Identification of Aspergillus Fungal Resistance Factors in a Plant Model System

Western Washington University Western CEDAR

WWU Graduate School Collection WWU Graduate and Undergraduate Scholarship

Fall 2016

Teresa C. De Sitter Western Washington University, [email protected]

Follow this and additional works at: https://cedar.wwu.edu/wwuet

Part of the Biology and Biomimetic Materials Commons

Recommended Citation De Sitter, Teresa C., "Identification of Aspergillus Fungal Resistance Factors in a Plant Model System" (2016). WWU Graduate School Collection. 548. https://cedar.wwu.edu/wwuet/548

This Masters Thesis is brought to you for free and open access by the WWU Graduate and Undergraduate Scholarship at Western CEDAR. It has been accepted for inclusion in WWU Graduate School Collection by an authorized administrator of Western CEDAR. For more information, please contact [email protected].

IDENTIFICATION OF ASPERGILLUS FUNGAL RESISTANCE FACTORS IN A PLANT MODEL SYSTEM

Teresa C. De Sitter

Accepted in Partial Completion of the Requirements for the Degree Master of Science

Kathleen L. Kitto, Dean of the Graduate School

ADVISORY COMMITTEE

Co-chair, Dr. Jeff C. Young

Co-chair, Dr. Marion Brodhagen

Dr. Gregory O’Neil

MASTER’S THESIS

In presenting this thesis in partial fulfillment of the requirements for a master’s degree at Western Washington University, I grant to Western Washington University the non- exclusive royalty-free right to archive, reproduce, distribute, and display the thesis in any and all forms, including electronic format, via any digital library mechanisms maintained by WWU.

I represent and warrant this is my original work, and does not infringe or violate any rights of others. I warrant that I have obtained written permissions from the owner of any third party copyrighted material included in these files.

I acknowledge that I retain ownership rights to the copyright of this work, including but not limited to the right to use all or part of this work in future works, such as articles or books.

Library users are granted permission for individual, research and non-commercial reproduction of this work for educational purposes only. Any further digital posting of this document requires specific permission from the author.

Any copying or publication of this thesis for commercial purposes, or for financial gain, is not allowed without my written permission.

Teresa C. De Sitter December 20, 2016

IDENTIFICATION OF ASPERGILLUS FUNGAL RESISTANCE FACTORS IN A PLANT MODEL SYSTEM

A Thesis Presented to The Faculty of Western Washington University

In Partial Fulfillment Of the Requirements for the Degree Master of Science

Teresa C. De Sitter December 2016

ABSTRACT

Aspergillus flavus is a saprophytic, mycotoxigenic fungus that contaminates agriculturally important seeds with the potently toxic and carcinogenic secondary metabolite, aflatoxin. Seed infection by fungi is often prevented by intact seed coats. Although

Arabidopsis thaliana is naturally resistant to Aspergillus infection, certain mutants in the flavonoid biosynthetic pathway are compromised in seed coat integrity. We hypothesized that these mutants might also permit Aspergillus infection. To that end, we systematically tested infectibility of mutants in the flavonoid biosynthetic pathway to identify those lacking resistance to Aspergillus fungal infection. Susceptible seeds included those mutated in the genes encoding for synthesis of the first flavonoid pathway precursor, chalcone, through leucocyanidin (CHS, F3’H, and DFR), indicating that the requisite compound is either leucocyanidin or a derivative of that compound. While preliminary observations suggested that older chs seeds might be more susceptible to A. nidulans than younger seeds, an experiment testing infectibility of seeds harvested at specific ages failed to reproduce the infection rates previously observed. Further investigation revealed that chs seed batches dominated by non-viable seeds are more infectible, as expected from a saprophytic fungus. A novel finding was that chs seeds formed during the final weeks of the parent plant’s development are more highly susceptible to A. nidulans. Our results suggest that wildtype

Arabidopsis seeds have a barrier to infection, which may be either mechanical or chemical.

ACKNOWLEDGMENTS

It is with appreciation that I thank my co-advisors, Dr. Jeff C. Young and Dr. Marion

Brodhagen, for all their support over the past three years in and out of the lab. Thank you to my third committee member, Dr. Greg O’Neil, for helping answer all of my organic chemistry questions. I would also like to thank everyone in the WWU Biology stockroom, especially Peter Thut, for their endless help in acquiring supplies, helping with safety protocols, and for their general good cheer. Thank you to Mary Ann Merrill for answering all the questions I’ve ever had about funding, registration, and graduate school nuances. An acknowledgment is given to Ryan Doucette for all his work on the protocol development for the seed infectibility assay, and for his work in completing the genetic dissection of the flavonoid biosynthetic pathway. Thank you to all the undergraduate assistants that have made this work possible by helping with the never-ending spore counting and plant care. Thank you to my family, friends, and fellow graduate students for all of their support throughout these past few years. Finally, a thank you to Western Washington University RSP for research funding, and to the Fraser Endowment for both a research grant and for two summer fellowships.

TABLE OF CONTENTS

ABSTRACT ...... iv

ACKNOWLEDGMENTS ...... v

LIST OF FIGURES ...... vii

LIST OF TABLES ...... viii

INTRODUCTION ...... 1-13

METHODS ...... 14-19

RESULTS ...... 20-23

DISCUSSION ...... 24-32

CONCLUSION ...... 33-34

REFERENCES ...... 35-43

FIGURES AND TABLES ...... 44-65

LIST OF FIGURES

Figure 1. The five cell layers on an Arabidopsis thaliana seed coat ...... 44

Figure 2. The flavan backbone and classes of flavonoids ...... 45

Figure 3. The flavonoid biosynthetic pathway in Arabidopsis ...... 47-48

Figure 4. Levels of expression of the seventeen laccase genes in Arabidopsis ...... 49

Figure 5. Molecular changes made by flavonoid modifying enzymes ...... 50

Figure 6. Ages of siliques along the branches of an Arabidopsis plant ...... 51

Figure 7. A two-model pathosystem: Arabidopsis thaliana and Aspergillus nidulans ...... 52

Figure 8. Infection of Arabidopsis seeds mutant in flavonoid pathway genes ...... 53

Figure 9. Rates of A. nidulans infection of laccase mutant seeds (lac15 and lac5) ...... 54

Figure 10. Rates of A. nidulans infection of chs mutant seeds of different ages ...... 55-56

Figure 11. Timecourse of A. nidulans infection of Arabidopsis Wt Col and chs seeds ...... 57

Figure 12. Comparison of chs seeds with different plant-care methods ...... 58-59

Figure 13. Comparison of chs seeds formed at different developmental stages ...... 60-61

Figure 14. Comparison of chs seeds with different genetic backgrounds ...... 62

Figure 15. Inhibition of A. nidulans in the presence of Wt Col aqueous seed extract ...... 63

Figure 16. Variability of chs seed infection, germination, and chemical inhibition ...... 64-65

vii

LIST OF TABLES

Table 1. Arabidopsis flavonoid biosynthetic pathway proteins, genes, and seed lines ...... 46

viii INTRODUCTION

Aspergillus flavus infects crops:

Aspergillus flavus is a ubiquitous, filamentous, saprophytic fungus that infects oil- containing seeds, including maize, peanut, and cottonseed (reviewed by Amaike and Keller,

2011). The fungus produces many secondary metabolites, including aflatoxins (reviewed by

Klich, 2007), which are the most carcinogenic naturally occurring molecules known

(reviewed by Amaike and Keller, 2011). Low-level, long-term ingestion of aflatoxins causes numerous chronic medical problems ranging from immunosuppression to cancer; ingestion of high levels of aflatoxins is lethal (reviewed by Yu et al., 2005; Shephard, 2008). A. flavus thrives in the tropical and subtropical regions of the world with temperatures between 25-

42˚C, and it proliferates in more temperate regions during times of excess heat and drought

(reviewed by Amaike and Keller, 2011; reviewed by Payne, 1998).

A. flavus colonization and aflatoxin contamination can be initiated in the field prior to harvest, mainly via cracked and damaged seeds, while subsequent fungal growth is often enhanced during post-harvest crop storage due to environmental storage conditions. Drought can cause cracking of seed coats, which in turn permits ingress of A. flavus germinating spores and hyphae into embryonic tissues (Keller et al., 1994; Cotty and Jaime-Garcia,

2007). Wounds caused by insect boring can also provide an entrance point for A. flavus hyphae (for example, see Sétamou et al., 1997). A. flavus can also colonize maize kernels via external growth of fungal hyphae down the length of the silk (Marsh and Payne, 1984; reviewed by Payne, 1998). Cottonseeds can be infected when the fungus gains ingress via the nectaries and grows into the developing boll (Klich et al., 1984; Klich and Chmielewski,

1985). In peanut, whose seeds grow below ground, floral invasion is not a significant source

of preharvest A. flavus contamination. Instead, infection occurs primarily through direct penetrance of the pod (Cole et al., 1986). Post-harvest A. flavus infection, which is the primary cause of aflatoxin contamination worldwide, often results from improper storage practices. If cracked and wounded seeds are infected by A. flavus prior to harvest, the fungus will spread rapidly in storage facilities with higher moisture levels and warm temperatures

(Cotty and Jaime-Garcia, 2007; Magan et al., 2003; Giorni et al., 2008; reviewed by Klich,

2007).

Current management methods are not sufficient for preventing A. flavus infection and aflatoxin contamination:

Efforts to manage A. flavus invasion and aflatoxin contamination of crops can target any of the three components involved in fungal infection: pathogen, environment, or host.

Pathogen: The A. flavus life cycle is not conducive to elimination of the ubiquitous soil pathogen. A. flavus forms sclerotia, which can survive for several years buried in the soil without a host (Wicklow et al., 1993). Under favorable conditions the sclerotia germinate to form mycelia that then form conidiophores. The powdery conidia are either dispersed by wind or insects and infect aboveground seeds, or are dispersed via rain and soil movement and infect belowground seeds (Cole et al., 1986; reviewed by Amaike and Keller, 2011).

Since A. flavus resists most registered fungicides, an emerging method for prevention of aflatoxin contamination is inoculating fields with non-aflatoxigenic strains of Aspergillus, thereby introducing a near-isogenic competitor into the niche that is usually filled by aflatoxin-producing A. flavus strains. This method must be implemented with caution since the non-aflatoxigenic strains could potentially mutate or sexually recombine to revert to toxin-forming strains. Additionally, this method is controversial because even strains that do

not produce aflatoxins can cause ill effects in humans, such as allergic reactions to spores, aspergillosis (recalcitrant fungal infections usually from inhalation or introduction via wounds) and toxicity from ingestion of other mycotoxins (reviewed by Klich, 2007).

Environment: Current disease management practices focus on creating pre- and post- harvest environmental conditions that are not favorable to A. flavus infection and aflatoxin contamination. Preventative methods begin with controlled cultural practices in the field, including sufficient irrigation to prevent seed cracking, control of insect pests to prevent seed wounding, and immediate harvesting after the crop reaches maturity to prevent fungal growth under uncontrolled environmental conditions of the field. In storage facilities, environmental conditions such as moisture levels, temperature, and atmospheric composition must be controlled to prevent post-harvest infection. A. flavus can grow and produce aflatoxin in moisture contents as low as 8-12% and 17-19%, respectively (reviewed by Sanchis and

Magan, 2004). Most storage facilities can only dehumidify to 14% moisture content, which reduces aflatoxin production but does not fully eliminate fungal growth (reviewed by Payne,

1998). Though the optimal temperature for A. flavus growth is between 25-42˚C, some strains can grow in temperatures as extreme as 12-48˚C (reviewed by Payne, 1998), making it difficult to maintain storage facilities that are cool enough to prevent A. flavus infection.

Artificially raising atmospheric CO2 levels in storage facilities to 25% can reduce A. flavus growth, but at least 50% CO2 is required to reduce aflatoxin production, though the exact mechanism of CO2 inhibition is not yet understood (Giorni et al., 2008). CO2 comprises only

0.04% of the gaseous compounds found in dry air (Dlugokencky and Tans, 2016), and maintaining high CO2 levels in storage facilities is not practical for most farmers around the world.

Host: Since managing the pathogen and the environment is not sufficient for fungal elimination, resistant plant crops provide another method for prevention of A. flavus infection and aflatoxin contamination. Several cultivars of both maize and peanut are naturally resistant to A. flavus infection (Kelley et al., 2012; Liang et al., 2006; reviewed by Fountain et al., 2015; reviewed by Brown et al., 1999) and are used by farmers in regions with high frequencies of contamination. Upregulation of stress-related genes, such as those involved in the biosynthesis of phenylpropanoids (Wang et al., 2016a), can aid in resistance to fungal infection and can specifically reduce aflatoxin contamination (Dolezal et al., 2014; Chen et al., 2004; Chen et al., 2006; Chen et al., 2010; Guo et al., 2006; Guo et al., 2008).

Understanding the specific genetic differences between plants that are susceptible to

Aspergillus fungal infection and those that are resistant is a crucial step in creating an effective method for preventing A. flavus invasion.

A two-model pathosystem could help identify a mode of seed resistance to A. flavus:

Aspergillus nidulans is a model fungus. This fungus has been extensively studied and its entire genome was sequenced in 2001 (Galagan et al., 2001). Unlike most Aspergillus species, A. nidulans can undergo sexual recombination during formation of ascospores

(Pontecorvo et al., 1953). While it contains most of the gene cluster responsible for aflatoxin production, A. nidulans only produces the penultimate precursor of the aflatoxin biochemical pathway, the mycotoxin sterigmatocystin (Keller and Adams 1995).

