction in cell death in yucca6-

3k cngc2 compared to cngc2. Cell death is indicated by blue stain and highlighted by

red arrows. Plants used were approximately 5 weeks old. Scale bar = 0.5 mm.

46

3.3.3 YUCCA6 over expressing line does not suppress dnd1-

conferred dwarf morphology

I have generated yucca6-1D cngc2 double mutants to test if over expression

of YUCCA6 can suppress cngc2-conferred dwarf morphology. yucca6-1D is an over-

expresser mutant of YUCCA6 which has a T-DNA insertion in 5’ UTR of YUCCA6

(Kim et al. 2007).

yucca6‐1D plants have a morphological phenotype characterized by long

petioles and extremely narrow leaves. This is a typical phenotype for auxin over-

accumulating mutants (Kim et al. 2007). 148 plants from a F2 population of a

yucca6-1D x cngc2 cross were analyzed for morphological segregation. If yucca6-

1D is able to suppress cngc2 the expected segregation ratio would be 3:9:1:3 (Col:

yucca6‐1D: cngc2: rdd1). As shown in Table 3-2, χ2 analysis rejects this hypothesis.

In the population of 148 plants analyzed, no plant was observed to have an rdd1-1D

cngc2 morphology, suggesting that yucca6‐1D does not suppress cngc2.

Interestingly, a plant with of comparable size to cngc2, but with high auxin

phenotypes was observed (Figure 3-3). This plant was genotyped to be

homozygous for cngc2 and contain a single or double copy of yucca6-1D.

47

48

49

Figure 3-3: Morphology of Col-0 and mutant plants overserved in yucca6-1D x

cngc2 segregation analysis. Plants are approximately 5 weeks and grown in short

day conditions (12L:12D). yucca6-1D is unable to suppress dnd1-conferred dwarf

morphology. yucca6-1D is an over expression line of YUCCA6 and accumulates

auxin. Scale bar = 1 cm.

50

3.3.4 yucca6-3k is able to suppress dnd1-conferred spontaneous

cell death formation

To further characterize the suppression of cngc2 by the loss-of-function

mutant yucca6-3k the spontaneous cell death formation phenotype of yucca6-3k

cngc2-3 was investigated. Leaves were taken from 4 week old Col-wt and mutant

plants grown in short day (L9:D15) and stained with trypan blue.. As mentioned,

dnd1 (and cngc2) plants exhibit conditional spontaneous lesion formation. When

trypan blue stained leaves of yucca6-3k cngc2-3 plants were observed there was

less spontaneous cell death observed compared to cngc2 (Figure 3-2B). This

experiment further supports our finding that rdd1-1D is a loss-of-function allele of

YUCCA6.

3.3.5 yucca6-3k is able to suppress dnd1-conferred delayed

flowering phenotype

Our lab recently discovered that dnd1 has delayed floral transition compared

to Col-0, and that rdd1-1D can suppress this phenotype (Chin et al., 2013). Through

monitoring floral transition it was found that the loss-of-function mutant yucca6-3k is

able to suppress dnd1-mediated delayed floral transition (Figure 3-4), providing

further evidence that rdd1-1D is a loss-of-function mutant of YUCCA6. In addition,

since we successfully isolated rdd1-1D single mutant (without dnd1 background), we

also investigated if rdd1-1D single mutant had alterations in floral transition. It was

observed that rdd1-1D and yucca6-3k single mutants did not display any statistically

significant alterations in floral transition compared to Col-wt.

51

52

Figure 3-4: Floral transition in Col-0 and mutant plants. A) morphology of 5 week old

plants grown in long day (16L:8D). B) Average days to bolting in Col-0 and mutant

plants. Both rdd1-1D and yucca6-3k are able to suppress cngc2-conferred delayed

floral transition. Additionally, both yucca6-3k cngc2-3 and rdd1-1D cngc2-3 exhibit

slightly earlier floral transition than Col-wild type. Both rdd1-1D and yucca6-3k single

mutants do not display any flowering delays compared to Col-0 . Error bars = SEM.

Bars marked with the same letter indicate no significant difference (Student’s t-test,

p < 0.05). Plants were grown in Sunshine Mix #1 with a photoperiod of 16 h light and

8 h dark. All experiments have been repeated at least three times with similar

results.

