BABA誘發免疫反應對抗細菌性病原菌之作用機制

(1)

i

國立臺灣大學生命科學院植物科學研究所碩士論文

Institute of Plant Biology College of Life Science

National Taiwan University Master Thesis

BABA 誘發免疫反應對抗細菌性病原菌之作用機制 Dissecting out the Role of β-Aminobutyric Acid in Plant

Immunity against Bacterial Pathogen Pst DC3000

研究生：朱柏威 Student: Po-Wei Chu

指導教授：金洛仁博士 Advisor: Laurent Zimmerli, Ph.D.

中華民國 99 年 10 月

October, 2010

(2)

ii

國立臺灣大學碩士學位論文口試委員會審定書

BABA 誘發免疫反應對抗細菌性病原菌之作用機制 Dissecting out the Role of β-Aminobutyric Acid in Plant

Immunity against Bacterial Pathogen Pst DC3000

本論文係朱柏威君(R97B42019)在國立台灣大學植物科學研究所完成之碩士學位論文，於民國 99 年 10 月 20 日承下列考試委員審查通過及口試及格，特此證明

口試委員:

台灣大學植物科學研究所助理教授

金洛仁博士 Dr. Laurent Zimmerli 台灣大學植物科學研究所副教授

吳克強博士台灣大學農業化學研究所助理教授

林乃君博士

(3)

i

誌謝

我曾經參加過許多學長姐的論文口試。那等待最終宣判的時刻，總是全程的最高潮。一旦門後的會議結束，學生的命運隨即抵定。踏進教室聆聽結果的這段路途，既漫長而又沉重。我總好奇裡頭等待學長姐的究竟為何？這是局外人無法窺探的秘密。如今，我終於親自踏上這條征途，為兩年的研究旅程畫上句點。

這兩年，充滿著扎實的研究與學習，指導教授 Dr. Laurent Zimmerli 著實功不可沒。在 Dr. Zimmerli 引領下，我踏出植物分子生物學的第一步；他並且教導，

如何令我每個步伐走得嚴謹而穩健。我體認到科學研究是門重視邏輯與思考的藝術，而非任意拼湊無數實驗的雜燴料理。我的論文嘗試集合兩年的研究成果，並

透過嚴謹的論述，將其鋪陳為一篇具啟發性的「故事」。也感謝兩位口試委員，

吳克強老師與林乃君老師，百忙中抽空協助修改文章，使論文更臻完美。

研究的路途必然有許多挫折，面對這些困境，多虧可愛的實驗室成員們願意傾聽與給予鼓勵，讓我在數不盡的實驗死巷裡重燃希望。尤要感謝敬偉學長、

Terry 學長，我們曾經組成默契絕佳的敢死隊，一起篩選過無數的突變株。沒有這段苦辛，必然沒有今日的 LRR-PK 小組(笑)。在敬偉學長、小喬與佳楠的共同協助下，我們一起令 LRR-PK 小組茁壯成長。同時感謝實驗室的博士後研究員，

Dominique，他熱心地指導我完成研究生涯的第一個基因轉殖阿拉伯芥。吳老師實驗室的峻惟學長與芳芳，提供許多實驗技術的協助，令我感念不已。這些共同奮鬥的戰友們，都是促成這篇論文的無名英雄。

兩年來，我將不少夜晚與假日的休息、陪伴親友的寶貴時間，奉獻於實驗中。

為此我想向遠在台中的父母親與女友，豆豆，致上無比歉意。進入研究所後，返鄉的次數明顯減少許多。對於不明瞭孩子遠在異鄉為何而忙的雙親而言，這必是

(4)

ii

一種難熬的等待。同時我非常感激豆豆，總是義無反顧的陪伴在我身邊，一起分享無數喜樂與哀愁。

兩年，對於研究工作而言實不算長，我僅能淺嚐「科學」這道佳餚。然而，

我從其中認識了鑽研任何課題應該掌握的要旨與不輕言放棄的態度。因此無論人生下一階段為何，我相信必能秉持這兩年所學，並全力以赴。

謹以此論文，獻給所有關心我的師長、家人、同學與朋友們。

朱柏威謹誌於台中家中 2010/10/25

(5)

iii

摘要

BABA (beta-aminobutyric acid)是一種無法用於合成蛋白質的胺基酸，但是卻被發現能有效促進植物對病原菌的抵抗性。植物經 BABA 處理後，一旦感懶細 菌性斑點病病原菌 (Pseudomonas syringae pv. tomato DC3000, Pst DC3000)，水楊 酸 (salicylic acid, SA) 訊息傳導途徑中的 PR1 基因將被快速誘發而表現。本論 文報導由 BABA 所引起的病原菌抵抗性，並非完全依賴於水楊酸訊息傳導途徑；

BABA 能誘發植物的早期免疫反應 (plant early defense response)。一經細菌 Pst DC3000 感染，BABA 誘發諸多免疫基因表現，且這些基因分別隸屬於不同訊息傳導途徑，例如，MAPK (mitogen-activated protein kinases) ，乙烯 (ethylene, ET)，

以及其他訊息途徑。同時，BABA 能有效誘發細菌感染時所引起的胝質累積 (callose deposition)。除此之外，如將水楊酸的突變株 sid2 以 RNA 干擾技術(RNA interference)造成 PR2 基因靜默(gene silencing)，不會改變 BABA 對該突變株的作 用，顯示 PR2 的存在與否對於 BABA 的作用沒有影響。為深入瞭解 BABA 如何 調節阿拉伯芥的免疫作用，金洛仁博士的實驗室成員們測試七十餘株突變株對於 BABA 的敏感性與細菌 Pst DC3000 的感病性。本論文發現某一 LRR 家族蛋白激 酶 (leucine-rich repeat protein kinase, LRR-PK)的突變株，不僅易受細菌感染、胝質的累積量減少，且對於 BABA 的敏感度降低。該 LRR-PK 表現於細胞膜。且於植物遭受細菌感染，或者感受到病原相關分子模式 (pathogen-associated molecular patterns, PAMPs)時，LRR-PK 基因能被快速誘導表現。初步證據顯示此 基因參與阿拉伯芥早期的免疫反應。除此之外，本研究亦發現四磷酸雙腺苷水解酶(Ap₄A hydrolase)對於阿拉伯芥抵抗腐生性病原菌(necrotrophic pathogen)具有 重要功能。該基因的突變株對於真菌性灰黴病病原菌(Botrytis cinerea)和細菌性軟

(6)

iv

腐病病原菌(Erwinia carotovora subsp. carotovora)較為感病。

關鍵字：BABA、阿拉伯芥、細菌性斑點病菌、水楊酸、植物早期免疫反應、胝質累積、蛋白激酶、四磷酸雙腺苷水解酶

(7)

v

Abstract

The non-protein amino acid beta-aminobutyric acid (BABA)has been known for years to be an effective inducer of resistance against pathogen infection. When BABA-pretreated Arabidopsis plants are challenged with the pathogenic bacteria Pseudomonas syringae pv. tomato DC3000 (Pst DC3000), a strong potentiation of PR1 expression is observed. PR1 is the marker gene for salicylic acid (SA)-related defense response. Here, we report that BABA-mediated bacterial resistance acts in a partly SA-independent manner to prime the early Arabidopsis defense response. Upon Pst DC3000 infection, mitogen-activated protein kinase (MAPK) cascade genes, ethylene (ET) and other signaling pathways were potentiated by BABA pretreatment in both Col-0 wild type and in the SA biosynthetic sid2 mutant. Callose deposition in response to Pst DC3000 inoculation was also primed by BABA. In addition, RNAi-mediated PR2 gene silencing in sid2 mutant demonstrated BABA-induced resistance (BABA-IR) to Pst DC3000 similar to sid2 indicating that PR2 is not critical for BABA-induced resistance in sid2.