Arabidopsis thaliana, a well-characterized plant in the Brassicaceae family, can be used as a plant model for studying the genetics underlying seed resistance to Aspergillus fungal infection. Wildtype (Wt) Arabidopsis seeds are largely resistant to fungal infection, indicating an underlying mechanism for protection against Aspergillus invasion. Arabidopsis

has a relatively small genome (~125Mb) that was sequenced in 2000 (The Arabidopsis

Genome Initiative, 2000). The plants are small at maturity and have an eight-week generation time. Arabidopsis flowers self-pollinate, reducing the labor requirements for plant care methods. Additionally, there are simple procedures for performing genetic crosses between

Arabidopsis lines. Many mutant Arabidopsis lines are available from the Arabidopsis

Biological Resource Center. Arabidopsis, as opposed to maize or peanut, provides a simpler system for studying plant genetics.

Arabidopsis seeds and A. nidulans fungus provide a promising two-model pathosystem. Most approaches to studying Aspergillus resistance in plants begin with a susceptible host and a search for resistant cultivars, then culminate in the identification of the genes involved in increased resistance (for examples, see Wilson et al., 2001; Pechanova et al., 2011; Kelley et al., 2012). Using Arabidopsis, a naturally resistant host, as a model organism allows us to take a reverse genetics approach. We can first identify genes of interest using the many web-based tools available for Arabidopsis genetics research (for example,

The Arabidopsis Information Resource (TAIR), The Bio-Analytic Resource for Plant

Biology (BAR), and Genevestigator). We can then assign susceptibility phenotypes using available mutant lines. This pathosystem allows researchers to determine if a specific gene or set of genes are necessary for conferring Aspergillus resistance in plants.

Arabidopsis seed coat (testa) development:

Angiosperm seeds are composed of the embryo, the endosperm, and the seed coat

(testa). Seed development is initiated with double fertilization of the ovule by the two sperm nuclei of a pollen grain. The ovule consists of maternal integument surrounding seven meiotic cells, six haploid and one that is double-nucleated. The haploid egg cell, situated at

the micropylar end between the two synergid cells, is fertilized to form the embryo. The double nucleated cell, between the micropylar and chalazal ends of the ovule, is fertilized to form the endosperm. The endosperm releases an unknown signal that initiates differentiation of the surrounding integument layers into a mature seed coat (reviewed by Figueiredo and

Kohler, 2014).

In Arabidopsis, the maternal integument in the ovule is made up of five layers of cells that grow via both cell division and cell growth after fertilization to form the testa. The differentiation process of the testa is complete by fifteen days after fertilization (Haughn and

Chaudhury, 2005). Two cell layers make up the outer integument (the epidermal and palisade layers) and three layers comprise the inner integument (Figure 1). The cells in the epidermal layer (Figure 1) produce and secrete large amounts of polysaccharide mucilage into the space between the plasma membrane and the cell wall. This secretion of mucilage forces the cytoplasm into the center of the epidermal cells, forming columella, after which thick secondary cell walls are formed (Windsor et al., 2000; Western et al., 2000; Haughn and

Chaudhury, 2005). The stored mucilage is not released from the outer epidermal layer until the mature seeds undergo imbibition at the initiation of germination (Western et al., 2000).

The palisade cells (Figure 1) form thick secondary cell walls on the inner side of the cell before undergoing apoptosis (Haughn and Chaudhury, 2005). The secondary cell walls produced by the outer integument cell layers provide the embryo with a physically strong encasement even after the cells are dead and the palisade layer has collapsed. The two outer layers of the inner integument do not seem to further differentiate before they undergo programmed cell death (Nakuane et al., 2005; reviewed by Hatsugai et al., 2015; reviewed

by Van Hautegam et al., 2015), after which they collapse and are compacted together with the palisade layer (Haughn and Chaudhury, 2005).

The cells in the innermost layer of the inner integument (the endothelium; Figure 1), synthesize phenylpropanoid molecules that accumulate for the first week after fertilization

(Haughn and Chaudhury, 2005). Some phenylpropanoid molecules (proanthocyanidins) then begin to oxidize, providing Arabidopsis seed coats with their characteristic brown pigmentation (reviewed by Winkel-Shirley, 2001). At the end of seed coat maturation, the phenylpropanoid molecules are released from the endothelial cells into the intracellular space as the cells undergo programmed cell death (Lepiniec et al., 2006).

Phenylpropanoids protect plants against Aspergillus infection:

Phenylpropanoids are a diverse set of molecules, produced via the phenylalanine ammonia lyase pathway, that provide seeds with protection from abiotic and biotic stressors

(reviewed by Tohge et al., 2013). Lignin molecules (polymerized phenolics) provide plants with mechanical protection by making cell walls tougher and more water resistant, thereby protecting against fungal penetration and against enzymes that can degrade the cell wall

(reviewed by Naoumkina et al., 2010; Kavousi et al., 2009). In maize, infection often occurs through cracks or wounds, indicating that lignin could be involved in providing kernels with a mechanical barrier to infection (Naoumkina et al., 2010). Lignin and its precursors may also act chemically as pathogen inhibitors (reviewed by Naoumkina et al., 2010). However, the involvement of lignin in inhibition of bacterial and fungal infection is disputed. For example, King and Scott claim that kernel hardness does not correlate with maize susceptibility to fungal infection (1982). Funnell and Pedersen showed that a reduction of

lignin content in sorghum leads to increased, rather than decreased, resistance to both bacterial and fungal infection (Funnell and Pedersen, 2006; reviewed by Naoumkina, 2010).

Phenylpropanoids likely contribute to phytoalexin chemical inhibition of fungal pathogens. Plants that are resistant to fungal infection, including resistant cultivars of maize and peanut, often show an upregulation of phenylpropanoid and flavonoid gene expression in green tissues after being challenged by fungal pathogens (reviewed by Naoumkina et al.

2010; Mierziak et al., 2014; Starr et al., 2014; Wang et al., 2016a). Stilbenes, though not present in many plants (Schröder, 1997; Dixon et al., 2002), are an important factor in peanut resistance to Aspergillus (Wotton and Strange, 1985; reviewed by Liang et al., 2006; Sobolev et al., 2007). Many of the flavonoid biosynthesis genes are upregulated during seed coat development (Winkel-Shirley, 2001; Abrahams et al., 2002; Haughn and Chaudhury, 2005;

Dean et al., 2011), and we hypothesize that flavonoids contribute to innate fungal resistance in mature seeds. It would be useful to discern which product of the flavonoid pathway is essential in Aspergillus resistance. A flavonoid gene conferring resistance in Arabidopsis could be present in maize and peanut as well, since parts of the phenylalanine ammonia lyase pathway (including flavonoid biosynthesis) are highly conserved amongst both monocots and dicots (reviewed by Dixon et al., 2002; Rawal et al., 2013); however, further analysis would be necessary in order to determine if the product was produced in agriculturally important crops, and if it was actively inhibitory in those plant systems.

Previous colleagues demonstrated that mutants in chalcone synthase, the first dedicated enzyme in flavonoid biosynthesis, were infectible by Aspergillus (Brodhagen and

Young, unpublished data). Flavonoids, a subset of phenylpropanoids, are derivatives of the fifteen-carbon flavan molecule (Figure 2a). They have been previously shown to chemically

inhibit Aspergillus mycelial growth (for examples, see Weidenbörner et al., 1990; Kanwal et al., 2010). Additionally, mature seeds that are resistant to fungal infection can show a higher expression rate of flavonoid biosynthesizing genes when compared to seeds that are susceptible to fungal infection. For example, flavonoid-specific enzymes such as chalcone synthase (CHS) are upregulated in mature peanut seeds that are resistant to Aspergillus fungal infection (Wang et al., 2016a). A detailed understanding of flavonoid biosynthesis will help identify the specific proteins that are involved in providing Arabidopsis seeds with resistance to Aspergillus.

Flavonoids: biosynthesis, modification, and transport in Arabidopsis:

The identification of mutant Arabidopsis lines with impaired flavonoid accumulation and altered seed coat pigmentation (Koornneef, 1990), named transparent testa mutants

(Table 1), has helped identify many of the genes and proteins involved in the accumulation of flavonoid molecules (Shirley et al., 1995; Abrahams et al., 2002). Arabidopsis seeds accumulate at least fifty-four flavonoid molecules (reviewed by Saito et al., 2013), including flavonols, flavan-3-ols, and epicatechin-based proanthocyanidins (Figure 2b,i, ii, iii, and

Figure 3) (Routaboul et al., 2006). There are four main functional categories of proteins involved in flavonoid production: enzymes involved in the biosynthesis of the flavonoid scaffolds, enzymes that modify the scaffold structures, transport proteins, and transcriptional regulatory factors (Figure 3) (reviewed by Saito et al., 2013; reviewed by Appelhagen et al.,

2014; reviewed by Lepiniec et al., 2006).

In Arabidopsis, there are nine enzymes known to be involved in the biosynthesis of flavonoid scaffold molecules (Figure 3). Chalcone synthase (CHS), the first enzyme in the flavonoid branch of the phenylpropanoid biosynthetic pathway, catalyzes the condensation

reaction of p-coumaroyl CoA with three molecules of malonyl CoA to form chalcone

(tetrahydroxychalcone) (Austin and Noel, 2003). After CHS initiates this first committed step in the production of flavonoids, chalcone isomerase (CHI) catalyzes the cyclization reaction of chalcone into naringenin (a flavanone, Figure 2b,iv). Naringenin is then oxygenated at the

3-position by flavanone 3-hydroxylase (F3H), to give dihydrokaempferol (dihydroflavonol)

(Pelletier et al., 1996). Dihydrokaempferol can undergo a desaturation at the 2,3 position by flavonol synthase (FLS) to produce kaempferol, or it can be hydroxylated at the 3’ position by flavonoid 3’ hydroxylase (F3’H) to produce dihydroquercetin. Dihydroquercetin can also undergo a desaturation at the 2,3 position by FLS to produce quercetin, or it can undergo a reduction of the keto group to a hydroxyl group at the 4 position by dihydroflavonol reductase (DFR) to produce leucocyanidin (a flavan-3,4-diol, Figure 2b,v). The branches initiated by FLS mark the first committed steps in the production of flavonols, while the branch initiated by DFR marks the first committed steps in the production of proanthocyanidins. Full desaturation of the C ring of leucocyanidin is catalyzed by leucoanthocyanidin dioxygenase (LDOX), resulting in cyanidin. Cyanidin is then reduced to epicatechin by the anthocyanidin reductase (BAN) enzyme (Abrahams et al., 2003). The final biosynthetic enzyme, laccase-like 15 (LAC15), is a laccase-like polyphenol oxidase that is involved in the oxidative polymerization of flavonoids that yields proanthocyanidins

(Pourcel et al., 2005), the majority of which are epicatechin-based in Arabidopsis (reviewed by Saito et al., 2013). One other laccase enzyme, LAC5, is upregulated during silique development (Figure 4), but it has no known involvement in flavonoid biosynthesis.

Accumulation of flavonoids in developing Arabidopsis seeds:

Flavonoids accumulate during Arabidopsis seed coat development. The main flavonol in mature Arabidopsis seed coats is quercetin-3-O-rhamnoside. It begins to accumulate in the endothelial cells about seven days after flowering and continues accumulating through the end of seed coat maturation (Routaboul et al., 2006). Epicatechin accumulates in high concentrations seven days after flowering, but decreases dramatically at the end of seed coat maturation, when it is polymerized to form proanthocyanidins. Soluble proanthocyanidins accumulate in the endothelial cells up to twelve days after flowering, after which their concentration quickly decreases, leaving very few soluble proanthocyanidins in the mature seed coat. Insoluble proanthocyanidins, on the other hand, begin to accumulate within four days after flowering, increase gradually in concentration, and are still present in high quantities in mature seed coats (Routaboul et al., 2006).

Modification of Arabidopsis flavonoids:

The flavonoid scaffold molecules are modified in different ways by several enzymes.

Flavonoid glycosyltransferase (TT15) catalyzes the glycosylation of flavonoids, using both sterol glucoside and UDP-glucose as substrates (Figure 5a). Other enzymes catalyze the prenylation, methylation, and acylation of flavonoid scaffold molecules (see Figure 5b for prenyl-, methyl-, and acyl- groups) (reviewed by Saito et al., 2013). Another enzyme, glutathione S-transferase PHI 12 (GSTF12), is homologous to glutathione transferases; however, instead of catalyzing the addition of glutathione molecules (Figure 5c) onto flavonoid scaffold molecules, GSTF12 is believed to be involved in transporting flavonoids into the vacuole and potentially in protecting them from oxidation (reviewed by Zhao et al.,

2010).

Transport and storage of flavonoids in Arabidopsis:

Several other proteins are involved in the transport and storage of flavonoids. During the early stages of seed coat development, proanthocyanidins are stored in the vacuoles of endothelial cells (Abrahams et al., 2003). Synthesis of proanthocyanidin precursors, however, is believed to occur in the cytosol near the endoplasmic reticulum, and these precursors must be transported into the vacuole (Debeaujon et al., 2001). The transparent testa 12 (TT12) enzyme, a multi-antimicrobial extrusion (MATE) family transporter

(Debeaujon et al., 2001), localizes to the membrane of endothelial cell vacuoles. It is involved in proanthocyanidin transport (Debeaujon et al., 2001), but the intermediates transported into the vacuole by TT12 are not yet known. Another transporter necessary for proanthocyanidin synthesis in endothelial cells is autoinhibited H+-ATPase 10 (AHA10)

(Baxter et al., 2005). AHA10 is a P-type H+-ATPase that could aid the TT12 enzyme in transporting proanthocyanidins, since MATE antiporters require an ion gradient across a membrane to drive their activity.

