Molecular Mechanisms for Regulating Seed Amino Acid Levels in Maize (Zea Mays) Marie Adams Hasenstein Iowa State University

Iowa State University Capstones, Theses and Retrospective Theses and Dissertations Dissertations

2007 Molecular mechanisms for regulating seed amino acid levels in maize (Zea mays) Marie Adams Hasenstein Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/rtd Part of the Agricultural Science Commons, Agronomy and Crop Sciences Commons, and the Genetics and Genomics Commons

Recommended Citation Hasenstein, Marie Adams, "Molecular mechanisms for regulating seed amino acid levels in maize (Zea mays)" (2007). Retrospective Theses and Dissertations. 14677. https://lib.dr.iastate.edu/rtd/14677

This Thesis is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Retrospective Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. Molecular mechanisms for regulating seed amino acid levels in maize (Zea mays)

Marie Adams Hasenstein

A thesis submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

MASTER OF SCIENCE

Major: Genetics

Program of Study Committee: Paul Scott, Co-Major Professor Erik Vollbrecht, Co-Major Professor Phil Becraft

Iowa State University

Ames, Iowa

2007

UMI Microform 1447529 Copyright 2008 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.

ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, MI 48106-1346 ii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ...... iii ABSTRACT ...... iv CHAPTER 1: GENERAL INTRODUCTION ...... 1 OBJECTIVE ...... 1 DISSERTATION ORGANIZATION ...... 2 LITERATURE REVIEW ...... 3 Importance of Maize for Food/Feed ...... 3 Lysine ...... 4 Methionine ...... 18 REFERENCES ...... 26 CHAPTER 2: MOLECULAR CHARACTERIZATION OF ZMBZIP1, A DOWNSTREAM TARGET OF THE OPAQUE-2 TRANSCRIPTION FACTOR ...... 35 ABSTRACT ...... 35 MATERIALS AND METHODS ...... 36 ZmbZIP1 Expression Levels ...... 36 Physical Mapping of the ZmbZIP1 Gene ...... 39 Creation of a ZmbZIP1 TILLING Mutant to Attempt to Alter ZmbZIP1 Protein Expression ...... 41 DNA Isolation ...... 41 RESULTS AND DISCUSSION ...... 46 ZmbZIP1 Has Varying Transcript Levels in Different Tissues ...... 46 Physical Mapping of the ZmbZIP1 Locus ...... 52 SNaPshot Genotyping ...... 57 CONCLUSION ...... 59 REFERENCES ...... 60 CHAPTER 3: MOLECULAR MECHANISM OF METHIONINE DIFFERENTIATION IN THE BS11 HIGH AND LOW MET MAIZE LINES ...... 62 ABSTRACT ...... 62 MATERIALS AND METHODS ...... 62 Differential Expression of Cys2 and Dzs10 via Quantitative Real-Time RT-PCR ...... 64 Cys2 and Dzs10 Sequencing and Allele Comparison ...... 65 Amino Acid Assay to Determine Kernel Methionine Levels ...... 66 RESULTS AND DISCUSSION ...... 67 The BS11HM Population Contains Significantly More Methionine than BS11LM ...... 67 Correlation Between Population Methionine Levels and Cys2 or Dzs10 mRNA Levels 69 Correlation Between Kernel Methionine and Cys2 or Dzs10 Haplotypes ...... 73 CONCLUSION ...... 76 REFERENCES ...... 78 GENERAL CONCLUSIONS ...... 79 iii

ACKNOWLEDGEMENTS

I would like to express my gratitude to a number of people who have been indispensible to me during my graduate career. First, to Dr. Paul Scott, my major professor, for his supervision, encouragement and enthusiasm for the work that I was doing, as well as for our many discussions on statistical analysis. I would also like to thank my committee

members Drs. Erik Vollbrecht and Phil Becraft for their guidance, concern about my research, and availability throughout my time at Iowa State. Special thanks are extended to

Anastasia Bodnar and Dr. Adrienne Moran-Lauter for their help with HPLC and amino acid analysis, both of which were instrumental for the completion of my project. On a more personal level, I would like to thank my husband, Jason Hasenstein, for his constant support through both my best and worst days, and to Bart and Michelle Adams, Lisa Haney, Jamie

O’Rourke, Oliver Couture, and Ryan Benson for their friendship and calming and positive influences on me as I was completing my research. iv

ABSTRACT

The majority of worldwide maize production is used for food or feed; as such it is imperative that its amino acid balance mirror the needs of humans and animals. However, lysine and methionine, two essential amino acids necessary for proper growth and development, are lacking in the maize endosperm. In this thesis, molecular mechanisms to increase both lysine and methionine deposition in the kernel are explored.