To further investigate how BABA regulates Arabidopsis immunity, members of Zimmerli’s laboratory tested over 70 T-DNA knock-out mutant lines for their sensitivity towards BABA and Pst DC3000 infection. Two T-DNA knock-out lines of a leucine-rich repeat protein kinase (LRR-PK) demonstrated enhanced bacterial susceptibility, reduced callose deposition, as well as impaired BABA sensitivity. This LRR-PK localizes to the plasma membrane. Upon Pst DC3000 infection and PAMPs treatment, this LRR-PK gene was rapidly induced. Hence, our preliminary data

(8)

vi

suggest that this gene is involved in Arabidopsis early defense responses. In addition, one gene was found important for plant resistance to necrotrophic pathogens.

Mutation of AtNUDX25, one Ap4A hydrolase, led to enhanced susceptibility to Botrytis cinerea and Erwinia carotovora subsp. carotovora strain WPP14 (Ecc WPP14).

Key words: BABA, Arabidopsis, Pseudomonas syringae, SA, early defense response, callose deposition, leucine-rich repeat protein kinase, Ap4A hydrolase

(9)

vii

List of Figures

Figure1. BABA induced resistance in the SA biosynthetic mutant sid2……….40 Figure 2. BABA pretreatment up-regulates plant innate immunity marker genes in basal and induced expressions independently of SA biosynthesis………...42 Figure 3. BABA pretreatment enhances callose deposition after Pst DC3000 infection independently of SA biosynthesis………45 Figure 4. BABA pretreatment does not change flg22-induced resistance…………...48 Figure 5. PR1 expression is highly dependent on SA biosynthesis……….49 Figure 6. Outline of the map of RNAi construct for PR2 gene silensing and

confirmation of construct by restriction enzyme digestion and PCR………...50 Figure 7. Transgenic plants are knock-down of PR2 expressions………...51 Figure 8. Bacterial growth assays for PR2 gene silencing transgenic lines…………52 Figure 9. AtNUDX25 structure and identification of atnudx25 mutant………….…..53 Figure 10. The atnudx25 mutant is more susceptible to fungal pathogen B. cinerea..54 Figure 11. The atnudx25 mutant is more susceptible to bacterial pathogen E.

carotonova………....……55 Figure 12. The atnudx25 mutant responses normally to bacterial pathogen P. syringae and bacterial PAMP flg22……….…...56 Figure 13. The putative LRR-PK organization and Identification of lrr-pk mutants...57 Figure 14. Mutation in the putative LRR-PK causes BABA partial insensitivity and enhanced susceptibility to Pst DC3000………58

(10)

viii

Figure 15. The putative LRR-PK is induced in response to bacterial infection and PAMP perception………...60 Figure 16. The putative LRR-PK localizes to the plasma membrane……….…61 Figure 17. Mutation in the putative LRR-PK reduces callose deposition during Pst DC3000 infection and flg22 treatment……….…62 Figure 18. Proposed model for the contribution of β-Aminobutyric acid (BABA) in PAMPs-triggered defenses………..….……64

(11)

ix

List of Tables

Table 1. Primers for Real-Time and semi-quantitative RT-PCR...65 Table 2. Primers for identifying homozygous T-DNA mutant lines...66 Table 3. Primers for PR2 fragment of RNAi Construct...66

(12)

x

Abbreviations

BABA β-aminobutyric acid

Pst DC3000 Pseudomonas syringae pv. tomato DC3000 Pst DC3000 hrcC Pseudomonas syringae pv. tomato DC3000 hrcC^- PAMP pathogen-associated molecular pattern

PTI PAMP-triggered immunity SA salicylic acid

JA jasmonic acid ET ethylene

PR1 PATHOGENESIS-RELATED 1 PR2 PATHOGENESIS-RELATED 2

SID2 SALICYLIC ACID INDUCTION DEFICIENT 2 NPR1 NONEXPRESSOR OF PR GENES 1

PDF1.2 PLANT-DEFENSIN 1.2 FLS2 FLAGELLIN SENSITIVE 2

BAK1 BRI1-ASSOCIATED RECEPTOR KINASE 1 FRK1 FLG22-INDUCED RECEPTOR-LIKE KINASE 1 CBP60g CAM-BINDING PROTEIN 60-LIKE G

LRR-PK leucine-rich repeat protein kinase

(13)

xi

Introduction

PAMP-Triggered Immunity against Bacteria: Pattern Recognition Receptors Pathogen perception and quick induction of defense response is important during plant innate immunity. This form of immune response is mediated by a group of membrane proteins termed “pattern recognition receptors” (PRRs), which recognize highly conserved molecular signatures characteristic of whole class of microbes defined as “pathogen-associated (or microbe-associated) molecular patterns” (PAMPs or MAMPs). Perception of PAMPs by PRRs leads to a chain of signaling events in plants, which is called PAMP-triggered immunity (PTI) (Boller and Felix, 2009;

Nicaise et al., 2009).

One well studied PAMP is flagellin, the protein component of bacterial motility organ flagellum. Flg22, a synthetic 22-amino-acid polypeptide that corresponds to a highly conserved region of eubacterial flagellin, elicits responses in most plant species and acts as potent elicitors at subnanomolar concentrations (Felix et al., 1999).

Flg22 treatment leads to strong growth inhibition in young seedling. This was used in a mutant screen, yielding a number of mutants that were insensitive to flg22. One of the mutated loci, FLAGELLIN-SENSING2 (FLS2) was found to encode a LRR receptor-like kinase (LRR-RLK) and later proved to be the PRR for flg22 (Gómez-Gómez and Boller, 2000; Chinchilla et al., 2006).

Besides flg22, the elongation factor Tu (EF-Tu) acts as a very potent bacterial PAMP in Arabidopsis as well (Kunze et al., 2004). EF-Tu is one of the most abundant and most conserved proteins in bacteria. Peptides corresponding to the acetylated N

(16)

2

terminus of EF-Tu, called elf18 and elf26, triggered PAMP responses in Arabidopsis also in subnanomolar concentrations. Although EF-Tu is mostly intracellular, its release from lysis of dying bacteria during plant colonizationshould be sufficient to trigger its subnanomolar recognition(Zipfel et al., 2006).

A targeted reverse-genetic approach was used to identify a receptor kinase essential for EF-Tu perception. The T-DNA insertion mutant efr was completely insensitive to elf18/elf26, although still responded normally to flg22. Nicotiana benthamiana, a plant unable to perceive EF-Tu, acquires EF-Tu responsiveness upon transient expression of EFR (Zipfel et al., 2006). These findings directly demonstrated that EFR is the PRR for EF-Tu (Zipfel et al., 2006).

FLS2 and EFR protein structures resemble each other, and are both classified into the family of leucine-rich repeat receptor-like kinases (LRR-RLKs). The Arabidopsis genome encodes more than 200 LRR-RLKs. FLS and EFR both belong to the subfamily LRR XII. FLS2 contains an extracellular domain of 28 LRR motifs, while EFR has 21 LRRs. LRR motifs are known from crystallographic studies to form a highly ordered horseshoe-like solenoid structure (Boller and Felix, 2009).

Site-directed mutational analysis and other approached were conducted to investigate how LRRs contribute to flagellin perception. The LRRs 9 to15 was identified important for flg22 binding, but the actual binding site remains unknown (Chinchilla et al., 2006; Dunning et al., 2007; Robatzek et al., 2007). Similar to FLS2, the exact elf18 binding site is still unknown in EFR.

Interestingly, recent study has shown that FLS2 interacts with another LRR-RLK, BAK1 (BRI1-associated receptor kinase 1), immediately (less than 2 min) after

(17)

3

treatment of ligand flg22 (Chinchilla et al., 2007). bak1 mutant still exhibits normal binding to flg22, but its early and late flg22 induced responses are strongly impaired.