Regulation of flavonoid biosynthetic pathway genes:

The expression of genes in the flavonoid biosynthetic pathway is under the regulation of six known transcription factors. Three proteins form a transcription complex: MYB domain protein 123 (MYB123), which is a myeloblastosis (MYB) family transcription factor; a basic helix-loop-helix transcription factor (BHLH42); and transparent testa glabra 1

(TTG1), which is a WD-repeat (WDR) family protein. The MYB-bHLH-WDR complex is well conserved, and MYB123-BHLH42-TTG1 regulates the transcription of the genes involved in the initiation of proanthocyanidin production (Debeaujon et al., 2003). A WRKY transcription factor (WRKY44) is thought to function downstream of TTG1 (reviewed by

Debeaujon et al., 2003; reviewed by Xu et al., 2015). Arabidopsis B-sister protein (ABS) is a

MADS-domain (MCM1, Agamous 1, Deficiens, and Serum response factor) protein (Nesi et al., 2002) involved in the initiation of the development of endothelial cells and their accumulation of proanthocyanidins in part by the regulation of anthocyanidin reductase

(BAN) (Dean et al., 2011). WIP domain protein 1 (WIP1), has a zinc finger domain and is also involved in the regulation of anthocyanidin reductase (BAN).

Summary:

With such a detailed understanding of flavonoid biosynthesis in Arabidopsis, and with mutant lines available for each known gene in the pathway, we aim to decipher which flavonoid biosynthetic, transport, and regulatory genes are directly involved in providing Wt

Arabidopsis seeds with resistance to Aspergillus fungal infection. This thesis will cover a dissection of the Arabidopsis flavonoid biosynthetic pathway. Since Arabidopsis does not produce anthocyanins or stilbenes, this thesis only covers the resistance mechanisms provided by a reduced set of phenylpropanoid molecules. This thesis will also take a detailed look into the variability in susceptibility phenotypes of Arabidopsis seed mutants, and will assess the robustness of the Arabidopsis and A. nidulans pathosystem. This work could have implications for understanding and improving crop seed resistance to A. flavus infection and aflatoxin contamination.

METHODS

Plant care for Arabidopsis thaliana:

Arabidopsis thaliana seeds of the Columbia and Landsberg ecotypes were received from the Arabidopsis Biological Resource Center (Ohio State University). Strains included wildtype Columbia (Wt Col) and wildtype Landsberg (Wt Ler), as well as strains with single gene mutations in flavonoid biosynthetic pathway genes (chs, fls, f3’h, dfr, ldox, ban, lac15, tt15, gstf12, tt12, aha10, bhlh42, ttg1, and wip1), in an additional laccase gene (lac5), and in the ERECTA gene (er) (Table 1). The seeds were vernalized at 4oC for seven days. They were then surface sterilized in 70% EtOH/0.01% Triton X-100 for five minutes, followed by a 95% EtOH rinse for one minute. The seeds were planted on 0.5X Murashige Skoog (MS)

1% agar media and placed at 4oC in the dark for three days, before being placed at 21oC under fluorescent light for 14 days. The seedlings were transplanted into a 2:1:1 sterilized mixture of SUNSHINE® Professional Growing Mix #2 media (SUN GRO Horticulture

Canada Ltd, Seba Beach, AB, Canada), Plant!t Super Coarse Perlite (Hydrofarm, Petaluma,

CA), and Hoffman® Horticultural Vermiculite (Good Earth™, Lancaster, NY) using Miracle

® Gro fertilizer (0.04g/L DI H2O; 30% N, 10% P, 10% K, less than 1% B, Cu, Fe, Mn, Mo,

Zn; product #1001791) (Scotts Miracle-Gro Products, Inc., Marysville, OH). The plants were grown at 21oC under fluorescent light (F4012/C50 SUPREME; 40 watts; Alto Technologies,

Phillips Lighting Company, Somerset, NJ) with a sixteen-hour day, and were watered with

Miracle Gro fertilizer (0.08g/L DI H2O) when the pots were no longer heavy with water.

Cohorts of seeds grown over two years ago were grown under a 24-hour light regimen.

Whole-plant-harvesting: Six weeks after transplanting, the plants were left unwatered and allowed to dry for two to eight months at 21˚C under fluorescent light. Seeds were

whole-plant harvested and stored in microcentrifuge tubes at room temperature. Each microcentrifuge tube held seeds from one parent plant, with a range in seed ages spanning up to seven weeks difference (Figure 6). It is of note that plants in the Ler genetic background often have less biomass than plants in the Col background. They are often slower to begin development, but faster to complete it, than Col ecotype plants. Under our experimental conditions, plants from different ecotypes were not segregated into separate watering trays, so the smaller Ler ecotype plants were likely overwatered.

Fungal care for Aspergillus nidulans:

Wildtype Aspergillus nidulans RDIT 9.32 was obtained from Dr. Nancy P. Keller, and stored at -80oC in glycerol. A serial dilution of A. nidulans was plated out on 1.6% (w/v) agar Glucose Minimal Media (GMM) (0.056 M glucose, 0.07 M NaNO3, 7E-3 M KCl, 2E-3

M MgSO4•7H2O, 0.011 M KH2PO4, 7.6E-5 M ZnSO4•7H2O, 1.8E-4 M H3BO3, 2.5E-5 M

MnCl2•4H2O, 1.8E-5 M FeSO4•7H2O, 7.3E-6 M CoCl2•5H2O, 6.4E-6 M CuSO4•5H2O, 1.7E-

7 M (NH4)6Mo7O24•4H2O, 1.3E-4 M Na4 EDTA; pH 6.5; 16 g agar in 1 L) (Bailes et al.,

2013) and incubated at 37oC in the dark. After 24-36 hours of growth, new GMM plates were point inoculated from single germinated spores and incubated at 37oC under broad-spectrum light (Halco Lighting Technologies, Norcross, GA). After seven days, the plates were stored at 4oC with Parafilm. From these stock cultures, A. nidulans was streaked on GMM agar and incubated for 48 hours at 37oC under broad-spectrum light. The spores were harvested in

0.01% (v/v) Triton X-100, stored at 4oC in a 15 mL conical vial, and used in bioassays within

12 hours.

Assessment of Arabidopsis seed susceptibility to A. nidulans:

Humidity chambers were aseptically prepared in NUNC® six-well plates (Thermo

Fisher Scientific, Roskilde, Denmark). Two pieces of #1 grade, 2.5 cm diameter Whatman™ filter paper (GE Healthcare Life Sciences, Little Chalfont, Buckinghamshire, UK) in each well were wetted with 100 µL of sterile Nanopure H2O. To help maintain high humidity, a microcentrifuge tube cap (removed from a 1.5 mL tube; Thermo Scientific, San Diego, CA) in each well was filled with 230 µL of sterile Nanopure H2O. Approximately 90 seeds by volume (plus or minus 10 seeds) were placed in microcentrifuge tubes for wetting. 250 µL of

0.01% Triton X-100 was added to each microcentrifuge tube before seeds were pipetted onto the filter paper of their corresponding wells in 150 µL of 0.01% Triton X-100. The seeds in each well were inoculated with 106 wildtype A. nidulans RDIT 9.32 spores in 15 µL of

0.01% Triton X-100, and incubated at 37oC in the dark. After an initial 16 hours, and then at

12 hour intervals, the filter papers in the four corner wells were rehydrated with 80 µL of sterile Nanopure H2O and the two center wells with 60 µL. After 48 to 72 hours, or until sporulation occurred on the positive control seeds (old chs seeds from the tt4 line in the Ler background), the top pieces of filter paper, along with the inoculated seeds, were placed in individual 50 mL Falcon® tubes (Corning Science, Reynosa, Tamaulipas, México) in 1 mL of sterile 0.01% Triton X-100. The tubes were vortexed for one minute; then, from the average of four counts on a hemocytometer, the average concentration of spores per seed was calculated for each replicate sample. A minimum of three biological replicates was run per seed type. The infection rates of individual mutant seed strains were compared to the infection rate of the wildtype seeds using an unpaired two-tailed t-test, with an alpha value of

0.05. Each experiment was repeated at least once.

Two controls were run with each experiment. First, there was a no-spore trial for each seed type to control for the fact that seeds were not surface-sterilized prior to the seed

infection assays; this control ensured that native epiphytic fungi were not present and were not contributing to the observed fungal growth. Each seed type had one well with 15 µL of

0.01% Triton X-100, instead of inoculation with 15 µL of fungal spore suspension (results not shown). Second, there was a no-seed control to ensure that the fungus was viable. Filter paper without any seeds was inoculated with 106 A. nidulans spores, in a minimum of three biological replicates, and quantification of fungal spores was calculated as described above.

Timecourse of Arabidopsis seed infectibility:

Buds of Arabidopsis thaliana Wt Col and chs plants were marked weekly, to yield cohorts of seeds harvested at similar days after flowering (DAF), with a resolution of one week (Figure 6). Three new seedlings from each genotype were planted every two weeks.

The plants were watered with Miracle Gro fertilizer (0.08g/L DI H2O) when the surface of the planting media started to dry.

Silique-harvesting: One week after peduncle collapse, seeds were harvested silique- by-silique and stored in microcentrifuge tubes at room temperature. Each microcentrifuge tube held seeds that were within a week of the same age, and which came from an average of nine different plants. Cohorts were marked for a ten-month period. Infectibility of older and younger seeds was compared, using the protocol for the assessment of seed infectibility described above.

Analysis of Arabidopsis seed germination rates:

Seeds were sterilized as described above in Plant Care. Twenty to sixty seeds were then planted in triplicate on 0.5X Murashige Skoog (MS) 1% agar media in 100 X 15mm

Petri plates using a pick sterilized with 70% ethanol. They were held at 4oC in the dark for

three days, before being placed at 21oC under fluorescent light. The percentage of germinated seeds per plate was quantified after ten days.

Analysis of the inhibitory properties of aqueous extracts from Arabidopsis seeds:

Aqueous seed extract was prepared with 0.1 mL (by volume) of each strain of

Arabidopsis thaliana seeds. Seeds were soaked in 600 µL of sterile ddH2O in microcentrifuge tubes for fifteen minutes with shaking at 30 rpm, followed by 15 seconds of vortexing. The seeds were ground using one hundred quick pulses from a microcentrifuge bit attached to a VirTis “45” drill (Gardiner, NY) on low. An additional 800 µL of DI H2O was added to each tube, followed by 15 seconds of vortexing. Tubes were centrifuged at maximum speed (14000 rpm) for three minutes, after which the aqueous extract supernatant was pipetted into a sterile microcentrifuge tube. An agar base of 1 mL solid Glucose Minimal

Media (1.6% agar w/v) was pipetted into capless 1.5 mL microcentrifuge tubes (Thermo

Scientific, San Diego, CA), and allowed to solidify. A soft agar overlay was prepared with 20

µL of aqueous extract added to 130 µL of semisolid Glucose Minimal Media (0.8% agar w/v) containing a suspension of 104 spores/mL of A. nidulans RDIT 9.32. The solution was vortexed for 30 seconds on maximum speed, after which 150 µL was spread evenly over the solid GMM base, into the respective capless microcentrifuge tubes. Another capless, sterile microcentrifuge tube was secured over each experimental tube, using micropore tape. The tubes were incubated agar-side-up at 37oC in the dark for 45 hours. The contents of each tube were added to 3 mL of 0.01% Triton X-100, then crushed with a pestle to a uniform consistency and vortexed for 1 minute at top speed. A hemocytometer was used to enumerate spores/cm2. Each treatment (seed extract from a particular genotype) consisted of five biological replicates. Each experiment was repeated a minimum of two times. Two controls

were run with each experiment: sterile Nanopure H2O instead of aqueous seed extract, and

0.01% Triton X-100 instead of spore suspension.

RESULTS

Validation of the pathosystem:

Wildtype Col and Ler Arabidopsis thaliana seeds are both resistant to Aspergillus nidulans infection (Figure 7a). Arabidopsis seeds mutant in chalcone synthase (chs) are susceptible to both A. flavus (Figure 7b) and A. nidulans (Figure 7c), with infection often occurring at the micropylar end of the seeds (data not shown). The susceptible chs seeds are in the Landsberg genetic background and have a second mutation in the ERECTA gene.

However, a single er mutation does not diminish the resistance of Col or Ler Arabidopsis seeds to A. nidulans fungus (Figure 7d).

A genetic dissection of the flavonoid biosynthetic pathway, to identify Aspergillus resistance factors:

A genetic dissection of the flavonoid biosynthetic pathway in Arabidopsis revealed that mutations in chs, f3’h, and dfr result in seeds that are more susceptible than Wt Col seeds to A. nidulans (Figure 8a, b, c; Figure 3), while seeds with mutations in fls, ldox, and ban remain resistant to infection (Figure 8d, e, f; Figure 3). Mutations in the known modifying enzymes (tt15 and gstf12) and the known transporters (tt12 and aha10) that are involved in flavonoid production do not result in susceptibility to A. nidulans (Figure 8g, h, i, j; Figure

3). A mutation in the wip1 transcription factor results in susceptibility, while a mutation in the bhlh42 transcription factor does not (Figure 8k, l; Figure 3). A mutation in lac15, a laccase-like enzyme known to act in the flavonoid biosynthetic pathway, results in seeds that are mildly susceptible to A. nidulans infection (Figure 8m, Figure 3; Figure 9). Seeds with a mutation in lac5, a laccase enzyme that is upregulated during silique development, but has no known involvement in flavonoid biosynthesis, remain resistant to A. nidulans (Figure 9). It is

of note that the seed infectibility assays described here were performed with seeds that were harvested from entire plants and allowed to mature to at least 133 days after flowering.