ZmbZIP1, a maize transcription factor, was identified through differential expression in a microarray experiment as a potential candidate to increase the lysine content of the endosperm while lessening the pleiotrophic effects of prior candidates, such as Opaque-2; however, RT-PCR shows this level of differential expression may be genotype specific.

ZmbZIP1 was physically mapped to maize chromosome 9, contig 380, and a rice orthologue,

RISBZ5 was found through synteny. Expression data shows that ZmbZIP1 mRNA is present in all tested maize tissues to some extent. An attempt to make a ZmbZIP1 knock- down/knock-out mutant did not result in any phenotypic changes in the maize plant as a whole. Further no changes were seen in kernel alpha-zein deposition, which is the main determinant of endosperm lysine content. These results suggest either a functionally redundant protein to ZmbZIP1 exists, the knock-out alleles retained sufficient activity to be functional, or phenotypes produced by disruption of ZmbZIP1 are not detectable by the methods we used. In all cases, these results suggest more work must be done to determine the efficacy of ZmbZIP1 in regulating endosperm lysine levels.

A molecular mechanism for increasing methionine deposition into maize endosperm was investigated with regards to two candidate genes: the 10kDa delta-zein (Dzs10), a v

methionine-rich storage protein in endosperm, and cysteine synthase (Cys2), a key regulator

in cysteine/methionine biosynthesis. mRNA expression levels and haplotypes were

determined for each population, and were correlated to the average methionine content. At

the mRNA level, Dzs10 was found to be differentially expressed in maize endosperm, while

Cys2 is likely not differentially expressed. Further, a correlation of 0.4535 was found

between Dzs10 mRNA and kernel methionine levels, suggesting part of the increase in kernel methionine deposition may be due to a change in Dzs10 expression. In Cys2, a strong relationship has been noted between the level of kernel methionine deposition and a G/A

SNP present in the gene, suggesting that different segregations of the Cys2 alleles may also alter the kernel methionine levels. 1

CHAPTER 1: GENERAL INTRODUCTION

Objective The objective of this thesis is to explore novel mechanisms to regulate kernel lysine

and methionine levels with the goal of improving maize grain quality. Since the opaque-2

knockout mutation strongly affects kernel lysine levels, but has a plethora of negative

pleiotrophic effects on plant hardiness, we aim to characterize a likely downstream target of

the Opaque-2 protein, ZmbZIP1, to determine its usefulness in altering lysine levels with

fewer adverse phenotypic side effects. We wish to determine where this gene is located in

the maize genome, as well as to further understand its tissue specificity, especially in relation

to the presence or absence of the opaque-2 transcription factor protein. We also wish to

determine whether mutations in the ZmbZIP1 gene created using different forms of maize

mutagenesis are able to create a visible phenotypic change in otherwise wildtype maize plants.

In our second objective we aim to characterize the mechanism responsible for controlling levels of methionine in two random-mated maize populations (BS11HM and

BS11LM) that have undergone six cycles of divergent selection for grain methionine content.

In order to do so, we have chosen to study cysteine synthase (Cys2), a major control point in

cysteine biosynthesis, and the 10 kDa delta-zein (Dzs10), the methionine rich endosperm

storage protein in maize as candidate genes. These genes were chosen because they represent

biological processes occurring at the source (Cys2) or sink (Dzs10). Two potential levels of

regulation that may be responsible for the differing methionine contents will be investigated:

allelic differences and variation in mRNA expression levels. These results will then be

correlated to methionine levels in the endosperm. 2

Dissertation Organization This thesis consists of a literature review and the materials and methods, results, and discussion section of two papers. These papers, when submitted, will also include a condensed version of all applicable sections of the literature review as an introduction section. The literature review gives a general discussion of maize quality traits with respect to two amino acids; lysine and methionine. In the lysine-related section, maize prolamin proteins, also known as zeins, are described, as well as the most well known zein-modifying gene, opaque-2. It also discusses unpublished work done by a prior lab member to introduce the gene central to this thesis, ZmbZIP1. The manuscript reports the location of the

ZmbZIP1 gene in the maize genome, and the results of analyses of the levels of ZmbZIP1

RNA present in a sampling of maize tissues from both wildtype and an opaque-2 maize lines.