Besides, bak1 mutant shows reduction in early elf18-triggered response of ROS production (Chinchilla et al., 2007). Thus, BAK1 is an important PTI regulator that acts concomitantly with PRRs.

Although BAK1 belongs to LRR II family, it contains only 4 LRR motifs in its extracellular domain. It is suggested that receptor dimerization between BAK1 and FLS2 may play essential role in flagellin signaling. The current model of FLS2 and BAK1 is that, upon binding of flg22, the LRR domain of FLS2 changes conformation, allowing it to heterodimerize with BAK1 to form a tight complex. Interaction of the ectodomains of FLS2 and BAK1 also brings the intracellular kinase domains into contact, thus allowing initiation of intracellular signaling, likely by transphosphorylation events (Boller and Felix, 2009).

PAMP-Triggered Immunity against Bacteria: Downstream Signaling

In Arabidopsis, PAMPs induced responses, including early responses like ion fluxes, oxidative burst, activation of MAPK cascade, calcium-dependent protein kinases (CDPKs) cascade, ethylene biosynthesis and late responses like callose deposition (Boller and Felix, 2009; Boudsocq et al., 2010).

MAPK networks are involved in various processes in eukaryote cells, including plant defense. Activation of MAPK cascade leads to activation of WRKY-type transcription factors, key regulators of plant defenses (Eulgem and Somssich, 2007).

Following activation of WRKYs, FRK1 (FLG22-INDUCED RECEPTOR-LIKE

(18)

4

KINASE 1), one MAPK-specific target LRR-RLK, is rapidly induced to turn on downstream signaling (Asai et al., 2002).

Ca²⁺ influx is another PAMP triggered early response. It was recently shown that CDPKs play important roles as Ca²⁺ sensors in flg22 or MAMP signaling, and were quickly activated following flg22 elicitation (Boudsocq et al., 2010).

Unexpectedly, CDPKs and MAPK cascades act differentially in PAMP-mediated regulatory programs to control early genes involved in the synthesis of defense peptides and metabolites, cell wall modifications and redox signaling (Boudsocq et al., 2010).

PAMPs also induce rapid and transient production of ROS in an oxidative burst.

Reactive oxygen species can act as antibiotic agents directly, as well established in macrophages in animals, or they may contribute indirectly to defense by causing cell wall cross-linking; in addition, reactive oxygen species may act as secondary stress signals to induce various defense responses (Apel and Hirt, 2004). However, the relative position of oxidative burst in the sequence of signaling events during PTI is still unclear (Nicaise et al., 2009).

The accumulation of callose, a β-1, 3-glucan polymer, between the cell wall and the plasma membrane is a classical marker of PTI after bacterial infection or PAMPs perception (Gómez-Gómez et al., 1999), although the biological foundation of this response is still unclear.

Stomatal closure is part of a plant innate immune response to restrict bacterial invasion. PAMPs treatment also induce stomatal closure (Melotto et al., 2006).

Stomatal guard cells perception of bacteria requires FLS2 receptor, and

(19)

5

guard-cell-specific OST1 kinase (Melotto et al., 2006).

The Nudix-motif Containing Protein Family

In plant, one group of proteins has recently drawn much attention due to its involvement in disease resistance. This protein family, Nudix hydrolase protein family, has been identified for a long time with few understandings on their physiological functions.

The Nudix hydrolase protein family contains the enzyme activity of hydrolyzing nucleotide derivatives (also called nucleoside diphosphates linked to some moiety X, hence the acronym “Nudix”). They usually catalyze cleavage of a diphosphate (PPi) bond, thus permitting nucleotide moieties to be recycled.

This family is characterized by a conserved Nudix motif, GX5EX7REVXEEXGU, where U represents a bulky hydrophobic amino acid such as Ile, Leu, or Val and X is any amino acid (Ogawa et al., 2008). Although Nudix motif is common to all of these proteins, they actually act upon a wide range of substrates, including nucleotide sugars (e.g. ADP-ribose), (d)NTPs, nicotinamide nucleotides (e.g. NADH), dinucleoside polyphosphates (e.g. ApnA; n= 3-6), CoA, and mRNA caps. Therefore, Nudix hydrolases can be further divided into subfamilies according to their substrate specificities.

In Arabidopsis, at least 27 genes encoding Nudix hydrolase homologues have been identified (Ogawa et al., 2008). Two of them were already found to play important role in plant resistance against certain stresses (Ogawa et al., 2005; Ge et al., 2007).

(20)

6

AtNUDX1 is a functional homologue of E. coli MutT in Arabidopsis, its expression in E. coli mutT^- strain highly reduced the frequency of spontaneous mutations in bacteria (Ogawa et al., 2005). AtNUDX1 protein has the ability to hydrolyze 8-oxo-dGTP, a mutagenic nucleotide, and thus plays an important role in protection against oxidative DNA and RNA damage in plant cells.

AtNUDX7 catalyzes the hydrolysis of NADH. It was first identified as a stress-induced gene, and the knock-out mutant was highly resistant to bacterial pathogen, Pst DC3000 (Jambunathan and Mahalingam, 2006). One study pointed out that atnudx7 mutation leads to perturbation of cellular redox homeostasis and a higher level of NADH in pathogen-challenged leaves. They also hypothesized, this alteration of cellular antioxidant status might “prime” the cells for the amplified defense response. In general, AtNUDX7 functions as a negative regulator of basal immunity in Arabidopsis.

On the other hand, another group of AtNUDXs, Nudix Ap4A hydrolase (hereafter referred to as Ap4Aase), is less known about their physiological functions.

Ap4Aase from the narrow leafed blue lupine, Lupinus angustifolius, has been well studied on the mechanism of enzymatic catalysis (Guranowski et al., 1994).

Three putative Ap4Aases in Arabidopsis showing high identity with lupine enzyme are designated as AtNUDX25, AtNUDX26, and AtNUDX27 respectively. Among them, AtNUDX25 was shown to efficiently hydrolyze Ap4A asymmetrically into ATP and AMP (Szurmak B, 2008), while AtNUDX26 and -27 preferentially hydrolyzed Ap4A or Ap5A (Ogawa et al., 2008). AtNUDX25 also differed from AtNUDX26 and -27 in subcellular localization.

(21)

7

The enzyme substrate, Ap4A, is found in organisms from Archeae to humans and is predominantly a by-product of protein synthesis, specifically by aminoacyl-RNA synthetases (Goerlich et al., 1982). So far, the biological significance of Ap4A is not fully understood. In prokaryote, Salmonella typhimurium, Ap4A accumulates under condition of oxidation and heat-shock stress, and is thus described as “alarmones”

(Lee et al., 1983). In human cell cultures, regulation of Ap3A/Ap4A ratio was suggested to be involved in cell differentiation and apoptosis (Vartanian et al., 1997).

One recent study on animal mast cell showed that, upon immunological challenge by IgE-antigen complex, Ap4A level elevated in a MAPK-dependent fashion. This accumulation leads to dissociation of the transcription factor, and caused downstream gene activation (Yannay-Cohen et al., 2009).

Induced Resistance and Priming with the SAR Inducer, BTH

Besides the essential and well studied PAMP-triggered immunity, plants also possess inducible defense responses against pathogen attack. Upon infection by pathogen, plant develops enhanced resistance to subsequent infections by a broad spectrum of pathogens, either in the area of primary infection or in the distal, uninfected organs (Durrant and Dong, 2004).

The mechanism of induced resistance is usually linked with the ability of plants to induce cellular defense responses more rapidly and to a stronger degree in response to a much lower level of stimulus, either pathogen infection or abiotic stresses (Kohler et al., 2002). This phenomenon is also termed as “priming”.

Although the priming has been known for years as a crucial part of induced

(22)

8

resistance, the molecular mechanisms and genetic basis of priming remained largely unclear.