Timecourse of Arabidopsis seed infectibility:

We noted that older whole-plant-harvested chs mutant seeds were sometimes more susceptible to A. nidulans, while younger mature seeds are less susceptible (example, Figure

10). Therefore, a timecourse assay was designed to measure seed susceptibility over time, comparing Wt Col and chs seeds from plants whose siliques were marked at one-week intervals and individually harvested one week after peduncle collapse. Developing seeds from both Wt Col and chs plant lines are susceptible to infection, but resistance develops in both seed types when they reach maturity between 8 to 18 days after flowering (Figure 11).

In contrast to our observations from whole-plant-harvested seeds, silique-harvested seeds remained resistant to A. nidulans regardless of age (Figure 11). The timecourse was only performed once, but its results were so unexpected as to prompt us to examine possible developmental differences between whole-plant- and silique-harvested seeds.

Effect of seed development on chs seed susceptibility to A. nidulans infection:

Ten months after flowering, silique-harvested chs seeds were not susceptible to A. nidulans (Figure 12b). However, ten-month-old chs seeds that were whole-plant-harvested after an eight-month drying period were highly susceptible (Figure 12b). The silique- harvested chs seeds are a vibrant, yellow color (Figure 12a), and have a germination rate over 99% (Figure 12c). The whole-plant-harvested chs seeds are a dull, grey-yellow color

(Figure 12a), and have a germination rate of about 40% (Figure 12c). While the silique- harvested seeds remain yellow during infection experiments, most of the whole-plant- harvested chs seeds visibly turn darker after 32 hours at 37˚C in the dark (data not shown).

The seeds that turn darker during incubation will not germinate, even if they are placed under fluorescent light at 21˚C (data not shown).

Closer inspection revealed that some cohorts of chs seeds have a mixture of phenotypes: larger, yellow seeds vs. smaller, darker seeds (Figure 13a). The smaller, darker chs seeds are those that are produced in the late-forming siliques that often develop in tight clusters after the plant has been taken off of water (Figure 6, week 7). These seeds are an average of 44% smaller than seeds produced in siliques from the main branches of an

Arabidopsis plant (Figure 13b). The yellow seeds are resistant to A. nidulans, while isolated late-forming seeds are highly susceptible to infection (Figure 13c). When late-forming chs seeds were mixed with either Wt Col seeds or with yellow chs seeds, inoculation with A. nidulans spores resulted in high infection of the late-forming seeds by 36 hours while at 48 hours all other seeds were clear of infection (data not shown). The yellow seeds have a germination rate of 97%, while the small, late-forming seeds have a reduced germination rate of only 15% (Figure 13d), and the resulting seedlings are stunted in growth (Figure 13e).

Effect of genotype on chs seed susceptibility to A. nidulans infection:

When four different chs mutant seed types were inoculated with A. nidulans, they showed variability in susceptibility, correlating with their genetic background (Figure 14a).

Seeds with mutations in the Ler genetic background (tt4 and tt4-1) were more highly infected than those in the Col genetic background (tt4-3 and tt4-TDNA). All seeds were used 70 days after flowering, and all parent plants were from the same cohort and were planted, watered, and whole-plant-harvested on the same dates; plants from different ecotypes were not segregated into separate watering trays. A germination test of these chs seeds showed that the

two mutant seed types that had a higher susceptibility to infection also had lower germination rates (Figure 14b).

Mode of wildtype Arabidopsis resistance to Aspergillus:

The aqueous extract from Wt Col seeds prevents A. nidulans growth, while the fungus grows and sporulates prolifically in a no-extract water control (Figure 15). However, aqueous extracts from chs seeds variably affect fungal growth, even when all seeds have the same allelic mutation and are grown and stored under the same temperatures and light frequencies. The effect of chs seed extracts on A. nidulans does not correlate with the infectibility, germination rates, or age of the seeds (Figure 16).

DISCUSSION

Creation of a two-model pathosystem:

Resistance of wildtype Arabidopsis thaliana seeds to Aspergillus fungi indicates the presence of a mechanical and/or chemical barrier to infection. The identification of mutant

Arabidopsis seeds that are susceptible to Aspergillus nidulans provides a two-model pathosystem for studying the mode of seed resistance to Aspergillus. Mutant Arabidopsis seeds in either the Col or the Ler genetic background can be used in this pathosystem to find genes that are necessary for fungal resistance.

Flavonoids involved in Arabidopsis seed resistance to A. nidulans:

Susceptibility of seeds mutant in the first enzyme of the flavonoid biosynthetic pathway, chalcone synthase (chs), suggests that a downstream molecular product(s) could be responsible for conferring resistance in wildtype Arabidopsis seeds. The genetic dissection of the flavonoid biosynthetic pathway shows that genes from CHS through DFR are necessary for conferring resistance to Aspergillus, while genes at the heads of the flavonol and proanthocyanidin branches (FLS and LDOX, respectively) may not be necessary. This would suggest that leucocyanidin, the molecular product of the last necessary gene (DFR), is required for Arabidopsis seed resistance to Aspergillus. It also suggests that desaturation of the C ring is not necessary for fungistatic activity of the active compound. We have not yet tested whether leucocyanidin is directly or indirectly inhibitory to Aspergillus, since our attempt to synthesize the molecule via a sodium borohydride (NaBH4) reduction of dihydroquercetin was unsuccessful. Future research should include an analysis of the inhibitory properties of the leucocyanidin molecule, to determine if it has direct fungicidal

properties against Aspergillus, or if leucocyanidin acts as a precursor for the active molecule(s) that provides wildtype Arabidopsis seeds with resistance.

If leucocyanidin is not directly active, susceptibility phenotypes of Arabidopsis seeds with mutations in known flavonoid modifying enzymes, transporters, and laccases provides insight into possible conversion steps of leucocyanidin from an inactive to an active molecule. Since tt15, gstf12, tt12, and aha10 mutant seeds are resistant to A. nidulans infection, conversion of leucocyanidin would not likely rely on glycosylation by TT15, glutathione transferase activity or transportation by GSTF12, or transportation into vacuoles by TT12 or AHA10. The susceptibility of lac15 mutant seeds to A. nidulans, however, suggests that the laccase is involved in providing seeds with Aspergillus resistance. LAC15 acts on several substrates and has known activity in the production of biflavonols and oxidized proanthocyanidins (Figure 3). However, since fls and ldox mutant seeds are resistant to A. nidulans, those branches of the pathway are not necessary for resistance. To create a chemical resistance factor, LAC15 could either act on an unknown substrate, or, in a novel branch of the flavonoid biosynthetic pathway, LAC15 may use leucocyanidin as a precursor in producing an active anti-fungal molecule (Figure 3). Since lac15 seeds are only mildly susceptible to A. nidulans, its activity can be assumed necessary, but not sufficient, for resistance; i.e., another factor(s) is necessary to achieve wildtype levels of resistance. A reasonable hypothesis is that another laccase is functionally redundant to LAC15. Though the biological role of LAC5 has not yet been identified, it is the only other laccase in Arabidopsis that is highly upregulated in mature siliques (Figure 4). Though a single mutation in LAC5 is not sufficient to cause susceptibility, a lac15/lac5 double mutation may be more susceptible

than either single mutant. Further investigation is necessary for determining the involvement of laccases in conferring resistance against Aspergillus.

Resistance of developing vs. mature Arabidopsis seeds:

Susceptibility of both Wt Col and chs seeds varies over time. Developing seeds from both plant lines are highly susceptible to A. nidulans until they reach maturity, at which point they both develop resistance. This susceptibility of young seeds, which to our knowledge has not yet been described, suggests that green seeds may rely on the parent plant for protection until they produce a mechanical and/or chemical barrier to infection (for a detailed review in seed development, see Chaudhury et al., 2001; Lepiniec et al., 2006; Bentsink and

Koornneef, 2008; Le et al., 2010). Mature Wt Col seeds remain resistant to A. nidulans, indicating that the barrier(s) are maintained over time. Mature chs seeds, however, show variability, sometimes remaining resistant to A. nidulans over time, but sometimes developing susceptibility. This indicates that a barrier in chs seeds may be inconsistently lost or broken down over time.

Variability in chs seed infection correlates with variation in germination rates:

The variability in susceptibility of chs seeds to A. nidulans correlates with seed viability: higher rates of infection correlate with lower germination rates. This indicates that

A. nidulans infects dead or weak seeds more readily than it infects living seeds, as is expected of a saprophytic fungus. These differences in seed susceptibility and viability appear under several changes of variables.

Growth conditions:

There are differences in seed viability amongst seeds grown in different conditions. A longer drying period for the parent plant before seed harvesting appears to result in reduced

viability of seeds, suggesting that long lengths of time under fluorescent light reduces fitness in chs seeds, but not in Wt Col seeds. Seed viability could be reduced by either ultraviolet light damage or via oxidative stress. The flavonoid biosynthetic pathway, rendered non- functioning in chs seeds, is important for protection of plant tissues against ultraviolet light damage (Li et al., 1993; Landry et al., 1995). Flavonoids absorb ultraviolet light, thereby preventing direct damage to DNA and RNA. They also function as antioxidants, protecting the cells from reactive oxygen species (reviewed by Agati and Tattini, 2010), which increase under ultraviolet light stress (Rozema et al., 1997). Additionally, in green tissues, oxidative damage to membranes caused by heat stress is induced by light exposure (Larkindale and

Knight, 2002), indicating that seeds not protected by flavonoids could be more vulnerable to oxidative damage. Since fluorescent lights can release ultraviolet light wavelengths

(Mironava et al., 2012), it is plausible that excessive light exposure under our experimental conditions damages chs seeds more readily than wildtype seeds (for examples, see Shaukat et al., 2013; Sullivan et al., 2014; Wang et al., 2016b), thereby causing lower rates of germination for chs, but not Wt Col seeds, from cohorts with long drying periods prior to harvest.

Water stress of the parent plant may also play a role in seed fitness. Under conditions of excess water during seed development, seeds from many plant species decrease in impermeability, and therefore increase in leakiness (reviewed by Jaganathan, 2016). Our watering regimen for whole-plant-harvested seeds was inconsistent from cohort to cohort, producing variability in levels of water stress and excess for the parent plants, and likely contributed to variability in seed desiccation and viability. The parent plants of silique- harvested seeds, however, were watered at optimal intervals, which may contribute to the

uniformity seen in their viability. Additionally, wildtype seeds have flavonoid proanthocyanidin polymers to help prevent leakage of solutes, but chs seeds have thinner and more permeable seed coats, and thus weaken at a faster rate (Debeaujon et al., 2000). This could result in increased chs seed desiccation over time, contributing to the lower rates of germination and higher rates of infection seen in some chs seeds compared to Wt Col seeds.

Our pathosystem is prone to variability in changing environmental conditions. Future research should only use Arabidopsis seeds that are harvested immediately after drying and which are watered at optimal intervals.

Parent plant’s developmental stage when seeds were produced:

The time at which a chs seed is produced during the parent plant’s development seems to affect seed viability, with late-forming seeds showing reduced rates of germination and growth. The late-forming chs seeds develop after the watering period has ended and the parent plant has begun to desiccate. This results in drought conditions during development, which is known to shorten the seed-filling period and reduce the availability for nutrient transfer from the parent plant into the developing embryo (Seiler et al., 2011; reviewed by

Mitchell et al., 2016). Additionally, the location of seeds on a parent plant can greatly influence the availability of nutrients for developing seeds, resulting in varied seed fitness and a wide-range of germination rates of seeds from a single plant (reviewed by Mitchell et al., 2016). Due to a reduction in both water and nutrient availability, it is possible that the late-forming chs seeds produce fewer chemical barriers to infection and/or have reduced seed coat integrity when compared to the chs seeds produced at the peak development of the parent plant, explaining in part why A. nidulans more readily infects late-forming seeds.

Analysis of flavonoid gene involvement in seed susceptibility would be more accurate if late-

forming seeds were excluded from the bioassays. This would help ensure that susceptibility phenotypes are not skewed by the varied abundance of late-forming, underdeveloped seed outliers.

Genetic backgrounds:

Seeds with different genetic backgrounds can show variability in seed viability. This variability amongst seeds grown in the same conditions, (with identical planting, transplanting, and harvesting dates) suggests that the genetic background of seed lines or the presence of a second mutation may influence the rate at which chs seeds die.

It appears that chs seeds in the Ler background have higher rates of infection and lower rates of germination than chs seeds in the Col background (Figure 14a). Wildtype Col and Ler seeds, however, are genetically very similar. They originated from the same non- homozygous population of Arabidopsis plants, collected near Limberg, Germany. This would suggest that the differences seen in seed viability are not dependent on the genetic background, but may be due to the presence of a second mutation.