53

3.3.6 yucca6-3k is able to suppress dnd1-conferred SA

accumulation

As discussed previously, dnd1 has an elevated basal level of SA compared to

Col-wt. Several of the dnd1-phenotypes have been shown to be SA-independent,

including its dwarf morphology and inability to produce HR, while the enhanced

resistance phenotype has been shown to be SA-dependent (Clough et al. 2000). As

discussed in Chapter 2, the novel delayed flowering phenotype of dnd1 has also

been shown to be SA-independent (Fortuna et al. 2015). Additionally, rdd1-1D has

been shown to suppress SA accumulation in our CNGC2 knockout line. To gain

further insight into the nature of the rdd1-1D mutation we compared SA

accumulation in rdd1-1D cngc2-3 to yucca6-3k cngc2-3. For this analysis we used a

SA biosensor which contained an SA inducible luminescent operon, allowing us to

compare relative fluorescence between our samples (DeFraia et al. 2008). In our

analysis it was observed that SA accumulation in rdd1-1D cngc2-3 and yucca6-3k

cngc2-3 was significantly reduced compared to cngc2 (Figure 3-5). This observation

further supports our finding that rdd1-1D is a loss-of-function mutation of YUCCA6.

This analysis was completed by the Yoshioka lab undergraduate student Ms.

Megumi Bachmann.

54

55

Figure 3-5: Endogenous free SA and total SA level (free SA + SAG) in Col-0 and

mutant leaves per milligram of fresh weight. Error bars indicate standard error of the

mean of three independent experiments (n=9).

56

3.3.7 rdd1-1D is unable to suppress cpr22-conferred phenotypes

To understand if rdd1-1D suppresses cngc2-conferred phenotypes

specifically, or can broadly suppress other CNGC related lesion mimic mutants we

generated double mutants of rdd1-1D (Col-0 background) with cpr22 (Ws

background). Since this double mutant is mixed background of two different

ecotypes, we generated additional control plants Col-0 wt crossed with cpr22 to

accurately compare the phenotypes. As both rdd1-1D and cpr22 are dominant

mutations we analyzed the F1 progeny. As shown in Figure 3-6A, F1 progenies of

rdd1-1 D and cpr22 are almost identical to F1 progeny of Col-0 and cpr22 indicating

that rdd1-1D is unable to suppress the cpr22 related curly leaf morphology.

Additionally, spontaneous cell death formation of these progenies was analyzed by

trypan blue staining. As seen in Figure 3-6B, the amount of cell death observed in

rdd1-1D cpr22 was similar to that of cpr22 and Col-0 cpr22. Taken together, these

results indicate that although dnd1 and cpr22 share similar autoimmunity

phenotypes, their respective wild type channels likely have distinct molecular

mechanisms underlying their mutant phenotypes.

57

58

Figure 3-6: Morphological and spontaneous cell death analysis in Col-0, Ws-2 and

mutant plants. (Left panel) Morphology of 5 week old Col-0, Ws-2 and mutant plants

grown in short day conditions (12L:12D). F1 progenies of a rdd1-1D cngc2 x cpr22

cross were analyzed for cpr22-conferred curly leaf morphology. rdd1-1D is unable to

supress this trait in cpr22 plants. Scale bar = 1 cm. (Right panel) spontaneous cell

death analysis in Col-0, Ws-2 and mutant plants. Cell death of 5 week old Col-0 and

mutant plants grown in short day conditions (12L:12D). F1 rdd1-1D cngc2 x cpr22

trypan blue staining reveals rdd1-1D is not able to suppress cpr22-conferred

spontaneous cell death. Scale bar = 500 µM.

59

3.3.8 Complementation analysis

As described previously, both proYUCCA6::YUCCA6rdd1 (genomic fragment

of YUCCA6 gene under the control of YUCCA6 native promoter construct; Table 3-

3) and CaM35S::YUCCA6rdd1 (YUCCA6 coding sequence under the constitutive

CaM35S promoter construct; Table 3-3) constructs have been created and

transformed into dnd1. Additionally, the CaM35S::YUC6rdd1 construct has been

transformed into Col-0. The outline for complementation analysis is described in

Figure 3-7. Kanamycin resistant T1 plants were selected in MS agar plate and

transplanted to soil where they grew to maturity and seeds were collected.