It also discusses the attempts made to determine a phenotype related to mutations in the

ZmbZIP1 gene. The methionine literature review discusses sulfate assimilation in plants, its

incorporation into methionine, and its resulting deposition in the maize kernel in zein storage proteins, as well as describing two candidate genes, Cys2 and Dzs10 that may play a role in this process. The related manuscript reports the associations between Cys2 and Dzs10 and methionine deposition into the maize kernel in high and low methionine maize populations; based upon these findings, Cys2 is hypothesized to have allelic differences that may encode enzymes with different activities in the two populations and Dzs10 is surmised to affect methionine deposition by increased mRNA production. 3

Literature Review

Importance of Maize for Food/Feed Cereal grains, such as maize, are both a protein- and starch-rich source of nutrition

due to their large energy stores in the endosperm; they therefore play an important role in

both human and livestock nutrition (NELSON 1969). Maize, along with wheat and rice, constitute the three most important cereals grown in the world and together these cereals

account for over 70% of the total cereal production (FAO 1999). Maize is a vital crop for the

nutrition of people around the world. Cereal grains such as maize make up a significant

portion of the diet in developing countries, making them a vital source of both carbohydrate

energy and protein (FAO 1992). People in 22 countries worldwide, mostly in Latin America

and sub-Saharan Africa, have a maize intake level between 300 and 1000 kcal per person per

day, thus providing a significant portion of daily calories and protein (FAOSTAT 2001).

Globally, rates of maize consumption have been rising since 1961 from 158 to 162 kcal/day

in 2001 (FAOSTAT 2001). Maize is also a necessary component of animal diets, comprising

up to 30% of protein and 90% of starch in dairy cows (DADO 1999). 70-80% of maize

produced is used as an animal feed ingredient, and the demand for maize products for livestock feed has been rising, especially in South East Asia, where it was predicted that an additional 4.1-16.1 Mt would be necessary by 2004 (KOBAYASHI 1998). This is mainly due to an increase in meat, egg and milk consumption in the developing world (DELGADO et al.

1999). Thus, due to both their growing economic and nutritional importance, maize proteins

continue to be a major subject for scientific study.

One of the limitations of maize grain is that it is not nutritionally compatible with the

nutritional requirements of monogastric animals, a category that includes most livestock 4

animals and humans, particularly in respect to the essential amino acids lysine and

methionine. Lysine levels in maize are below the required minimums set for both human and

monogastric livestock growth and development (FAO/WHO/UNU 1985). Methionine levels,

especially when maize is combined with low-methionine containing legumes such as soy in

animal feeds, are also substandard in animal diets (KRISHNAN 2005). Plant breeding efforts

are being focused on increasing these amino acids to levels compatible with both

monogastric livestock and human nutritional needs.

Lysine

Maize Prolamins – the Zeins The major determinants of amino acid content in maize are the storage proteins located in the maize endosperm (LARKINS 1986). The most abundant storage proteins found

in cereals are the alcohol-soluble storage proteins, or prolamins. Maize prolamins consist of

a group of polypeptides called zeins. Zeins are synthesized in the rough endoplasmic

reticulum and are deposited in the seed as relatively large, insoluble protein bodies that make

up about 10-15% of the endosperm by volume (LARKINS and HURKMAN 1978). Storage

bodies, including zeins, have no enzymatic activities in the developing seed; instead, they are

hypothesized to provide nitrogen for the developing plant through protein catabolism

(LARKINS et al. 1984). Amino acid contents vary between each zein; the typical zein is

particularly rich in proline, leucine, glutamine, alanine and serine, with these amino acids

making up 70% of the total content. Conversely, the zeins are deficient in the essential amino

acids lysine and tryptophan and α- and γ- zeins are particularly low in methionine (KIRIHARA

1988). 5

Zeins are classified by their molecular mass: proteins found at 19 and 22 kDaltons

are denoted α-zeins, at 15 kDa are found the β-zeins, at 16, 27 and 50 kDa are the γ-zeins

(PRAT 1987) and at 10 kDa are the δ-zeins (LARKINS et al. 1993). High performance liquid

chromatography (HPLC) has been suggested as a robust and precise way to separate and

characterize zeins in maize endosperm (WILSON 1991).

Αlpha-zeins, the most abundant group of maize prolamins, are encoded by two large gene families. The 22kDa protein is encoded by 23 genes, 22 of which are located in tandem along a stretch of chromosome 4s, with the 23rd located ~20cM closer to the centromere on

the same chromosome. However, only seven of these genes have been shown to be

transcriptionally active. (SONG et al. 2001). The 19kDa family is even more diverse and is comprised of over 50 related genes that span three of the 10 maize chromosomes. Only half

of these genes seem to be active (HEIDECKER and MESSING 1986). Both the 19 and 22kDa

proteins have a unique structure that contains a repetitive amino acid sequence

(LQQF/LLPA/FNQLA/LA/VANSPAYLQQ). This motif is predicted to form α-helices to

create an easy to pack, rod shaped protein molecule in the protein bodies (ARGOS et al.