Irradiation of some seeds from the non-homozygous Limberg population created the erecta (er) mutation, and Landsberg erecta (Ler) was used as the genetic background for several subsequent mutant Arabidopsis lines, including each susceptible Arabidopsis line from chs through dfr, and lac15. Though the er gene mutation in Landsberg genetic background mutants is not sufficient for seed susceptibility to Aspergillus (Figure 7d), an er mutation in green tissues is known to increase plant susceptibility to pathogens such as the bacterium Ralstonia solanacearum, the necrotrophic fungus Plectosphaerella cucumerina, and the oomycete Pythium irregulare (Godiard et al., 2003; Llorente et al., 2005; Adie et al.,

2007; Sánchez-Rodríguez et al., 2009; reviewed by Miedes et al., 2014). Susceptibility of

green tissues of er mutant plants appears to correlate with an altered cell wall composition

(Sánchez-Rodríguez et al., 2009; reviewed by Miedes et al., 2014). It is formally possible that the er mutation in Ler genetic background mutants decreases seed viability over time by further altering the composition of the seed coat cell walls during development. The er mutation could therefore augment the infectibility phenotype of flavonoid biosynthetic pathway mutants. Furthermore, the er mutation results in plants that are shorter, and tend to have a decreased biomass. Under our experimental conditions, with plants of all ecotypes sharing watering trays, smaller plants tended to get overwatered. Excess water could contribute to an increased permeability (reviewed by Jaganathan, 2016) of chs seeds in the

Ler genetic background. Further investigation is needed in order to determine if the er mutation effects the susceptibility phenotypes and viability of other flavonoid mutations.

Arabidopsis seeds may have both mechanical and chemical barriers to A. nidulans infection:

Since all wildtype seeds and some chs seeds are resistant to A. nidulans, we hypothesize that a mechanical barrier to Aspergillus infection, consisting of the outer cell layers of the seed coat along with the micropylar plug, is initially present in Arabidopsis seeds. Environmental differences that result from varying plant care methods, such as light damage and excessive watering during seed development, can result in seeds that do not develop as strong of a mechanical barrier during development and are therefore more susceptible to embryonic damage and desiccation (Debeaujon et al., 2000; reviewed by

Mitchell et al., 2016). These stressors put on the parent plant could thereby contribute to decreased germination rates and increased susceptibility to Aspergillus infection. More

research is required in order to fully understand why the mechanical barriers of some seeds are weaker than others.

Our data suggest that further resistance of Arabidopsis seeds is provided by water- soluble, chemical resistance factor(s) that prevent A. nidulans proliferation in the absence of mechanical barriers. The aqueous seed extract from Wt Col seeds reduces the germination rates of A. nidulans spores (Brodhagen, unpublished data), indicating that the chemical resistance factor(s) are a growth retardant, but are not fungicidal. The chemical barrier(s) are consistently present in Wt Col seeds, but variably so in chs seeds (Figure 16c). Although abiotic and biotic stressors can result in fluctuations in flavonoid biosynthesizing genes (for examples, see Shaukat et al., 2013; Wang et al., 2016a; Wang et al., 2016b), flavonoids are produced at basal levels in Arabidopsis under normal growth conditions (reviewed by Saito et al., 2013). This indicates that a flavonoid inhibitory compound would always be present, to some degree, in wildtype seeds. However, chs seeds, which lack flavonoids, can also inhibit fugal growth. This supports that there are both flavonoid and non-flavonoid inhibitory, water- soluble resistance factors present in Arabidopsis seeds. The genes responsible for producing non-flavonoid inhibitory compound(s) may only be turned on under specific growth conditions or stressors during the development of the plant, explaining in part the variability seen in chs seed susceptibility; however, further investigation is required before definitive conclusions can be drawn.

One remaining confounding factor is that we have not yet formally ruled out the potential effects of fungal growth enhancer(s) such as pectins, celluloses, and lipids

(reviewed by Bennett, 2010). Mucilage, for example, is a water-soluble polysaccharide secreted after seed imbibition, composed primarily of pectin in Arabidopsis (Goto, 1985).

Though preliminary data indicates that mucilage does not affect fungal growth and sporulation (Young, unpublished data), analysis of Arabidopsis seeds mutant in mucilage production genes (reviewed by Moïse et al., 2005) would help confirm these results. Some positive enhancers, such as oils, have an increased concentration in flavonoid-deficient

Brassicaceae species (Simbaya et al., 1995), and can fluctuate with variation in environmental conditions. For example, increased light can result in an increased lipid concentration in Arabidopsis seeds (Li et al., 2006). It is possible that variability in A. nidulans growth is due to fluctuations in both fungal growth enhancers and inhibitors. Future research should aim to identify the role that fungal growth enhancers play in Aspergillus growth in the presence of Arabidopsis seed extracts. For example, bioinformatics tools such as Genevestigator could help identify genes that are upregulated during seed development, and which are also involved in pectin, cellulose, or lipid production. Seeds mutant in these genes could be obtained from the Arabidopsis Biological Resource Center, and could be crossed with flavonoid mutants. Double mutant susceptibility phenotypes could be compared to the susceptibility of single mutants. This assessment would enhance our Arabidopsis and

A. nidulans pathosystem.

CONCLUSION

Wildtype Arabidopsis seeds are resistant to A. nidulans infection via mechanical and water-soluble chemical barriers. Pathogen resistance is typically stacked so it is not surprising to find multiple antifungal mechanisms (reviewed by Palma et al., 2009; Bednarek and Schulze-Lefert, 2009; and Fountain et al., 2015). In Arabidopsis, it is possible that a mechanical barrier may be responsible for protecting young mature seeds, while the flavonoid biosynthetic pathway may be responsible for protecting older mature seeds whose mechanical barriers have lost integrity. However, we cannot confidently assign specific flavonoid resistance factors until our pathosystem has been optimized.

The pathosystem we developed is prone to variability. Our results suggest that chs mutant seeds are more prone to differences in development due to age and maternal effects than wildtype seeds, which is consistent with previous research (Debeaujon et al., 2000).

Under our experimental conditions, susceptibility of seeds was inversely correlated with germination rates. Therefore, our methods must be optimized to separate the effects of chs seed viability from fungal susceptibility, before a confident assignment of the flavonoid fungal resistance factor(s) can reliably be assigned. All developing Arabidopsis plants should be watered at optimal intervals, and should be separated by ecotype and accession to prevent overwatering of smaller plants; all seeds should be harvested immediately after the siliques begin to dry out. Additionally, under our experimental conditions, seeds in the Landsberg erecta genetic background were more susceptible to A. nidulans than seeds in the Col background. Our pathosystem must be further analyzed to determine if a second mutation in the ERECTA gene will consistently increase susceptibility of flavonoid mutant seeds.

Ultimately, the assessment of the chemical inhibitory properties of Arabidopsis seeds mutant

in different flavonoid pathway genes could help identify flavonoid molecules that actively inhibit Aspergillus fungus. This knowledge could aid in understanding the mechanisms underlying seed resistance to Aspergillus in agriculturally important crops.

REFERENCES

Abrahams, S., G.J. Tanner, P.J. Larkin, and A.R. Ashton (2002). Identification and biochemical characterization of mutants in the proanthocyanidin pathway in Arabidopsis. Plant Physiology 130, 561-576.

Abrahams, S., E. Lee, A. R. Walder, G. J. Tanner, P.J. Larkin, and A.R. Ashton (2003). The Arabidopsis TDS4 gene encodes leucoanthocyanidin dioxygenase (LDOX) and is essential for proanthocyanidin synthesis ad vacuole development. The Plant Journal 35, 624-636.

Adie, B.A.T., J. Pérez-Pérez, M.M. Pérez-Pérez, M. Godoy, J.J. Sánchez-Serrano, E.A. Schmelz, and R. Solano (2007). ABA is an essential signal for plant resistance to pathogens affecting JA biosynthesis and the activation of defenses in Arabidopsis. The Plant Cell 19, 1665-1681.

Amaike, S., and N.P. Keller (2011). Aspergillus flavus. Annual Review of Phytopathology 49, 107-133.

Appelhagen, I., K. Thiedig, N. Nordholt, N. Schmidt, G. Huep, M. Sagasser, B. Weisshaar (2014). Update on transparent testa mutants from Arabidopsis thaliana: characterisation of new alleles from an isogenic collection. Planta 240, 955-970.

Austin, M.B., and J.P. Noel (2003). The chalcone synthase superfamily of type III polyketide synthases. Natural Product Reports 20(1), 79-110.

Bailes, G., M. Lind, A. Ely, M. Powell, J. Moore-Kucera, C. Miles, D. Inglis, M. Brodhagen (2013). Isolation of native soil microorganisms with potential for breaking down biodegradable plastic mulch films used in agriculture. Journal of Visualized Experiments 75, e50373, DOI:10.3791/50373.

Baxter, I.R., J.C. Young, G. Armstrong, N. Foster, N. Bogenschutz, T. Cordova, W.A. Peer, S.P. Hazen, A.A. Murphy, and J.F. Harper (2005). A plasma membrane H+-ATPase is required for the formation of proanthocyanidins in the seed coat endothelium of Arabidopsis thaliana. Proceedings of the National Academy of Sciences (PNAS) 102(7), 2649-2654.

Bednarek, P., and P. Schulze-Lefert (2009). Role of plant secondary metabolites at the host- pathogen interface. Annual Plant Reviews 34, 220-260.

Bennett, J.W. (2010). An overview of the genus Aspergillus, p.1-17. In Masayuki Machinda, and Katsuya Gomi (eds.), Aspergillus: Molecular Biology and Genomics. Caister Academic Press, Norfolk, UK.

Bentsink, L., and M. Koornneef (2008). Seed dormancy and germination. In American Society of Plant Biologists, The Arabidopsis Book, DOI:10.1199/tab.0119.

Brown, R.L., Z.-Y. Chen, T.E. Cleveland, and J.S. Russin (1999). Advances in the development of host resistance in corn to aflatoxin contamination by Aspergillus flavus. Phytopathology 89(3), 113-117.

Chaudhury, A.M., A. Koltunow, T. Payne, M. Luo, M.R. Tucker, E.S. Dennis, and W.J. Peacock (2001). Control of early seed development. Annual Review of Cell and Developmental Biology 17, 677-699.

Chen, Z.-Y., R.L. Brown, K.E. Damann, and T.E. Cleveland (2004). Identification of a maize kernel stress-related protein and its effect on aflatoxin accumulation. Phytopathology 94, 938-945.

Chen, Z.-Y., R.L. Brown, K. Rajasekaran, K.E. Damann, and T.E. Cleveland (2006). Identification of a maize kernel pathogenesis-related protein and evidence for its involvement in resistance to Aspergillus flavus infection and aflatoxin production. Phytopathology 96, 87- 95.

Chen, Z.-Y., R.L. Brown, K.E. Damann, and T.E. Cleveland (2010). PR10 expression in maize and its effect on host resistance against Aspergillus flavus infection and aflatoxin production. Molecular Plant Pathology 11(1), 69-81.

Cole, R.J., R.A. Hill, P.D. Blankenship, and T.H. Sanders (1986). Color mutants of Aspergillus flavus and Aspergillus parasiticus in a study of preharvest invasion of peanuts. Applied and Environmental Microbiology 52(5), 1128-1131.

Cotty, P.J., and R. Jaime-Garcia (2007). Influences of climate on aflatoxin producing fungi and aflatoxin contamination. International Journal of Food Microbiology 119(1-2), 109-115.

Dean, G., Y.G. Cao, D.Q. Xiang, N.J. Provart, L. Ramsay, A. Ahad, R. White, G. Selvaraj, R. Datla, and G. Haughn (2011). Analysis of gene expression patterns during seed coat development in Arabidopsis. Molecular Plant 4(6), 1074-1091.

Debeaujon, I., K.M. Léon-Kloosterziel, and M. Koornneef (2000). Influence of the testa on seed dormancy, germination, and longevity in Arabidopsis. Plant Physiology 122, 403-413.

Debeaujon, I., A.J. Peeters, K.M. Leon-Kloosterziel, and M. Koornneef (2001). The TRANSPARENT TESTA12 gene of Arabidopsis encodes a multidrug secondary transporter- like protein required for flavonoid sequestration in vacuoles of the seed coat endothelium. The Plant Cell 13, 853-871.

Debeaujon, I., N. Nesi, P. Perez, M. Devic, O. Grandjean, M. Caboche, and L. Lepiniec (2003). Proanthocyanidin-accumulating cells in Arabidopsis testa: regulation of differentiation and role in seed development. The Plant Cell 15, 2514-2531.

Dixon, R.A., L. Achnine, P. Kota, C.-J. Liu, M.S. Srinivasa Reddy, and L. Wang (2002). The phenylpropanoid pathway and plant defence – a genomics perspective. Molecular Plant Pathology 3(5), 371-390.

Dlugokencky, E., and P. Tans (2016). Global greenhouse gas reference network: trends in atmospheric carbon dioxide. National Oceanic & Atmospheric Administration/ Earth System Research Laboratory (www.esri.noaa.gov/gmd/ccgg/trends/).

Dolezal, A.L., X. Shu, G.R. OBrian, D.M. Nielsen, C.P. Woloshuk, R.S. Boston, and G.A. Payne (2014). Aspergillus flavus infection induces transcriptional and physical changes in developing maize kernels. Frontiers in Microbiology 5(384), 1-10.

Funnell, D.L., and J.R. Pedersen (2006). Reaction of sorghum lines genetically modified for reduced lignin content to infection by Fussarium and Alternaria spp. Plant Disease 90, 331- 338.

Figueiredo, D.D., and C. Köhler (2014). Signaling events regulating seed coat development. Biochemical Society Transactions 42, 358-363.