One well studied type of induced resistance (IR) is systemic acquired resistance (SAR), which requires the signal molecule salicylic acid (SA) and is associated with accumulation of pathogenesis-related (PR) proteins (Durrant and Dong, 2004). SAR can also be induced by exogenous application of SA or its synthetic analogs, 2,6-dichloroisonicotinic acid (INA) and benzothiadiazole (BTH) (Kohler et al., 2002).

In recent years, SAR and its potent inducer, BTH, were largely used for investigating mechanisms of cell priming.

As mentioned, a strong PR1 gene activation was observed in BTH-pretreated plants. Besides, BTH pretreatment also greatly augmented the PHENYLALANINE AMMONIA-LYASE (PAL) gene expression and callose deposition following subsequent Pst DC3000 infection, wounding with forceps, or infiltration of water into leaves (Kohler et al., 2002). PAL is a key enzyme in the phenylpropanoid pathway that leads to a variety of defense-related plant secondary metabolites such as SA, phytoalexins, and lignin-like polymers (Hahlbrock and Scheel, 1989).

Furthermore, a recent publication revealed the involvement of MAPK cascade in the process of priming (Beckers et al., 2009). Using BTH as inducer, MPK3 and MPK6 transcripts and inactive proteins accumulated within 72 h. BTH-pretreated plants showed enhanced activation of kinases activity in response to subsequent virulent Psm ES4326 infection. In mpk3 and mpk6 mutant, BTH-mediated priming for enhanced PAL and PR1 gene expression was highly attenuated (Beckers et al., 2009).

However, the BTH potentiation of PAL gene expression, callose deposition and

(23)

9

MPK3 and MPK6 activation were all absent in Arabidopsis non-expressor of PR genes 1 (npr1) mutant. NPR1 is a positive regulatory protein which induces defense gene expression and activating SAR in response of SA (Beckers et al., 2009).

Induced Resistance and Priming with BABA

Other than SA and BTH, one distinct chemical compound called β-aminobutyric acid (BABA) was also reported to be an effective inducer of resistance to bacterial pathogens (Zimmerli et al., 2000), fungal pathogens (Zimmerli et al., 2001; Flors et al., 2008), insects, and abiotic stresses like drought, salt and heat (Jakab et al., 2005;

Zimmerli et al., 2008). Interestingly, BABA is a synthetic non-protein amino acid, which does not resemble any biologically active molecule in plants, except for some amino acids.

Research on the mechanism of BABA-induced resistance (BABA-IR) in Arabidopsis has shown that it is mostly based on priming for different stress-inducible defense responses. In the case of bacterial pathogen Pst DC3000, although BABA pretreatment does not directly induce expression of PR1 (which was induced in the case of BTH), PR1 gene expression was strongly potentiated in BABA-IR against bacteria (Zimmerli et al., 2000). Hence, the signaling pathway controlling BABA-IR is partially different from that of SAR.

However, BABA was found inactive in npr1 mutant and transgenic NahG plants.

NahG transgenic plants overexpressing a salicylate hydroxylase gene (NahG) have low levels of SA and are unable to undergo SAR. Hence, BABA-IR against the bacteria Pst DC3000 resembles SAR in that it requires endogenous accumulation of

(24)

10

SA (Zimmerli et al., 2000).

On the other hand, it seems that BABA-IR to other pathogens depends on signaling pathways other than SA. For example, BABA is also protective against two necrotrophic pathogens, Alternaria brassicicola and Plectosphaerella cucumerina, while BTH has no significant effect on resistance against them (Ton and Mauch-Mani, 2004). BABA-IR against A. brassicicola and P. cucumerina was unaffected in the JA-insensitive mutant coi1-1 and the camalexin-deficient mutant pad3-1. Moreover, the expression of BABA-IR was not associated with enhanced accumulation of camalexin or enhanced transcription of the JA-inducible PDF1.2 gene. Instead, BABA-IR was blocked in the ABA-deficient mutant aba1-5, the ABA-insensitive mutant abi4-1 and the callose-deficient mutant pmr4-1. This suggests that ABA signaling and related callose deposition are involved in the regulation of BABA-induced priming against necrotrophic pathogens.

Last year in our lab, we found that BABA-IR to heat stress and Pst DC3000 was abolished by L-glutamine treatment (Wu et al., 2010). As reported, all amino acids, except for L-glutamine, cause one unfamiliar phenomenon called general amino acid inhibition, which is prevented by L-glutamine (Bonner et al., 1996; Bonner and Jensen, 1997).

Together, these studies suggest that BABA pretreatment probably induced a stressed state in plants by yet unknown mechanisms, which might be linked with activation of various downstream defense responses against different biotic or abiotic stresses.

(25)

11

Objectives

During this master, I investigated BABA-IR to the bacterial pathogen Pst DC3000 and the mechanism of BABA priming upon bacterial infection. Related researches on this topic had only focused on SA signaling and downstream PR1 expression (Zimmerli et al., 2000). To understand what might happen upstream or earlier than these events, plant innate immunity under BABA pretreatment was examined, including PTI-responsive genes expression and callose deposition. In addition, to further understand molecular basis of BABA-IR and to use BABA as a tool to identify novel genes involved in the Arabidopsis immunity, mutants screening experiments were conducted. A brief analysis of the candidate mutants is also presented.

(26)

12

Material and Methods

Plant Material and Growth Conditions

Arabidopsis thaliana plants, wild type Columbia (Col-0), sid2-1 (AT1G74710), atnudx25 (At3g10620; SALK_00093), lrr-pk-1 (AT3G21340; SALK_000388), and lrr-pk-2 (CS000992) Arabidopsis plants were grown on soil mix of soil/ perlite (3:2) with one plant per pot (6 × 6 × 5cm), in a growth chamber at 22°C day and 18°C night with 9-h photoperiod per 24 h for 5 weeks.

Screening of Homozygous Mutant Plants

T-DNA insertion mutants in this study were derived from the Arabidopsis Biological Resource Center (ABRC). To isolate homozygote of T-DNA insertion mutants, at least 10 seeds of each line were grown in long day condition for later genotype assay and seeds collection.

Genomic DNA was prepared as follow. Rosette leaves from two-week-old plants were collected from each plant and DNA was then extracted as described previously (Edwards et al., 1991). Briefly, leaf samples were extracted with 400 µl of extraction buffer (200mM Tris-HCl pH7.5, 250mM NaCl, 25mM EDTA, 0.5% SDS) and were vortexed shortly. The mixture was then centrifuged at 14,800 rpm for 1 min. 300 µl of supernatants were transferred to new 1.5-mL microcentrifuge tubes and mixed with equal volume of isopropanol. The samples were left at room temperature for 2 min and centrifuged at 14,800rpm for 5 min. The pellets were air dried at 55°C. Fifty µl of filtered water was added to dissolve the pellet. DNA samples were kept in -80°C for long term storage or 4°C for immediate genotype assay.

(27)

13

The genotype assay was done with two sets of PCR analyses. In the first set of PCR, gene-specific primers (LP and RP, designed by Salk Institute Genomic Analysis Laboratory (SIGnAL) website http://signal.salk.edu/) were used to amplify the gene fragment without T-DNA insertion. The second set of PCR using RP primer and T-DNA left border of the vector as a primer, which were to identify the presence of T-DNA insertion. The homozygous lines have insertions in both chromosomes, and won’t get any product in first PCR, but one single band in second PCR. The seeds of homozygous plants were then collected for further experiments. The primers used in this study are listed in Table 2.

Plants Treatments

Five-week-old Arabidopsis plants were used for all treatments. For BABA-induced resistance assay, BABA (Sigma-Aldrich) treatments were performed as soil drench applications at 225 µM final concentration. Control plants were treated with an equal volume of water. Pathogen infection experiments were performed 48 h after BABA treatment.

For flg22-induced resistance assay, treatments with flg22 (Biomer Technology) were performed by pressure infiltration (needle-less) of 1 µM peptide solution into leaves. Control plants were infiltrated with 10 mM MgSO4 buffer solution. Pathogen infection experiments were performed 24 h after flg22 treatment.