1982).

In contrast to α-zeins, β-zein proteins are encoded by only one gene (PEDERSEN et al.

1986), which is located on the long arm of chromosome 6 (www.maizegdb.org). They are

also structurally anomalous with the other zeins, with very little sequence homology and a 3-

dimensional structure that is predicted to form β sheets (PEDERSEN et al. 1986). It has been

suggested that these β-zeins provide stability to the protein bodies to facilitate the deposition

of α- and δ-zeins (BAGGA et al. 1997). 6

Gamma–zeins are the second most abundant group of zein proteins; however, unlike

alpha-zeins, each protein is encoded by only one gene. The coding sequence for the 16 kDa

peptide is located on chromosome 2L and the sequences for the 27 and 50 kDa proteins are

located on chromosome 7L. Structurally and functionally, these proteins are very similar to

the 15kDa β-zein except that each γ–zein contains a highly conserved six-cysteine residue

motif, along with several other conserved polypeptides (WOO et al. 2001).

The 10 and 18 kDa δ-zeins are also part of the same superfamily as the β- and γ-zeins;

their single coding sequences are located on chromosomes 9L and 6L respectively (CHUI and

FALCO 1995; KIRIHARA 1988). They are sulfur amino-acid rich; however, they are also the

least transcribed of all the maize prolamins (WOO et al. 2001).

The expression of the multi-gene zein family seems to be coordinately regulated both

spatially and temporally during seed development. Comparisons among the promoter

regions of the different zein genes have identified a bifactorial, highly conserved sequence

motif ~300 bp upstream of the transcriptional start site with the consensus sequence 5’

TGTAAAG (N9) TGA(G/C)TCAT 3’ . The 5’ motif has been denoted the prolamin (P-) box, and the 3’ motif is called the GCN4 sequence (MULLER et al. 1995) due to its similarity to

yeast GCN4 motifs (HILL et al. 1986). The P-box, especially, is conserved among other

cereal prolamins, including the barley hordeins, the wheat glutenins, and the Coix coixins

(KREIS et al. 1985). Due to its high level of conservation and the binding specificity of trans-

elements to the P-box, it is highly likely that it is important in the coordinated activation of

zeins during endosperm development (VICENTE-CARBAJOSA et al. 1997). Two other zein

promoter motifs have been described, ACGT, a core sequence located ~20 bp downstream of

the bipartite motif (SCHMIDT et al. 1992) that has also been shown to recruit trans-acting 7

factors such as bZIP proteins (FOSTER et al. 1994) and AACA, which has been characterized

in the promoter of the rice analog to zeins, the glutelins (TAKAIWA et al. 1996). It has been

demonstrated that these two elements, along with the GCN4 motif, are sufficient to confer

endosperm-specific expression (WU et al. 2000). Several trans-acting factors controlling

zein expression have also been characterized, including the bZIP proteins Opaque 2 and

Opaque 2-Heterodimerizing Protein (OHP) that bind to the ACGT core sequence(PYSH et al.

1993; SCHMIDT et al. 1992), and the DOF protein Prolamin Binding Factor (PBF) that binds

to the P-box (VICENTE-CARBAJOSA et al. 1997).

Opaque-2 There have been many mutations that have altered zein deposition in the maize

kernel, either through a change in deposition of the α- or γ-zeins (DANNENHOFFER et al.

1995; GEETHA et al. 1991) or through the structural change of protein bodies (COLEMAN et al. 1997; FONTES et al. 1991). One of the first of these mutations to be characterized was

Opaque 2 (MERTZ et al. 1964). The Opaque 2 mutation causes an extreme reduction of the

22kD alpha zein gene transcripts, which results in a lower level of proteins of this class

(KODRZYCKI et al. 1989) and a 50-70% reduction in zeins overall (MERTZ et al. 1964). The

phenotypic result of this reduction is a soft, “opaque” kernel when compared to the vitreous, translucent endosperm of wild type seed. opaque 2 kernels also have an altered amino acid profile, with 69% more lysine than their wild-type counterparts (MERTZ et al. 1964), making

the opaque 2 mutation a potentially valuable resource for more nutritionally balanced maize.