Fountain, J.C., P. Khera, L. Yang, S.N. Nayak, B.T. Scully, R.D. Lee, Z.-Y. Chen, R.C. Kemerait, R.K. Varshney, and B. Guo (2015). Resistance to Aspergillus flavus in maize and peanut: Molecular biology, breeding, environmental stress, and future perspectives. The Crop Journal 3, 229-237.

Galagan, J.E., et al. (2005). Sequencing of Aspergillus nidulans and comparative analysis with A. fumigatus and A. oryzae. Nature 438, 1105-1115.

Giorni, P., P. Battilani, A. Pietri, and N. Magan (2008). Effect of aw and CO2 level on Aspergillus flavus growth and aflatoxin production in high moisture maize post-harvest. International Journal of Food Microbiology 122, 108-113.

Godiard, L., L. Sauviac, K.U. Torii, O. Grenon, B. Mangin, N.H. Grimsley, and Y. Marco (2003). ERECTA, an LRR receptor-like kinase protein controlling development pleiotropically affects resistance to bacterial wilt. The Plant Journal 36, 353-365.

Goto, N. (1985). A mucilage polysaccharide secreted from testa of Arabidopsis thaliana. Arabidopsis Information Service 22, 143-145.

Guo, B.Z., G. Xu, Y.G. Cao, C.C. Holbrook, and R.E. Lynch (2006). Identification and characterization of phospholipase D and its association with drought susceptibilities in peanut (Arachis hypogaea). Planta 223, 512-520.

Guo, B., X. Chen, P. Dang, B.T. Scully, X. Liang, C.C. Holbrook, J. Yu, and A.K. Culbreath (2008). Peanut gene expression profiling in developing seeds at different reproduction stages during Aspergillus parasiticus infection. BioMed Central Developmental Biology 8(12), DOI:10.1186/1471-213X-8-12. 37

Hatsugai, N., K. Yamada, S. Goto-Yamada, and I. Hara-Nishimura (2015). Vacuolar processing enzyme in plant programmed cell death. Frontiers in Plant Science 6(234), DOI:10.3389/fpls.2015.00234.

Haughn, G., and A. Chaudhury (2005). Genetic analysis of seed coat development in Arabidopsis. TRENDS in Plant Science 10(10), 472-477.

Hruz, T., O. Laule, G. Szabo, F. Wessendorp, S. Bleuler, L. Oertle, P. Widmayer, W. Gruissem, P. Zimmermann (2008). Genevestigator V3: A reference expression database for the meta-analysis of transcriptomes. Advances in Bioinformatics 2008 (420747), DOI:10.11 55/2008/420747.

Jaganathan, G.K. (2016). Influence of maternal environment in developing different levels of physical dormancy and its ecological significance. Plant Ecology 217, 71-79.

Kanwal, Q., I. Hussain, H.L. Siddiqui, and A. Javaid (2010). Antifungal activity of flavonoids isolated from mango (Mangifera indica L.) leaves. Natural Product Research 24(20), 1907-1914.

Kavousi, H.R., H. Marashi, J. Mozafari, and A.R. Bagheri (2009). Expression of phenylpropanoid pathway genes in chickpea defense against race 3 of Ascochyta rabiei. Plant Pathology Journal 8(3), 127-132.

Keller, N.P., R.A.E. Butchko, B. Sarr, and T.D. Phillips (1994). A visual pattern of mycotoxin production in maize kernels by Aspergillus spp. Phytopathology 84, 483-488.

Keller, N.P and T.H. Adams (1995). Analysis of a mycotoxin gene cluster in Aspergillus nidulans. National Center for Biotechnology Information 8, 14-21.

Kelley, R.Y., W.P. Williams, J.E. Mylroie, D.L. Boykin, J.W. Harper, G.L. Windham, A. Ankala, and X. Shan (2012). Identification of maize genes associated with host plant resistance or susceptibility to Aspergillus flavus infection and aflatoxin accumulation. PLoS ONE 7(5), DOI:10.1371/journal.pone.0036892.

King, S.B., and G.E. Scott (1982). Field inoculation techniques to evaluate maize for reaction to kernel infection by Aspergillus flavus. Phytopathology 72, 782-785.

Klich, Maren A., S. H. Thomas, and J. E. Mellon. Field studies on the mode of entry of Aspergillus flavus into cotton seeds. Mycologia 76(4), 665-669. 1984.

Klich, M.A., and M.A. Chmielewski (1985). Nectaries as entry sites for Aspergillus flavus in developing cotton bolls. Applied and Environmental Microbiology 50(3), 602-604.

Klich, M.A. (2007). Aspergillus flavus: the major producer of aflatoxin. Molecular Plant Pathology 8(6), 713-722.

Koornneef, M. (1990). Mutations affecting the testa colour in Arabidopsis. Arabidopsis Information Service 28, 1-4.

Landry, L.G., C.C.S. Chapple, and R.L. Last (1995). Arabidopsis mutants lacking phenolic sunscreens exhibit enhanced ultraviolet-B injury and oxidative damage. Plant Physiology 109, 1159-1166.

Larkindale, J., and M.R. Knight (2002). Protection against heat stress-induced oxidative damage in Arabidopsis involves calcium, abscisic acid, ethylene, and salicylic acid. Plant Physiology 128, 682-695.

Le, B.H., C. Cheng, A.Q. Bui, J.A. Wagmaister, K.F. Henry, J. Pelletier, L. Kwong, M. Belmonte, R. Kirkbride, S. Horvath, G.N. Drews, R.L. Fischer, J.K. Okamuro, J.J. Harada, and R.B. Goldberg (2010). Global analysis of gene activity during Arabidopsis seed development and identification of seed-specific transcription factors. Proceedings of the National Academy of Sciences 107(17), 8063-8070.

Lepiniec, L., I. Debeaujon, J.-M. Routaboul, A. Baudry, L. Pourcel, N. Nesi, and M. Caboche (2006). Genetics and biochemistry of seed flavonoids. Annual Review of Plant Biology 57, 405-430.

Li, J., T.-M. Ou-Lee, R. Raba, R.G. Amundson, and R.L. Last (1993). Arabidopsis flavonoid mutants are hypersensitive to UV-V irradiation. The Plant Cell 5, 171-179.

Li, Y., F. Beisson, M. Pollard, J. Ohlrogge (2006). Oil content of Arabidopsis seeds: the influence of seed anatomy, light and plant-to-plant variation. Phytochemistry 67(9), 904-915.

Liang, X.Q., M. Luo, and B.Z. Guo (2006). Resistance mechanisms to Aspergillus flavus infection and aflatoxin contamination in peanut (Arachis hypogaea). Plant Pathology Journal 5(1), 115-124.

Llorente, F., C. Alonso-Blanco, C. Sánchez-Rodriguez, L. Jorda, and A. Molina (2005). ERECTA receptor-like kinase and heterotrimeric G protein from Arabidopsis are required for resistance to the necrotrophic fungus Plectosphaerella cucumerina. The Plant Journal 43, 165-180.

Magan, N., R. Hope, V. Cairns, and D. Aldred (2003). Post-harvest fungal ecology: impact of fungal growth and mycotoxin accumulation in stored grain. European Journal of Plant Pathology 109, 723-730.

Marsh, S.F., and G.A. Payne (1984). Preharvest infection of corn silks and kernels by Aspergillus flavus. Phytopathology 74, 1284-1289.

Miedes, E., R. Vanholme, W. Boerjan, and A. Molina (2014). The role of the secondary cell wall in plant resistance to pathogens. Frontiers in Plant Science 5:358, DOI:10.3389/fpls. 2014.00358. 39

Mierziak, J., W. Wojtasik, K. Kostyn, T. Czuj, J. Szopa, A. Kulma (2014). Crossbreeding of transgenic flax plants overproducing flavonoids and glucosyltransferase results in progeny with improved antifungal and antioxidative properties. Molecular Breeding 34, 1917-1932.

Mironava, T., M. Hadjiargyrou, M. Simon, and M.H. Rafailovich (2012). The effects of UV emission from compact fluorescent light exposure on human dermal fibroblasts and keratinocytes in vitro. Photochemistry and Photobiology 88, 1497-1507.

Mitchell, J., I.G. Johnston, and G.W. Bassel (2016). Variability in seeds: biological, ecological, and agricultural implications. Journal of Experimental Biology, DOI: 10.1093/ jxb/erw397.

Moïse, J.A., S. Han, L. Gudynaite-Savitch, D.A. Johnson, and B.L.A. Miki (2005). Seed coats: structure, development, composition, and biotechnology. In Vitro Cellular & Developmental Biology – Plant 41, 620-644.

Naoumkina, M.A., Q. Zhao, L. Gallego-Giraldo, X. Dai, P.X. Zhao, and R.A. Dixon (2010). Genome-wide analysis of phenylpropanoid defense pathways. Molecular Plant Pathology 11(6), 829-846.

Nesi, N., I. Debeaujon, C. Jond, A.J. Stewart, G.I. Jenkins, M. Caboche, and L. Lepiniec (2002). The TRANSPARENT TESTA 16 locus encodes the ARABIDOPSIS BSISTER MADS domain protein and is required for proper development and pigmentation of the seed coat. The Plant Cell 14, 2463-2479.

Palma, K., M. Wiermer, and X. Li (2009). Marshalling the troops: intracellular dynamics in plant pathogen defense. Annual Plant Reviews 34, 177-219.

Payne, G.A. (1998). Process of contamination by aflatoxin-producing fungi and their impact on crops, p.279-306. In Kaushal K. Sinha, and Deepak Bhatnagar (eds.), Mycotoxins in Agriculture and Food Safety. Marcel Dekker, Inc., New York, USA.

Pechanova, O., T. Pechan, W.P. Williams, and D.S. Luthe (2011). Proteomic analysis of the maize rachis: Potential roles of constitutive and induced proteins in resistance to Aspergillus flavus infection and aflatoxin accumulation. Proteomics 11, 114-127.

Pelletier, M.K., and B.W. Shirley (1996). Analysis of flavanone 3-hydroxylase in Arabidopsis seedlings: coordinate regulation with chalcone synthase and chalcone isomerase. Plant Physiology 111, 339-345.

Pontecorvo, G., J.A. Roper, L.M. Chemmons, K.D. Macdonald, and A.W.J. Bufton (1953). The genetics of Aspergillus nidulans. Advances in Genetics 5, 141-238.

Pourcel, L., J.-M. Routaboul, L. Kerhoas, M. Caboche, L. Lepiniec, and I. Debeaujon (2005). TRANSPARENT TESTA10 encodes a laccase-like enzyme involved in oxidative polymerization of flavonoids in Arabidopsis seed coat. The Plant Cell 17, 2966-2980. 40

Rawal, H.C., N.K. Singh, and T.R. Sharma (2013). Conservation, divergence, and genome- wide distribution of PAL and POX A gene families in plants. International Journal of Genomics, DOI:10.1155/2013/678969.

Routaboul, J.-M., L. Kerhoas, I. Debeaujon, L. Pourcel, M. Caboche, J. Einhorn, L. Lepiniec (2006). Flavonoid diversity and biosynthesis in seed of Arabidopsis thaliana. Planta 224, 96- 107.

Saito, K., K. Yonekura-Sakakibara, R. Nakabayashi, Y. Higashi, M. Yamazaki, T. Tohge, and A.R. Fernie (2013). The flavonoid biosynthetic pathway in Arabidopsis: structural and genetic diversity. Plant Physiology and Biochemistry 72, 21-31.

Sánchez-Rodríguez, C., J.M. Estévez, F. Llorente, C. Hernández-Blanco, L. Jordá, I. Pagán, M. Berrocal, Y. Marco, S. Somerville, and A. Molina (2009). The ERECTA receptor-like kinase regulates cell wall-mediated resistance of pathogens in Arabidopsis thaliana. Molecular Plant-Microbe Interactions 22(8), 953-963.

Sanchis, V., and N. Magan (2004). Environmental conditions affecting mycotoxins, p.174- 189. In N. Magan and M. Olsen, (eds.), Mycotoxins in food: Detection and control. Woodhead Publishing Limited, Cambridge, England.

Schröder, J. (1997). A family of plant-specific polyketide synthases: facts and predictions. Trends in Plant Science 2(10), 373-378.

Seiler, C., V.T. Harshavardhan, K. Rajesh, P.S. Reddy, M. Strickert, H. Rolletschek, U. Scholz, U. Wobus, and N. Sreenivasulu (2011). ABA biosynthesis and degradation contributing to ABA homeostasis during barley seed development under control and terminal drought-stress conditions. Journal of Experimental Biology, DOI:10.1093/jxb/erq446.

Sétamou, M., K.F. Cardwell, F. Schulthess, and K. Hell (1997). Aspergillus flavus infection and aflatoxin contamination of preharvest maize in Benin. Plant Disease 81, 1323-1327.

Shaukat, S.S., M.A. Farooq, M.F. Siddiqui, and S. Zaidi (2013). Effect of enhanced UV-B radiation on germination, seedling growth and biochemical responses of Vigna mungo (L.) Hepper. The Pakistan Journal of Botany 45(3), 779-785.

Shephard, G.S. (2008). Impact of mycotoxins on human health in developing countries. Food Additives and Contaminants 25(2), 146-151.

Shirley, B.W., W.L. Kubasek, G. Storz, E. Bruggemann, M. Koornneef, F.M. Ausubel, and H.M. Goodman (1995). Analysis of Arabidopsis mutants deficient in flavonoid biosynthesis. The Plant Journal 8(5), 659-671.