In the combined assay of BABA and flg22-induced resistance, plants were first treated with BABA for 48 h, and then infiltrated with flg22 for 24 h before pathogen infection experiment.

(28)

14

For gene expression analysis, flg22 at 100 nM were applied for the indicated times.

Pseudomonas syringae Infection and Bacterial Growth Assay

Pseudomonas syringae pv tomato strain DC3000 and DC3000 hrcC^- mutant were grown during 16-18 h at 28 °C in King’s B medium with 100 µg ml^-1 of rifampicin (Pst DC3000) or 100 µg ml^-1 of rifampicin and kanamycin (Pst DC3000 hrcC^-). Bacteria were then resuspended in 10 mM MgSO4 with OD600 value at 0.2 corresponding to concentration of 10⁸ cfu ml^-1. For dipping inoculation, bacterial densities were adjusted to 5 × 10⁶ c.f.u. ml^-1 or 2 × 10⁷ c.f.u. ml^-1 for bacterial growth assays or gene expression analysis, respectively, or as indicated in the figure legends.

For infiltration inoculation, bacterial densities were adjusted to 10⁵ c.f.u. ml^-1 for bacterial growth assays, gene expression analysis, and callose depositions, as indicated in the figure legends.

Bacterial growth assays were performed as previously described (Zimmerli et al., 2000) with the following modifications. Pst DC3000 bacterial suspensions were either dip- or infiltrated-inoculated into 5-week-old plants after indicated pretreatment of flg22, BABA, or mock control. Each sample consisted of nine leaf discs (5 mm diameter) from three plants (three leaves each). Leaf discs were homogenized in 1.5 ml of 10 mM MgSO4 and dilution series were made. Fifty µl of appropriate dilutions were applied on King’s B plates containing 100 µg ml^-1 of rifampicin. Tissue samples were harvested from inoculated leaves at 2, or 3 dpi for dip-inoculation or at 0, 2, or 3 dpi for infiltration-inoculation experiments, as indicated in the figure legends.

(29)

15

Botrytis cinerea Infection

Botrytis cinerea was grown on 1/2 PDA-agar plates for 14 d. Spores were washed out from plates with proper amounts of 1/2 PDB medium and quantified with a hemocytometer. Spores were then diluted to 10⁵ spores ml^-1 in 1/2 PDB medium.

Ten µl of spore suspension were inoculated on leaf surfaces of five-week-old plants (three leaves per plants). Disease symptoms were determined at 3 or 5 dpi, as indicated in the figure legends.

Erwinia carotonova Infection

Erwinia carotovora subsp. carotovora strains WPP14 was grown during 16-18 h at 28°C in Luria-Bertani (LB) medium without antibiotics. Bacteria were then resuspended in 10 mM MgSO4 with OD600 value at 0.25 corresponding to concentration of 10⁸ c.f.u. ml^-1. Bacterial densities were adjusted to 3 × 10⁵ c.f.u. ml^-1 for dipping inoculation of 5-week-old plants. Disease symptoms were determined at 2-4 dpi, as indicated in figure legends.

Callose Deposition

Five-week-old plants were injected with 10⁷ c.f.u. ml^-1 of Pst DC3000 suspension or 1 µM of flg22, and samples were collected at indicated time points.

Control plants were infiltrated with 10 mM MgSO4 buffer solution. Leaf discs (5 mm diameter) excised from infiltrated leaves were cleared overnight in 96% ethanol until leaf discs appeared white and slightly transparent. Leaf discs were washed twice with

(30)

16

water for 2 h each time. After washing, leaf discs were incubated with 70 mM Na2HPO4 (pH 8.0) solution containing 0.05% aniline blue overnight. Callose deposits were visualized using Olympus EX51 microscope under UV illumination with broadband DAPI filter set.

Pictures of random areas of leaf discs were taken and callose deposits were counted using the “analyze particles” of ImageJ (http://rsbweb.nih.gov/ij/). Nine leaf discs from different leaves of three plants were analyzed for each treatment or genotype.

RNA Extraction and Gene Expression Analysis

Total RNA was extracted from leaf tissue or whole plant with the RNeasy Plant Mini Kit (Qiagen). Contaminating DNA was removed with an RNase-free DNase Set (Qiagen), and RNA concentrations were quantified with a Nanodrop Spectrophotometer (Thermo Scientific). RNA samples were diluted to equal concentrations of 2 µM before synthesis of cDNA with M-MLV reverse transcriptase (Invitrogen).

Real-Time and semi-quantitative RT-PCRs were performed using the primers listed in Table 1. EF1α (AT5G60390) was used as the internal reference. Real-Time RT-PCR experiments were carried out using an iQ5 Real-Time PCR Detection System (Bio-Rad) and the KAPA SYBR^® FAST qPCR Kits (Kapa Biosystems), following the manufacturer’s protocol. The thermal cycling program used was an initial 95 °C for 3 min, followed by 40 cycles at 95 °C for 3 s, 60 °C for 30 s. Melting curve was run from 65 °C to 95 °C with 10 s time intervals to ensure the specificity of the product.

(31)

17

Data were analyzed using the Bio-Rad iQ5 software (version 2.0).

Mesophyll Protoplast Transient Expression Assay

Protoplasts were prepared essentially as described previously (Yoo et al., 2007).

Leaves from 5-week-old plants were cut into strips and transferred immediately into fresh prepared enzyme solution (20 mM MES (pH 5.7) containing 1.5% cellulose R10 (Yakult Pharmaceutical Ind. Co.), 0.3% mecerozyme R10 (Yakult Pharmaceutical Ind.

Co.), 0.4 M mannitol, 20 mM KCl, 10 mM CaCl2 and 0.1% BSA). Leaf strips were incubated for 2-2.5 h. Protoplasts were collected carefully into 5-ml centrifuge tubes and centrifuged at 100 g for 1 min at room temperature. Supernatant was removed and 3-4 ml of W5 solution (2 mM MES (pH 5.7) containing 154 mM NaCl, 125 mM CaCl2 and 5 mM KCl) was added to wash the protoplasts. Centrifugation was repeated once and protoplasts were resuspended in 2-5 ml W5 solution, and then kept on ice for 30 min (or 4 °C overnight). After protoplasts were settled at bottom of tube by gravity, W5 solution was removed and resuspended with equal volume of MMG solution (4 mM MES (pH 5.7) containing 0.4 M mannitol and 15 mM MgCl2). DNA plasmid and 100 µl protoplasts were added into a 2-ml centrifuge tube and mixed with equal volume of PEG by gently tapping the tube. Mixture was incubated at room temperature for 15 min, and then diluted with 440 µl W5 solution, and centrifuged at 100 g for 2 min at room temperature. Supernatant was removed and added with 100 µl of W5 solution. Protoplasts was incubated overnight and kept under light at room temperature. The fluorescence was observed with a TCS SP5 confocal fluorescence microscope (Leica).

(32)

18

DNA Constructs

The vector for the expression of self-complementary RNA (termed hairpin) targeted to PR2 (AT3G57260) were constructed using Gateway cloning technology (Invitrogen). 393 bp of cDNA fragment encoding partial open reading frame of PR2 was chosen following the rule that blocks of sequence identity with non-target gene of over 20 bases should be avoided. This short fragment were subcloned into the donor vector, pDONR221, and then recloned into the destination vector, pB7GWIWG2(I), in which inverted repeat of PR2 gene fragments separated by a spacer sequence was expressed as hairpin structure under the control of the cauliflower mosaic virus 35S promoter. The PR2 fragment sequence and specific primers for PCR amplification are listed in Table 3. PCR and in vitro BP and LR recombination reactions were carried out according to the manufacturer’s instructions (Invitrogen).