For both the genomic and overexpressor complementation 7 T1 plants were

identified. From 7 T1 lines, T2 seeds were sown onto 0.5X MS plates containing 50

µg/ml Kanamycin for segregation analysis, allowing us to determine whether each T1

line possessed single or multiple copy insertions (Table 3-4). T2 plants were then

transplanted off selection media for morphological analysis. Due to the stress of

transplanting as well as the variable nature of the cngc2 phenotype I am unable to

complete complementation analysis at this time. I am currently waiting to harvest T2

seeds.

In parallel to our search for single-copy homozygous lines, I have sown T2

seeds directly to soil for morphological analysis. This may give us an idea about

complementation before we are able to harvest T3 seeds from our T2 individuals.

These T2 lines are being grown in short day conditions (8L:16D) with appropriate

controls: Col-0, cngc2, rdd1-1D cngc2, yucca6-3k, and yucca6-1D (Table 3-4).

60

61

Figure 3-7: Schematic diagram of complementation analysis 1) Antibiotic resistant

T1 transformants have been selected from Kanamycin selection media. 2) These T1

individuals were transplanted to soil where they were grown for seed collection. 3)

Once T2 seeds were collected they were sown again onto Kanamycin selection

media where segregation was analyzed. 4) Antibiotic resistant individuals were

again transplanted to soil for morphological analysis and seed collection. Due to the

variable nature of the cngc2 phenotype we could not conclude on complementation

in this generation. 4) In parallel positive T2 seeds have been sown directly to soil for

morphological analysis. 5) T3 seeds sown onto selection media for analysis of single

and multi-copy insertion status. 6) Single copy homozygous individuals will be sown

onto soil for phenotypic analysis (not illustrated).

62

63

64

3.3.9 Pathogen analysis

As discussed previously, dnd1 plants display an enhances resistance to Hpa

Noco2 and Pst DC3000 (Yu et al. 1998; Chin et al. 2013). Previous work completed

in our lab has determined that rdd1-1D is able to suppress this enhanced resistance

(Chin et al. 2013). To investigate if yucca6-3k can suppress pathogen resistance

comparable to rdd-1D pathogen resistance analysis of rdd1-1D cngc2 alongside

yucca6-3k cngc2 is in progress. To date I have been unable to replicate the

breakdown in resistance previously observed in rdd1-1D cngc2-3. This is likely due

to the difference of concentration of spores. I am currently repeating pathogen

analysis using a titer of 8 x105 spores/ml which was used previously.

3.4 Discussion

Ca2+ signal transduction is one of the earliest events in pathogen defense in

plants. Yet, the molecular identify of ion channels involved in this critical signal

transduction pathway remain elusive. As a calcium permeable channel, CNGC2 has

been extensively studied in the context of pathogen defense activation. CNGC2 null

mutant dnd1 has several immunity related phenotypes as discussed, including broad

spectrum disease resistance and inability to induce HR upon infection with avirulent

pathogens. It can be difficult to reconcile these two phenotypes, as dnd1 has

enhanced immunity while simultaneously negatively regulating a pathway commonly

associated with immunity signaling, HR. Another feature which makes dnd1

interesting is how certain phenotypes are SA-dependent while others are SA-

65

independent. As discussed in Chapter 2, the delayed flowering phenotype is SA-

independent. Additionally the dwarf morphology and inability to induce HR are SA-

independent, while the enhanced resistance phenotype is SA-dependent. It is

therefore likely CNGC2 has wide ranging physiology roles related to pathogen

defense, growth and development.

As discussed in this chapter, the first suppressor of dnd1, rdd1-1D has been

identified. rdd1-1D is able to suppress many of the dnd1-conferred phenotypes,

including dwarf morphology, enhanced resistance, delayed flowering, and

spontaneous cell death. Here, I have presented data which indicates that rdd1-1D is

a loss-of-function mutation in the auxin biosynthesis gene YUCCA6, as the loss-of-

function mutant yucca6-3k is able to complement cngc2 in a manner nearly identical

to rdd1-1D.