Simbaya, J., B.A. Slominski, G. Rakow, L.D. Campbell, R.K. Downey, and J.M. Bell (1995). Quality characteristics of yellow-seeded Brassica seed meals: protein, carbohydrates, and dietary fiber components. Journal of Agricultural and Food Chemistry 43, 2062-2066. 41

Sobolev, V.S., B.Z. Guo, C.C. Holbrook, and R.E. Lynch (2007). Interrelationship of phytoalexin production and disease resistance in selected peanut genotypes. Journal of Agricultural and Food Chemistry 55, 2195-2200.

Starr, J.L., W. Yang, Y. Yan, F. Crutcher, and M. Kolomiets (2014). Expression of phenylalanine ammonia lyase genes in maize lines differing in susceptibility to Meloidogyne incognita. Journal of Nematology 46(4), 360-364.

Sullivan, J.H., D.S. Muhammad, K.M. Warpeha (2014). Phenylalanine is required to promote specific developmental responses and prevents cellular damage in response to ultraviolet light in soybean (Glycine max) during the seed-to-seedling transition. PLoS ONE 9(12), DOI: 10.1371/journal.pone.0112301.

The Arabidopsis Genome Initiative. (2000). Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408, 796-815.

Tohge, T., M. Watanabe, R. Hoefgen, and A.R. Fernie (2013). The evolution of phenylpropanoid metabolism in the green lineage. Critical Reviews in Biochemistry and Molecular Biology 48(2), 123-152.

Wang, H., Y. Lei, L. Wan, L. Yan, J. Lv, X. Dai, X. Ren, W. Guo, H. Jiang, and B. Liao (2016a). Comparative transcript profiling of resistant and susceptible peanut post-harvest seeds in response to aflatoxin production by Aspergillus flavus. BioMed Central, Plant Biology 16(54), DOI:10.1186/s12870-016-0738-z.

Wang, Q.-W., S. Nagano, H. Ozaki, S.-I. Morinaga, J. Hidema, K. Hikosaka (2016b). Functional differentiation in UV-B-induced DNA damage and growth inhibition between highland and lowland ecotypes of two Arabidopsis species. Environmental and Experimental Botany 131, 110-119.

Weidenbörner, M., H. Hindorf, H.C. Jha, and P. Tsotsonos (1990). Antifungal activity of flavonoids against storage fungi of the genus Aspergillus. Phytochemistry 29(4), 1103-1105.

Western, T.L., D.J. Skinner, and G.W. Haughn (2000). Differentiation of mucilage secretory cells of the Arabidopsis seed coat. Plant Physiology 122, 345-355.

Wicklow, D.T., D.M. Wilson, and T.C. Nelsen (1993). Survival of Aspergillus flavus sclerotia and conidia buried in soil in Illinois or Georgia. Phytopathology 83, 1141-1147.

Wilson, R.A., H.W. Gardner, and N.P. Keller (2001). Cultivar-dependent expression of a maize lipoxygenase responsive to seed infesting fungi. Molecular Plant-Microbe Interactions 14(8), 980-987.

Windsor, J.B., V.V. Symonds, J. Mendenhall, and A.M. Lloyd (2000). Arabidopsis seed coat development: morphological differentiation of the out integument. The Plant Journal 22(6), 483-493. 42

Winkel-Shirley, B. (2001). Flavonoid biosynthesis: A colorful model for genetics, biochemistry, cell biology, and biotechnology. Plant Physiology 126, 485-493.

Wotton, H.R., and R.N. Strange (1985). Circumstantial evidence for phytoalexin involvement in resistance of peanuts to Aspergillus flavus. Journal of General Microbiology 131, 487-494.

Xu, W., C. Dubos, and L. Lepiniec (2015). Transcriptional control of flavonoid biosynthesis by MYB-bHLH-WDR complexes. Trends in Plant Science 20(3), 176-185.

Yu, J., T.E. Cleveland, W.C. Nierman, and J.W. Bennett (2005). Aspergillus flavus genomics: gateway to human and animal health, food safety, and crop resistance to disease. Revista Iberoamericana de Micología 22, 194-202.

Zhao, J., Y. Pang, and R.A. Dixon (2010). The mysteries of proanthocyanidin transport and polymerization. Plant Physiology 153, 437-443.

FIGURES AND TABLES

Epidermal* OUTER*INTEGUMENT* Palisade* INNER*INTEGUMENT* Endothelium*

SEED*COAT*(testa)**

ENDOSPERM* DEVELOPING*EMBRYO*

Figure 1. The five cell layers of a developing Arabidopsis thaliana seed coat, five days after fertilization. Two layers make up the outer integument (the epidermal and palisade layers), and three cell layers make up the inner integument (the endothelium being the innermost layer) (adapted from Haughn and Chaudhury, 2005).

Figure 2. The flavan backbone and classes of flavonoids. a) Flavan backbone: flavonoids are derivatives of the fifteen carbon flavan molecule, which has two phenyl rings (A and B) and one heterocyclic ring with an oxygen at the 1 position (C). b) Classes of flavonoids: i. Flavonols: 3-hydroxy derivatives of flavan with a 2,3-double bond and a 4-carbonyl group. ii. Flavan-3-ols: 3-hydroxy derivatives of flavan. iii. Proanthocyanidins: 4,8 polymerized flavan derivatives. iv. Flavanones: 4-carbonyl derivatives of flavan. v. Flavan-3,4-diols: 3,4- hydroxy derivatives of flavan.

Table 1. Arabidopsis flavonoid biosynthetic pathway proteins, their genes and corresponding mutant seed lines, and the mutagen used to create each mutant line. Both the conventional and the transparent testa gene names are provided. All wildtype and mutant plant lines were acquired from the Arabidopsis Biological Resource Center. Columbia (Col) genetic background (blue); Landsberg erecta (Ler) background (green); Wassilewskija background (purple). None = No mutant line used. EMS = ethylmethane sulfonate; NMU = nitrosomethyl urea; ZFN = zinc finger nuclease.

Seed Lines Protein Name Gene Name(s) Mutagen Gene Mutation; Name Chalcone Synthase CHS; TT4 chs; tt4† (actually alb2•tt4-1) EMS tt4-1† EMS tt4-3† ZFN tt4-TDNA* T-DNA Chalcone Isomerase CHI; TT5 None None Flavone-3-Hydroxylase F3H; TT6 None None Flavonol Synthase FLS fls; fls* T-DNA Flavone-3’-Hydroxylase F3’H; TT7 f3’h; tt7† (actually alb2•tt7-1) EMS Dihydroflavanol Reductase DFR; TT3 dfr; tt3† X-rays Leucocyanidin Dioxygenase LDOX; TT18 ldox; tt18† T-DNA Anthocyanidin Reductase (Banyuls) BAN ban; ban† T-DNA Laccase-Like 15 LAC15; TT10 lac15; tt10† EMS Transparent Testa 15 TT15 tt15; tt15† T-DNA Glutathione S-Transferase PHI 12 GSTF12; TT19 gstf12; tt19† T-DNA Transparent Testa 12 TT12 tt12; tt12† T-DNA Autoinhibited H+-ATPase Isoform 10 AHA10; TT13 aha10; aha10-4† T-DNA MYB Domain Protein 123 MYB123; TT2 None None Basic Helix-Loop-Helix BHLH42; TT8 bhlh42; tt8† Unknown Transcription Factor 42 Transparent Testa Glabra 1 TTG1 None None Arabidopsis B-Sister Protein ABS; TT16 None None WIP Domain Protein 1 WIP1; TT1 wip1; tt1† X-rays WRKY Domain Transcription Factor 44 WRKY44; TTG2 None None Laccase 5 LAC5 lac5; lac5-1† T-DNA lac5-2† T-DNA Erecta ER er; Ler† Unknown er-1† T-DNA † Homozygous mutant seed line * Heterozygous mutant seed line

PAL p-Coumaroyl-CoA

CHS (tt4)

Krebs 3X Malonyl-CoA Cycle Chalcone

CHI (tt5)

Naringenin

F3H (tt6)

FLS Dihydrokaempferol Kaempferol FLAVONOLS (fls)

F3’H (tt7)

FLS Dihydroquercetin Quercetin FLAVONOLS (fls) LAC15 BIFLAVONOLS DFR (tt10) (tt3)

LDOX BAN LAC15 Epicatechin-based Leucocyanidin Cyanidin Epicatechin (tt18) (ban) (tt10) Oxidized Proanthocyanidins

LAC15 (tt10)

MODIFIERS TRANSPORTERS REGULATORS

TT15 (tt15) TT12 (tt12) MYB123 (tt2) ABS (tt16) GSTF12 (tt19) AHA10 (tt13) BHLH42 (tt8) WIP1 (tt1) TTG1 (ttg1) WRKY44 (ttg2)

Figure 3 (previous page). The flavonoid biosynthetic pathway in Arabidopsis (Figure adapted from Winkel-Shirley, 2001; Routaboul et al, 2006). Molecular changes are highlighted red within each molecular structure. Mutant seed lines are indicated in parentheses below each protein name. Mutant lines whose seeds are infected by A. nidulans (red); resistant seed lines (blue); untested seed lines (black). A proposed novel branch of the pathway is indicated in grey.

Figure 4. Levels of expression of the seventeen laccase genes in Arabidopsis (Hruz et al., 2008). Laccase-like 15 (LAC15) and laccase 5 (LAC5) genes increase in expression at the end of seed maturation, and have higher expression rates in mature siliques than other known laccase enzymes. LAC15 is involved in flavonoid biosynthesis, but the function of LAC5 is not yet known.

Sterol Glucoside UDP-Glucose

Acyl Prenyl Methyl

Glutathione

Figure 5. Molecular changes made by flavonoid modifying enzymes. a) Substrates for flavonoid glycosyltransferase (TT15). b) Molecular groups added on to flavonoid scaffold molecules. c) Substrate for glutathione S-transferase PHI 12 (GSTF12).

5 3 4

5 3 2 4 1 2 3 6 5 2 4 6 5 7

Figure 6. Ages of siliques and seeds along the branches of an Arabidopsis thaliana plant. Black numbers represent the week after transplanting in which that silique began to develop. When seeds are whole-plant-harvested, all seeds from the entire parent plant are mixed together, and seed ages span a range of about a seven-week difference. When seeds are silique-harvested, the age range of seeds in each collection tube spans only a single week.

a) b) c) A. flavus A. nidulans * 6E+4 6E+4 6E+4

SEM) 5E+4 5E+4 5E+4 - * 4E+4 4E+4 4E+4

3E+4 3E+4 3E+4

2E+4 2E+4 2E+4

1E+4 1E+4 1E+4 Spores/Seed (+/ 0E+0 0E+0 0E+0 Wt Col Wt Ler No-seed WtWt ColCol chschs Wt ColWt Col chschs Control

d) 6E+4 *

5E+4 SEM) - 4E+4

3E+4

2E+4

1E+4

Spores/Seed (+/ 0E+0 Wt Col er (in Col) Wt Ler er (in Ler) chs Wt Col er (in Col) Wt Ler er (in Ler) chs

Figure 7. A two-model pathosystem: Arabidopsis seeds and Aspergillus nidulans. a) Rates of A. nidulans infection of wildtype (Wt) Arabidopsis seeds from both the Columbia (Col) and Landsberg (Ler) ecotypes. Seeds were whole-plant-harvested and used about 850 days after flowering. N=3 for each treatment. b) Rates of Aspergillus flavus infection of Arabidopsis Wt Col and mutant chs seeds (in the Ler genetic background). Seeds were whole-plant- harvested and used about 371 days after flowering. N=5 for each treatment. c) Rates of A. nidulans infection of Arabidopsis Wt Col and mutant chs seeds. Seeds were whole-plant- harvested and used about 371 days after flowering. N=5 for each treatment. d) Rates of A. nidulans infection of Arabidopsis seeds mutant in the erecta (er) gene, in both the Col and Ler genetic backgrounds, compared to a chs positive control. Seeds were whole-plant- harvested and used about 70 days after flowering. N=5 for all treatments. All assays were carried out in the dark at 37˚C for A. nidulans, or at 28˚C for A. flavus. Quantification of infection is represented by the number of spores per seed, plus or minus the standard error of the mean (SEM). A two-tailed t-test with an alpha value of 0.05 was run, comparing each infection rate to that of the no-seed control (a) or to that of Wt Col (b, c, d). P-values below 0.05 (*) are indicated above each treatment. Each no-spore control was clear of infection (data not shown). Each experiment was repeated twice with similar results.

a) * b) * c) 6E+4 6E+4 6E+4

5E+4 5E+4 5E+4 SEM)

- 4E+4 4E+4 4E+4 * 3E+4 3E+4 3E+4 2E+4 2E+4 2E+4 1E+4 1E+4 1E+4 0E+0 Spores/Seed (+/ 0E+0 0E+0 Wt Col chs Wt Col Wt Col f3'h Wt ColWt Col dfr d) e) f) 6E+4 6E+4 6E+4

5E+4 5E+4 5E+4 SEM)

- 4E+4 4E+4 4E+4 3E+4 3E+4 3E+4 2E+4 2E+4 2E+4 1E+4 1E+4 1E+4

Spores/Seed (+/ 0E+0 0E+0 0E+0 Wt Col Wt Col flsfls Wt Col ldox Wt ColWt Col ban g) h) i) 6E+4 6E+4 6E+4 5E+4 5E+4

SEM) 5E+4 - 4E+4 4E+4 4E+4 3E+4 3E+4 3E+4 2E+4 2E+4 2E+4 1E+4 1E+4 1E+4 0E+0 0E+0

Spores/Seed (+/ 0E+0 Wt Col Wt Col tt15 Wt Col Wt Col gstf12gstf12 Wt Col Wt Col tt12tt12 j) k) l) 6E+4 6E+4 6E+4

SEM) 5E+4 5E+4 5E+4 - 4E+4 4E+4 * 4E+4 3E+4 3E+4 3E+4 2E+4 2E+4 2E+4 1E+4 1E+4 1E+4 0E+0 0E+0 0E+0 Spores/Seed (+/ Wt Col Wt Col aha10 Wt Col Wt Col wip1wip1 Wt Col Wt Col bhlh42bhlh42 m) 6E+4

SEM) 5E+4 - 4E+4 3E+4 2E+4 * 1E+4

Spores/Seed (+/ 0E+0 Wt Col Wt Col lac15

Figure 8 (previous page). A. nidulans infection of Arabidopsis seeds mutant in flavonoid biosynthetic pathway genes, compared to the infection rate of Wt Col seeds. All seeds were whole-plant-harvested. a) chs seeds, 371 days after flowering; N=5. b) f3’h seeds, 133 days after flowering; N=3. c) dfr seeds, 133 days after flowering; N=5. d) fls seeds, 260 days after flowering; N=3. e) ldox seeds, 260 days after flowering; N=3. f) ban seeds, 140 days after flowering; N=5. g) tt12 seeds, 154 days after flowering; N=5. h) aha10 seeds, 581 days after flowering; N=5. i) tt15 seeds, 140 days after flowering; N=3. j) gstf12 seeds, 140 days after flowering; N=3. k) wip1 seeds, 847 days after flowering; N=3. l) bhlh42 seeds, 140 days after flowering; N=4. m) lac15 seeds, 581 days after flowering; N=5. A two-tailed t-test with an alpha value of 0.05 was run, comparing the infection rate of each mutant seed type to that of Wt Col seeds. P-values below 0.05 (*) are indicated above each treatment. All experiments were repeated a minimum of two times with similar results.