The DNA plasmids which express LRR-PK protein fused with GFP (S3G21340HGF) or Flag epitope tag (S3G21340BFF) at their C termini under the control of the cauliflower mosaic virus 35S promoter were obtained from the Arabidopsis Biological Resource Center (ABRC) and were as described (Gou et al., 2010).

Agrobacterium tumefaciens (strain GV31), which was transformed with constructs obtained by electroporation, was used for the transformation of Col-0 wild-type plants (with LRC1, S3G21340HGF, S3G21340BFF), sid2 (with LRC1), and lrr-pk-2 (with S3G21340HGF, S3G21340BFF) mutant lines. The seeds with successful transformation were screened on 1.5% MS agar plates containing 50 µM

(33)

19

glufosinate-ammonium (Fluka). T2 plants were transferred onto soil for bacterial growth assay and seed collection.

(34)

20

Results

BABA-Induced Resistance to Pst DC3000 in the SA Biosynthetic Mutant sid2 As suggested by Zimmerli et al., 2000, BABA-induced resistance to Pst DC3000 essentially depends on potentiation of the SA-signaling. SID2 (SALICYLIC ACID INDUCTION DEFICIENT 2) encodes an isochorismate synthase required for SA synthesis. It has been reported that PR1 expression in sid2 is highly reduced after infection with Pst DC3000 avirulent strain (Nawrath and Metraux, 1999).

Interestingly, when sid2 was inoculated with low concentration (5×10⁶ c.f.u. ml^-1) of Pst DC3000 with dipping method, BABA still provided protection in the mutant (Figure 1A). In bacterial growth assay, BABA pretreated sid2 demonstrated a 2-7 times reduction in bacterial growth (Figure 1B), while in wild-type Col-0, the BABA-pretreated plants demonstrated 8-10 times reduced bacterial growth (Figure 1B). We also examined bacterial growth by syringe infiltration of 10⁵ c.f.u. ml^-1 Pst DC3000. In two biologically independent replicates, although BABA-pretreated wild-type Col-0 still exhibited less bacterial growth, BABA-pretreated sid2 demonstrated a similar bacterial growth compared to water-pretreated controls (Figure 1C). Therefore, the difference of bacterial growth in sid2 mutant caused by BABA pretreatment was predominately observed when bacteria were inoculated with the dipping method.

BABA Pretreatment Alters Gene Expression during Plant Innate Immunity To investigate BABA effects on plant innate immunity and its relationship with

(35)

21

SA signaling, we examined the expression levels of several innate immunity marker genes in wild type and sid2 mutant. The marker genes chosen included genes specific for the MAPK cascade, FRK1 and WRKY53 (Asai et al., 2002), genes reflecting activation of CDPK cascade, CYP81F2 and NHL10 (Boudsocq et al., 2010), and gene upstream of SA production and signaling, CBP60g (Wang et al., 2009). Basal gene expression levels were analyzed 48 h after BABA pretreatment. In wild type and sid2, most of the five selected genes were induced by BABA pretreatment (Figure 2A).

Expression levels of these genes were examined following Pst DC3000 hrcC or flg22 treatment. Pst DC3000 hrcC lacks functional type three secretion system (TTSS) to overcome plant innate immunity, and largely induces PAMP-triggered immunity during infection (Tsuda et al., 2008). In these experiments, buffer infiltrated samples were used as controls. Following either Pst DC3000 hrcC or flg22 treatment, all the five genes demonstrated higher expression levels in BABA pretreated plants (Figure 2B to G). BABA pretreated sid2 demonstrated similar expression pattern as Col-0 wild type.

Hence, basal and induced levels of plant innate immunity marker genes were both positively regulated by BABA pretreatment. And this regulation was independent of SA signaling.

BABA Pretreatment Alters Callose Deposition During Pst DC3000 Infection Callose deposition, another classical marker of plant innate immunity, was also analyzed following inoculation with Pst DC3000. Callose accumulated in bacteria-inoculated plants but not in buffer-infiltrated mocks (Figure 3A and B). At 6

(36)

22

hpi and 9 hpi with bacteria, water pretreated sid2 always exhibited less callose deposition than water pretreated wild type (Figure 3A and B). Notably BABA pretreated wild type and sid2 both showed higher callose deposition than water pretreated controls (Figure 3A and B).

Beside difference in density of callose deposits, morphologically distinct type of callose deposition was also observed in BABA pretreated plants (Figure 3C, D and E).

BABA pretreated plants contained normal and big callose deposits. To analyze this phenomenon, diameter of callose was analyzed using 20 µm as criterion to distinguish between small and big callose, as described by (Ham et al., 2007). Deposits of callose greater than 20 µm in diameter are classified as big, those less than 20 µm are classified as small. Big callose accumulated only in BABA pretreated plants, and was most prominent at 9 hpi with bacteria, but the diameter of big callose never exceeded 40 µm (Figure 3C, D and E).

MYB51 encodes a transcription factor downstream of ethylene (ET) signaling and is induced in response to PAMPs treatment (Clay et al., 2009). MYB51 is usually considered as a marker gene for callose deposition (Clay et al., 2009; Millet et al., 2010). Interestingly, higher expression was observed in BABA pretreated wild type and sid2 after either flg22 or Pst DC3000 hrcC inoculation (Figure 3F and G).

Hence, enhanced callose accumulation with BABA pretreatment was related to BABA priming of plant innate immunity marker gene.

BABA Pretreatment Does Not Change Flg22-Induced Resistance

Flg22 pretreatment can induce resistance to subsequent bacterial infection, a

(37)

23

phenomenon very similar to BABA-IR. In previous result, we observed that BABA positively regulates flg22 or Pst DC3000 hrcC induced gene expression and callose deposition. Hence BABA pretreatment may also demonstrate some effects on flg22-IR. Therefore, plants were pretreated with BABA and then treated with flg22 and subsequently inoculated with Pst DC3000. At 4 dpi with bacteria, plants with flg22 or BABA or both pretreatment were less infected than non-pretreated plants (Figure 4A). But disease severity among the three pretreatment conditions was nearly identical. To more precisely examine the symptom, bacterial growth at 2 dpi was assessed. Compared to BABA pretreatment, flg22 pretreatment more efficiently reduced bacterial growth (Figure 4B). And BABA pretreatment did not induce more flg22 protection (Figure 4B). This result suggests that signaling events of BABA-IR and flg22-IR may overlap with each other.

Generation of PR2 Gene Silencing Transgenic Line in sid2 Background

As reported, PR1 expression levels are strongly reduced in sid2 after pathogen attack, while PR2 expression levels remain normal (Figure 5) (Nawrath and Metraux, 1999). The BABA-IR observed in sid2 may thus be due to BABA up-regulation of PR2, which in turn slightly increase resistance to pathogen infection. To test this hypothesis, we reduced PR2 expression in sid2 by using the gene silencing approach.

The Gateway destination vector pB7GWIWG2(I) was used to design gene silencing constructs (Figure 6A). Over-expression of PR2 inverted repeat separated by intron spacer was cloned into pB7GWIWG2(I), and named as LRC1 (Figure 6A).

Restriction enzyme EcoRI and XbaI were used to examine insertion of PR2 fragments.

(38)

24

In LRC1, both recognition sites of EcoRI were replaced by PR2 fragments, hence the plasmid was not cleaved by enzyme (Figure 6B). On the other hand, XbaI digested product of LRC1 demonstrated reduced size as predicted (Figure 6B). Furthermore, insertion of PR2 fragments was examined by two sets of PCR analysis. In first set of PCR, 35S promoter and PR2 primers were used to amplify first PR2 fragment (Figure 6C). The second set of PCR used PR2 and 35S terminator primers to identify the

reversely inserted PR2 fragment (Figure 6C). Twenty-six putative T1 lines were selected for segregation analysis. Among them, 21 T2 lines had survival rates more than 70 % on MS medium containing BASTA, which were either heterozygous or homozygous for PR2 silencing construct. These lines were then assayed for their PR2 expression in response to Pst DC3000 infection. At least 5 lines (line number 1, 7, 8, 10, 25) demonstrated reduced PR2 expression (Figure 7A). The transgenic lines possess more widely expanded leaves and shorter petioles (Figure 7B).