Several of the pleiotropic phenotypes of the dnd1 mutant, as discussed, are

independent of SA signaling. As rdd1-1D is a loss-of-function mutation in the auxin

biosynthesis gene YUCCA6, dnd1-conferred phenotypes may be related to auxin

signaling. It has been reported that these two hormones, SA and auxin, are

antagonistic in defense responses. For example, the bacterial PAMP flg22 induces

the accumulation of several microRNAs which target auxin signaling components,

including TIR and AFB proteins, which in turn leads to a down regulation of auxin

responsive genes (Navarro et al. 2005). However, as rdd1-1D is able to suppress

SA accumulation in dnd1 and cannot suppress cpr22-conferred leaf morphology and

spontaneous cell death it is likely rdd1-1D suppression is independent of a simple

SA-antagonism pathway.

66

The role of auxin in plant defense responses is not well studied and remains

to be elucidated, especially in the context of CNGC-mediated signaling. Future

investigation is required to provide novel insight into the role of CNGC and Ca2+ in

auxin signaling during plant defense activation.

67

Chapter 4

Investigating a Role for Cyclic Nucleotide Gated Ion Channel 2

in Auxin Homeostasis

68

Abstract

Ca2+ is an important secondary messenger in plants which plays a critical role

in growth and development. Changes in cytosolic Ca2+ concentration is a central

mechanism by which plants are able to respond to biotic and abiotic stress. For

example, one of the earliest responses to pathogen recognition is Ca2+ influx across

the plasma membrane. CNGC2 is a Ca2+ permeable channel and has been

connected pathogen defense through the well-studied autoimmune mutant dnd1.

The first novel suppress of dnd1, repressor of defense, no death 1 (rdd1-1D),

has a loss-of-function mutation in the auxin biosynthesis gene YUCCA6. Auxin is a

plant hormone with important and wide ranging roles in growth and development,

which has also been connected to pathogen signaling. However, how auxin is able

to produce such a range of downstream signaling responses remains to be fully

elucidated. One possibility is that auxin utilizes Ca2+ as a signaling component. As

CNGC2 is a Ca2+ permeable channel, and RDD1 is likely the auxin biosynthesis

gene YUCCA6, I explore in this chapter the possibility that CNGC2 is involved in

auxin signaling.

69

4.1 Introduction

Auxin is an important plant hormone studied extensively for its role in growth

and development. Auxin, which in Greek means “to grow”, has been studied in

plants for more than 120 years, and Charles Darwin was one of the first researchers

to study auxin when he observed that canary grass coleoptiles grow toward

unidirectional light (Darwin 1880). Although a growth promoting molecule was known

to exist, it was not for some time that auxin was isolated from plant extracts (Went

1926). The most common naturally occurring auxin in plants is Indole-3-acetic acid

(IAA). However, several other growth promoting molecules with similar structures

occur in plants, all of which are commonly referred to as auxins (Sugawara et al.

2015).

The isolation of a complete auxin biosynthesis pathway has remained elusive

for plant researchers until quite recently. This is presumably due to the complexity

and redundancy of the auxin biosynthesis pathway. YUCCA genes were identified in

an activation-tagging screen for auxin biosynthesis mutants (Zhao et al. 2001).

YUCCA enzymes are hypothesized to catalyze the rate-limiting step in auxin

biosynthesis, the conversion of indole-3-pyruvic acid (IPA) to IAA (Mashiguchi et al.

2011; Dai et al. 2013). The overexpression of YUCCA6 has been connected to a

variety of phenotypes, including drought tolerance, reduced ROS accumulation,

increased plant height, narrowed leaves and delayed leaf senescence (Zhao et al.

2001; Kim et al. 2011; Kim et al. 2013)

70

Auxin is synthesized in fast growing tissues, such as the root tip and

hypocotyls, where it can be transported throughout the plant as required. However,

localized auxin biosynthesis is also important and may help to explain the

developmental defects observed in YUCCA double knockout lines. As discussed

previously, yucca1 yucca4 mutants which have severe defects related to sterility and

vein patterning cannot be rescued by exogenous application of auxin, but can be

rescued by expressing the auxin biosynthesis gene iaaM driven by the native

YUCCA1 or YUCCA4 promoter (Cheng et al. 2006; Zhao 2008). Thus, this data

supports the importance of localized auxin biosynthesis.

The auxin signal was believed to be perceived by one of two coreceptor

complexes, one involvi