6E+3 * 5E+3

SEM) * - 4E+3

3E+3

2E+3

Spores/Seed (+/ 1E+3

0E+0 Wt ColWt Col lac15lac15 lac5lac5--11 lac5lac5--22 chschs

Figure 9. Rates of A. nidulans infection of laccase mutant seeds (lac15 and lac5), compared to Wt Col and a chs positive control. Seeds were whole-plant-harvested and used 70 days after flowering. Each no-spore control was clear of infection (data not shown). N=3 for each treatment. A two-tailed t-test with an alpha value of 0.05 was run, comparing the infection rate of each mutant seed type to that of Wt Col seeds. P-values below 0.05 (*) are indicated above each treatment. The experiment was repeated twice with similar results.

6E+4 * 5E+4 SEM) - 4E+4

3E+4 * 2E+4

Spores/Seed (+/ 1E+4

0E+0 Wt Col Wt Col young chschs old chschs (younger) (older)

Figure 10. Rates of A. nidulans infection of chs mutant seeds of different ages. Wt Col and younger chs seeds were used 50 days after flowering, and older chs seeds were used 350 days after flowering. All seeds were whole-plant-harvested. Each no-spore control was clear of infection (data not shown). N=5 for each treatment. A two-tailed t-test with an alpha value of 0.05 was run, comparing the infection rate of each chs mutant seed cohort to that of Wt Col seeds. P-values below 0.05 (*) are indicated above each treatment. The experiment was repeated twice with the same seed cohorts, with similar results.

4E+4

3E+4

Wt Col 3E+4 chs SEM) - Positive Control 2E+4 No-seed Control

2E+4

1E+4 Spores/Seed (+/

5E+3

0E+0 0 20 40 60 80 100 120 140 Days After Flowering

Figure 11. Timecourse of A. nidulans infection of Arabidopsis Wt Col and chs seeds at different ages of maturity (between zero and 140 days after flowering). Wt Col and chs seeds were marked weekly and were harvested silique-by-silique one week after peduncle collapse. The positive control was older chs seeds (about 645 days after flowering) that were whole- plant harvested. Even the oldest mature seeds of both Wt Col and chs remained resistant to A. nidulans, while the positive control chs seeds were mildly susceptible. Each no-spore control was clear of infection (data not shown). N=3 for each treatment.

chs (silique-harvested) chs (whole-plant-harvested)

b) 6E+4

5E+4 SEM) - 4E+4 * 3E+4

2E+4 Spores/Seed (+/ 1E+4

0E+0 Silique-harvested Wt Col Silique-harvested chs Whole-plant-harvested Whole-plant-harvested Wt Col chs Wt Col Wt Col chs chs (silique- (silique- (whole-plant- (whole-plant- harvested) harvested) harvested) harvested)

c) 100 90 80 70 60 50 40 30 Percent of seeds 20 10 ND ND 0 1 chs 2 3 chs 4 (silique-harvested) (whole-plant-harvested)

Fully germinated Radical emergence only Failed to germinate

Figure 12 (previous page). Comparison of chs seeds of the same age, but with different plant- care methods. a) A comparison of silique-harvested and whole-plant-harvested chs seeds (about 315 days after flowering). The yellow silique-harvested seeds were harvested one week after peduncle collapse; the grey-yellow whole-plant-harvested seeds were harvested after eight months of drying on the parent plant. Image was acquired with a magnified iPad camera, with a dissecting pin separating the two seed cohorts. b) A. nidulans infection rates of the chs seeds pictured above. Each no-spore control was clear of infection (data not shown). N=5 for each treatment. A two-tailed t-test with an alpha value of 0.05 was run, comparing the infection rate of chs mutant seeds to that of Wt Col seeds from the same cohort. P-values below 0.05 (*) are indicated above each treatment. The experiment was repeated two times, with similar results. c) Germination rates of the chs seeds shown above. N=3, with 30-40 seeds per plate. The germination test was repeated twice with similar results. Data that was not collected is indicated as not done (ND).

a) c)

* 6E+4

SEM) 5E+4 - 4E+4

3E+4

b) 2E+4

1E+4 Spores/Seed (+/ 0E+0 Wt Col tt4 (y)chs tt4 (s/d)chs (yellow) (late-forming)

d) 100 90 80 70 (middles of branches) chs 60

50 40 Percent of seeds 30 20 10 ND 0 1 chs 2 chs3 (yellow) (late-forming)

Fully germinated Radical emergence only Failed to germinate chs (ends of branches)

e) chs (yellow) chs (late-forming) 97% Germination 15% Germination

chs (late-forming branches)

Figure 13 (previous page). Comparison of chs seeds from the same plant, but formed at different developmental stages of the parent plant. a) A yellow chs seed (left) compared to a smaller, darker chs seed (right). Both seed phenotypes can occur on a single chs parent plant. Image acquired using a magnified iPad camera. b) Dissection of a chs parent plant. Seeds from the middles of the parent plants’ main branches (top); Seeds from the last five siliques at the ends of the parent plants’ main branches (middle); Seeds from late-forming siliques, which often form in tight clusters after the plant has been taken off of water (bottom). The late-forming seeds are an average of 44% smaller than seeds from other branches of the plant. Seed sizes were measured using ImageJ surface area analysis. Images were acquired with a SpotBasic camera and software, under the same lighting and magnification. The scale bar represents one millimeter. c) Rates of A. nidulans infection of sorted chs seeds (645 DAF), with yellow seeds separated from the late-forming seeds, which are smaller and darker. Each no-spore control was clear of infection (data not shown). N=3 for each treatment. A two-tailed t-test with an alpha value of 0.05 was run, comparing the infection rate of each chs seed type to that of Wt Col seeds. P-values below 0.05 (*) are indicated above each treatment. Experiment was repeated twice with similar results. d) Germination rates of the sorted chs seeds. N=3 for each seed type, with 35-50 seeds per plate. Germination test was repeated twice with similar results. Data that was not collected is indicated as not done (ND). e) The chs seedlings from yellow seeds (left) grew well, with a diameter up to 1.3 centimeters after two weeks of growth. The chs seedlings from the late-forming seeds did not grow past three millimeters in diameter. The scale bar represents one centimeter.

6.E+4

5.E+4 SEM) - 4.E+4

3.E+4 *

2.E+4 * Spores/Seed (+/ 1.E+4

0.E+0 Wt Col Wt LerWt Ler chs-1 (Ler)chs chs chs chschs-2 (Ler) chs-3 (Col) chs-4 (Col) (Ler; tt4) (Ler; tt4-1) (Col; tt4-3) (Col; tt4-TDNA)

b) 100 90 80 70 60 50 40 30 Percent of seeds 20 10 0 Wt Col Wt Ler Wt Ler chs-chs chs chs chs1 (Ler) chs-2 (Ler) chs-3 (Col) chs-4 (Col) (Ler; tt4) (Ler; tt4-1) (Col; tt4-3) (Col; tt4-TDNA) Fully germinated Radical emergence only Failed to germinate

Figure 14. Comparison of chs seeds with identical plant-care methods, but with different genetic backgrounds. Seeds were whole-plant-harvested after five weeks of drying, and were used in the assays 70 days after flowering. a) Rates of A. nidulans infection of four different chs mutant seeds compared to Wt Col and Wt Ler. Each no-spore control was clear of infection (data not shown). N=5 for each treatment. A two-tailed t-test with an alpha value of 0.05 was run, comparing the infection rate of each seed type to that of Wt Col seeds. P-values below 0.05 (*) are indicated above each treatment. Experiment was repeated twice with similar results. b) Germination rates of each seed type. N=3, with 20-60 seeds per plate. Experiment was repeated twice with similar results.

6E+7 *

5E+7 SEM)

- 4E+7 (+/

2 3E+7 cm 2E+7

Spores/ 1E+7

0E+0 Wt Col NE

Figure 15. Inhibition of A. nidulans growth in the presence of Wt Col aqueous seed extract, compared to a no-extract (NE) water control. A. nidulans growth was quantified by spores/cm2. The experiment was run for forty-five hours at 37˚C, in the dark. N=5 for each treatment. A two-tailed t-test with an alpha value of 0.05 was run, comparing A. nidulans growth in the presence of water compared to growth in the presence of Wt Col seed extract. A p-value below 0.05 (*) is indicated. The experiment was repeated three times with new seed extract, with similar results.

a) 6E+4

5E+4 SEM) - 4E+4

* * 3E+4 * * 2E+4 *

Spores/Seed (+/ * 1E+4 * * * *

0E+0 Wt Col - TC (F:150910)tt4 (0.5 yrs) tt4 (1yr) tt4 (1.5 yrs) tt4 (2 yrs) tt4 (2.5 yrs) tt4 (3 yrs) tt4 (3.5 yrs)tt4 - TC (F:151008)tt4 - TC (F:150910)tt4 - TC (F:150813)tt4 Bulk (2.5 yrs)

b) 100

Percent of seeds 20 ND ND 0 Wt Col - TC tt4 (0.5 yrs) tt4 (1yr) tt4 (1.5 yrs) tt4 (2 yrs) tt4 (2.5 yrs) tt4 (3 yrs) tt4 (3.5 yrs) tt4 - TC tt4 - TC tt4 - TC tt4 - Bulk (F:150910) (F:151008) (F:150910) (F:150813) (2.5 yrs) Failed to Germinate Radical Emergence only Fully Germinated

c) 6E+7 * * 5E+7 *

SEM) 4E+7 - (+/

2 3E+7

2E+7 *

Spores/cm 1E+7 * * * * * 0E+0 Wt Col - TC tt4 (0.5 yrs) tt4 (1 yr) tt4 (1.5 yrs) tt4 (2 yrs) tt4 (2.5 yrs) tt4 (3 yrs) tt4 (3.5 yrs) tt4 TC - tt4 TC - tt4 TC - tt4 Bulk - NE (F:150910)Wt Col WPH WPH WPH WPH WPH WPH WPH (F:151008)SH (F:150910)SH (F:150813)SH (2.5 yrs)BH NE chs chs chs chs chs chs chs chs chs chs chs (0.5yr) (1yr) (1.5yr) (2yr) (2.5yr) (3yr) (3.5yr) (1yr) (1yr) (1yr) (2.5yr)

Figure 16 (previous page). Variability of chs seed infection, germination, and chemical inhibition of A. nidulans. Comparisons are made between whole-plant-harvested (WPH) chs seeds of different ages, silique-harvested (SH) chs seeds, and bulk-harvested (BH) seeds that served as a positive control. a) Rates of A. nidulans infection of chs seeds compared to Wt Col seeds. Each no-spore control was clear of infection (data not shown). N=5 for each treatment. A two-tailed t-test with an alpha value of 0.05 was run, comparing the infection rate of each chs mutant seed type to that of Wt Col seeds. P-values below 0.05 (*) are indicated above each treatment. Experiment was repeated twice for each seed type, with similar results. b) Germination rates of chs and Wt Col seeds. N=3 for each seed type, with 30-70 seeds per plate. Germination test was repeated twice for each seed type, with similar results. Data that was not collected is indicated as not done (ND). c) Inhibitory properties of aqueous extracts from Wt Col and chs seeds, compared to a no-extract (NE) water control. The experiment was run for forty-five hours at 37˚C, in the dark. N=5 for each treatment. A two-tailed t-test with an alpha value of 0.05 was run, comparing A. nidulans growth in the presence of chs seed extracts compared to A. nidulans growth in the presence of Wt Col seed extract. A p-value below 0.05 (*) is indicated above each treatment. The experiment was repeated twice with the same seed extracts, showing similar results.