Bacterial growth was assayed to determine whether PR2 plays an important role in the BABA-IR observed in the sid2 mutant. Preliminary tests indicated that most of the BABA-pretreated transgenic lines demonstrated a response similar to sid2 (Figure 8A and 8B). This preliminary observation suggests that BABA-IR in sid2 can function independently of PR2.

Knock-out Mutant Lines Screening

To evaluate the effects of BABA on Arabidopsis, gene expression was monitored 24 h after BABA treatment with the microarray approach (Zimmerli et al., 2008).

Among the genes with altered expression levels, 678 genes were found to be

(39)

25

up-regulated in response to BABA, whereas only 83 genes were repressed (Zimmerli et al., 2008). To understand whether these candidate genes might be involved in BABA-IR against P. syringae or other pathogens, and also to investigate their roles in the plant innate immunity, a group of genes encoding receptor like kinases and enzymes were selected from the microarray data, and their T-DNA mutant lines were ordered from ABRC for further studies. Bacterial infection symptoms were observed 3 days after infection. The disease rates of wild-type and mutant plants were compared. Mutant sensitivities towards BABA pretreatment were also examined.

With massive screening of near 70 mutant lines, two T-DNA mutants demonstrated interesting phenotype. The first is a knock-out mutant line of AtNUDX25 gene; the second interesting line is a knocked out of a putative leucine-rich repeat protein kinase (LRR-PK). Both mutant lines were described in detail in the following sections.

Sequence Analysis of AtNUDX25 and Identification of atnudx25 mutant

AtNUDX25 consists of five exons. The full length CDS contains 528 bp which encodes a 175-residue peptide. The conserved Nudix motif (residues 40 to 62) is shown in Figure 9A and together compared with the sequence of the well- characterized plant Ap4Aase from Lupinus angustifolius. A Tyr residue 15 amino acids downstream of the Nudix box is considered responsible for the preference of Ap4A as enzyme substrate (Dunn et al., 1999) (Figure 9A).

T-DNA mutant lines of AtNUDX25 were ordered from ABRC. T-DNA insertion sites are as indicated (Figure 9B). SALK_00093, has a T-DNA insertion in its third

(40)

26

intron at 240 bp 3’ to the initial codon. Another T-DNA mutant (SALK_078741) has T-DNA insertion in 5’ UTR. Knock-out of AtNUDX25 was examined by RT-PCR.

AtNUDX25 transcript was knocked out in SALK_00093 (hereafter referred to as atnudx25), but not in SALK_078741 (Figure 9C). Therefore, for the following experiments, atnudx25 was used as the only loss-of-function mutant of AtNUDX25.

Mutation in AtNUDX25 Result in Enhanced Susceptibility to Necrotrophic Pathogens

atnudx25 were more severely infected by two necrotrophic pathogens, the fungal pathogen B. cinerea, and the bacterial pathogen, Ecc WPP14 (Figure 10A and 11).

The lesion size of B. cinerea infected leaves was determined 3 days after inoculation.

The diameter was larger in atnudx25 mutant than wild type (Figure 10B).

Both of the pathogens kill host cells at very early stages in the infection and cause extensive tissue damage. They induce jasmonates (JA)-dependent defense responses in plant (Glazebrook, 2005), therefore we analyzed the marker gene for JA signaling, PDF1.2, following pathogen inoculation. The preliminary studies indicated that the basal expression of PDF1.2 (0 hpi) was higher in atnudx25 (Figure 10C). But after inoculations of B. cinerea, PDF1.2 expression was normally induced (Figure 10C). Therefore, although enhanced susceptibility was observed, reduction in JA downstream marker gene, PDF1.2, expression was not detected in our test. This preliminary result suggests that AtNUDX25 functions independently of JA signaling.

(41)

27

Mutation in AtNUDX25 Does Not Affect Response to flg22 and Pst DC3000

According to the web database, BAR (The Bio-Array Resource for Plant Functional Genomics) (http://bar.utoronto.ca/), AtNUDX25 is induced in response to bacterial PAMP, flg22. At 1hpi with flg22, AtNUDX25 was highly induced in wild type, but not in its loss-of-function mutant (Figure 12B). To test whether AtNUDX25 is important for perception of flg22, expression of the PTI marker gene FRK1, following flg22 treatment was analyzed. But no significant differences in the gene expression were observed between wild type and mutant (Figure 12C).

Sensitivity of atnudx25 mutant to hemi-biotrophic bacteria such as Pst DC3000 was also assessed. Pst DC3000 contains flg22 PAMPs in their flagella. The disease rates of wild type and atnudx25 were quite similar (Figure 12A). These results suggested, while AtNUDX25 responses quickly upon perception of flg22, it might not be involved in plant resistance to Pst DC3000.

The Putative LRR-PK organization and Identification of lrr-pk mutants

The putative LRR-PK consists of twelve exons. The full length CDS contains 2685 bp which encodes an 894-residue peptide. As predicted by the Pfam protein families database (http://pfam.sanger.ac.uk/) (Finn et al., 2010), the N terminus contains a secretion signal peptide (residues 1 to 23) followed by an extracellular domain of unknown function and 2 LRR motifs (residues 409 to 455). A transmembrane domain (residues 512 to 536) is predicted to separate the extracellular domain from intracellular Ser/ Thr kinase domain (residues 587 to 854) (Figure 13A).

T-DNA insertion mutant lines were ordered from ABRC. The insertion sites are

(42)

28

as indicated (Figure 13B). Both lrr-pk-1 (Salk_000388) and lrr-pk-2 (CS000992) have T-DNA insertions in regions encoding the kinase domain. RT-PCR was performed to confirm that the T-DNA and transposon insertion inhibited the accumulation of putative LRR-PK transcript. Both lrr-pk-1 and lrr-pk-2 mutant lines were knocked out of LRR-PK (Figure 13C), and were used in the following experiments.

Mutation in the Putative LRR-PK Causes Partial BABA Insensitivity and Enhanced Susceptibility to Pst DC3000.

We investigated the possible role of the LRR-PK under study by testing the phenotype of lrr-pk-1 and lrr-pk-2. At 3 day post dip-inoculation with Pst DC3000 both of the mutant lines showed more yellowing and necrosis than wild-type Col-0 (Figure 14A). In addition, BABA pre-treatment on mutant lines did not provide significant protection against bacterial infection (Figure 14A). In addition, both lrr-pk-1 and lrr-pk-2 demonstrated higher bacterial growth than wild type at 2 dpi, (Figure 14B). Interestingly, BABA pretreatment on both mutant lines induced less protection than on Col-0 wild type (Figure 14B). This result suggested that the two mutant lines still respond to BABA pretreatment, but the sensitivity to BABA is reduced.

To understand whether lrr-pk mutant lines were defected in early stage of plant innate immunity, bacterial growth following infiltration inoculation of Pst DC3000 was assessed. The preliminary result indicated that, when Pst DC3000 was directly infiltrated into leaves, lrr-pk-1 mutant and wild type had similar sensitivity to bacteria (Figure 14C). This suggests that the putative LRR-PK may function during early

(43)

29

stages of Arabidopsis innate immunity.

The Putative LRR-PK is Induced in Response to Bacterial Infection and PAMP Perception.

To investigate the role of the putative LRR-PK in plant innate immunity, its gene expression in response to bacterial infection and flg22 perception was analyzed. At 3 hpi with bacteria, the putative LRR-PK was highly induced by normal Pst DC3000 and also TTSS mutated Pst DC3000 hrcC^- (Figure 15). Besides, LRR-PK was quickly induced by flg22 treatment at 1 hpi (Figure 15). These results suggest that the putative LRR-PK has immediate response to bacterial invasion and PAMPs perception as well.

The Putative LRR-PK Localizes to the Plasma Membrane

Protein subcellular localization also may provide information to gene functions.

As mentioned, the predicted protein structure of this putative LRR-PK contains one secretion signal peptide targeted to the membrane system and also one transmembrane domain (Figure 13A), therefore we wondered whether this protein really localizes on the plasma membrane in plant cells, which is one feature for many important LRR-PKs involved in plant innate immunity signaling (e.g. FLS2, EFR, BAK1) (Boller and Felix, 2009). The DNA construct, S3G21340HGF, containing the putative LRR-PK CDS fused with a C-terminal GFP, controlled by 35S promoter, was used to transformed Arabidopsis wild-type protoplast.

Under confocal microscope, a clear GFP signal lying on borderline of protoplast was observed, like an annular eclipse (Figure 16). This result suggests that the

(44)

30

putative LRR-PK localizes on the plasma membrane in vivo, and further suggests that this protein may function in early signaling of plant innate immunity.

Mutation in the Putative LRR-PK Reduces Callose Deposition During Pst DC3000 Infection and Flg22 Perception

Callose deposition, another classical marker of plant innate immunity, was also analyzed following inoculation of Pst DC3000 or flg22. Buffer infiltrated samples were used as mock controls. At 9 hpi with Pst DC3000, callose deposition was strongly reduced in both lrr-pk-1 and lrr-pk-2 mutants (Figure 17A and B). Besides, at 12 hpi with flg22, both mutant lines demonstrated reduced callose accumulation (Figure 17A and C). These results suggest that the putative LRR-PK may function in PTI signaling pathway relating to callose deposition.

(45)

31

Discussion

BABA Primes Early Stage of Plant Innate Immunity through Potentiation of Multiple Signaling Pathways

We observed a slightly reduced bacterial growth in BABA pretreated sid2 mutant when plants were infected by dipping but not by infiltration method (Figure 1).

Infiltration of bacteria with syringe might bypass the first steps of the natural infection process, like entering the leaves through natural opening such as stomata (Zipfel et al., 2004). Therefore, we hypothesized that BABA-induced resistance may not only primes the SA signaling but also early steps of plant innate immunity, the so called PTI. It seems that PTI becomes prominent when effects of BABA on SA-signaling were removed, as in the case of sid2 mutant.

How the early steps of plant innate immunity or PTI interact with downstream SA signaling is one noticeable issue regarding plant defense responses. It is suggested that during PAMPs perception, SA-independent signaling is firstly activated at early stage, which leads to SA accumulation. In turn, plant exploits SA-mediated signaling to maintain PTI (Tsuda et al., 2008). Investigating how BABA primes PTI and probably in turn activating SA signaling may provide new knowledge to this issue.

(A) BABA Primes PTI-Responsive Genes expression

In this study, expression of PTI marker genes was first analyzed, and with BABA pretreatment, 6 PTI-responsive genes demonstrated higher expression levels independently of SA signaling (Figure 2 and 3). SA dependency of these genes is

(46)

32

compared with one previous study which identified several clusters of PAMPs induced genes (PIGs) with characteristic regulatory patterns (Tsuda et al., 2008).

Some BABA-primed genes such as FRK1, WRKY53, CYP81F2 and CBP60g were also reported in this study (Tsuda et al., 2008).

FRK1, WRKY53 and CBP60g are activated independently of SA signaling at early stage of infection, but need SA accumulation to maintain their late expression (Tsuda et al., 2008). Hence, it is reasonable that BABA primes these genes independently of SA signaling at 1 or 3 hpi with flg22 or bacteria. And according to Tsuda et al., expression of CYP81F2 during infection is totally independent of SA, as observed in our study. Besides, PAMPs induce ethylene (ET) production (Felix et al., 1999), Tsuda et al. also suggested that a cluster of PAMPs induced genes are regulated by JA/ ET signaling. Previously, BABA was shown to enhance mRNA accumulation of ET early signaling intermediates (Zimmerli et al., 2008). In our study, BABA also primed expression of MYB51, the transcription factor downstream of ET signaling (Figure 3F and G). Therefore, BABA pretreatment alters expression of various PTI genes in different signaling pathways. Especially, CBP60g primed by BABA is an important component upstream of SA accumulation (Wang et al., 2009), and BABA may prime SA signaling via priming of this PTI gene. In addition, BABA may also manipulate early stage of plant innate immunity through ET signaling, as MYB51 was also primed by BABA.

(B) BABA Primes Callose Deposition in a SA-independent manner

Other than gene expression, BABA pretreated plants demonstrated enhanced

(47)

33

callose deposition in the sid2 mutant, thus independently of SA signaling (Figure 3).

The early signaling network for callose deposition functions independently of SA.

PAMPs or flg22-induced callose can potentiate SA-responsive callose deposition (Clay et al., 2009; Adams-Phillips et al., 2010). Notably, MYB51 and CYP81F2, the genes representing two independent pathways required for flg22-induced callose deposition (Clay et al., 2009), were both primed by BABA. These results thus suggest that BABA enhanced callose deposition is related with MYB51/ET-dependent and CYP81F2/I3G-dependent portions of callose deposition network. Hence, BABA may prime PAMPs or flg22-induced callose deposition and in turn potentiate SA-dependent callose deposition.

On the other hand, BABA pretreated plants frequently demonstrated bigger callose deposits with diameter larger than 20 µm (Figure 3). This morphologically distinct type of callose was also recorded in plant infected with non-host bacterial pathogen, Pseudomonas syringae pv. phaseolicola (Pph) (Ham et al., 2007). Diameter of Pph induced big callose ranged from 20 µm to 50 µm. The biological foundation of this phenomenon is not clear, but seems to be independent of SA signaling, as big callose deposition was observed even in SA mutant, sid2 and npr1 (Ham et al., 2007).

These results suggest that BABA-IR may participate in plant non-host resistance, although the biological mechanism remains elusive.

In summary, BABA primes many flg22 or PAMPs induced defense responses, including gene expression and callose deposition, which are SA-independent in early stage of bacterial infection. Hence, we propose that BABA up-regulates plant defense responses through priming of PTI, and in turn induces SA signaling to continuously

(48)

34

activate plant defense system.

(C) BABA and flg22 Induced Resistance to Bacteria

Although we hypothesized that BABA primes flg22 induced defense responses, however in our preliminary results, BABA pretreatment did not induce more flg22 protection (Figure 4). There are two possible explanations. First, flg22 pretreatment more efficiently reduced bacterial growth than BABA (Figure 4B), and thus BABA-IR is overwhelmed by flg22-IR. Secondly, BABA- and flg22- responsive signaling may be highly overlapped. For example, flg22 induces the production of ET (Felix et al., 1999), and flg22-IR does not strictly depend on SA (Zipfel et al., 2004).

Therefore BABA does not provide more protection in flg22-pretreated plants.

AtNUDX25 Regulates Plant Resistance to Necrotrophic Pathogen

Plant responses to necrotrophic pathogens appear to be mediated by a complex web of signaling dominated by JA and ET- signaling pathway (Glazebrook, 2005). In our study, although atnudx25 mutant demonstrated enhanced susceptibility to necrotrophic pathogens, B. cinerea and Ecc WPP14, expression of PDF1.2 expression was similar between atnudx25 and wild type (Figure 10). This suggests that defect in other signaling pathway leads to more diseased phenotype in atnudx25. On the other hand, flg22 highly induced expression of AtNUDX25 (Figure 12). However, whether atnudx25 mutant is impaired in flg22 induced responses is yet undetermined. In the future, the relationship between flg22 induced responses and enhanced susceptibility to necrotrophic pathogen in atnudx25 should be analyzed.