Unfortunately, the mutation has pleiotrophic effects, including an increase in RNAse activity

(DALBY and DAVIES 1967), alterations in amino acid(YUNES et al. 1994b), nitrogen, and

sugar metabolism (GIROUX et al. 1994), accelerated timing of male sexual maturation 8

(GUPTA 1979), differences in the photosynthetic capability of seedlings (MOROT-GAUDRY et al. 1979), and increased susceptibility to plant pathogens (LOESCH et al. 1976). All of these

factors lower yields and thus limit utility of opaque 2 in a maize breeding program. To

ameliorate these negative effects of the opaque 2 phenotype, modifier genes have been

introduced into opaque 2 lines creating quality protein maize (QPM) lines. These lines

maintain the increased levels of lysine and tryptophan present in opaque 2 maize, but have a

vitreous, translucent endosperm and an increased yield (BJARNASON and VASAL 1992;

GEVERS and LAKE 1992).

The opaque 2 locus encodes a DNA binding protein in the basic leucine zipper (bZIP)

family. It has been characterized as a transcriptional activator (HARTINGS et al. 1989;

SCHMIDT et al. 1990) that is expressed solely in the endosperm as early as 10 days after pollination. Opaque 2 activates transcription of several genes by binding to specific promoter sequences with its basic amino acid filled DNA binding region, including the

ACGT core family element TCCACGTAGA in the promoters of 22kD α-zein (SCHMIDT et

al. 1992) and 15kD gamma zein genes, the GACATGTC and CCACATCATC motifs in the

promoter of the maize b32 gene (YUNES et al. 1994a), as well as to several non-maize

promoters such as the vicilin core sequence GCCACCTCAT in brazil-nut (VINCENTZ et al.

1997) and AP-1 site present in the pea lectin promoter in in vitro studies (DE PATER et al.

1994). These varied binding sites suggest that the opaque 2 protein has a wide range of affinity and could potentially be effective as long as the majority of the ACGT core site is intact (VINCENTZ et al. 1997). Opaque 2 also tends form either a homo- or heterodimer via it’s leucine rich zipper region (YUNES et al. 1994a). Using in vitro expression, opaque 2 has

been shown to bind to the prolamin binding factor (PBF), a zinc-finger protein (VICENTE- 9

CARBAJOSA et al. 1997) and the opaque-2 heterodimerizing protein (OHP1), another bZIP

protein (PYSH et al. 1993).

bZIP Proteins Basic leucine zipper (bZIP) proteins were first discovered by noting the amino acid

similarities among the mammalian c-Fos, c-Jun and c-Myc oncogenes, the rat liver protein

C/EBP, and the yeast regulatory protein Gcn4 (LANDSCHULZ et al. 1988; VOGT et al. 1987).

These proteins were all known to recognize and bind DNA in a sequence-specific manner

and contain two distinct subdomains: a basic region that contacts the DNA and a leucine

zipper region that promotes protein-protein dimerization (PABO and SAUER 1992).

The DNA binding region of bZIP proteins is comprised of ~30 amino acids, the

majority of which are the basic amino acids arginine and lysine (KERPPOLA and CURRAN

1991). As has been shown by swapping the basic and zipper regions of different proteins, this region is responsible for the DNA binding specificity of the bZIP proteins (AGRE et al.

1989). NMR spectroscopy has shown that, when in solution, the basic region is found as a

random coil. However, when binding to its specific DNA sequence, it undergoes a

significant three-dimensional conformational change into an alpha-helical structure, which

suggests that the protein is stabilized by binding to its DNA partner (WEISS et al. 1990).

The leucine zipper region is found directly C-terminally of the basic DNA binding region (KERPPOLA and CURRAN 1991). It is characterized by ~35 residues that form a

leucine heptad repeat region (JOHNSON and MCKNIGHT 1989). It has been shown that this

region forms a left-handed alpha-helix, which places each leucine residue on the same face of

the helix. This area serves as a dimerization domain in which the leucines of two bZIP

proteins “zip up” to form a coiled-coil quaternary protein structure using hydrostatic side 10

chain interactions (PABO and SAUER 1992). Also found in this region are a high density of

oppositely charged amino acids placed in such a manner that there is likely intrahelical

pairing in order to help stabilize the alpha-helical structure (JOHNSON and MCKNIGHT 1989).

The bZIP proteins may form either homo- or heterodimers, allowing them to combine their singular DNA binding domains to create novel ones. This allows bZIP proteins to play a myriad of roles in development and differentiation: they may limit activity, as is the case with the cAMP response regulator CREB, which is antagonized by its partner CREM

(FOULKES et al. 1991); they may alter their target specificity or create new target specificities

(CURRAN and FRANZA 1988; HAI and CURRAN 1991), or it may change regulation by

combining an activation and repression domain (PABO and SAUER 1992).

Structurally, the bZIP dimer complex has been shown by co-crystalline structures to form a Y-shaped, “scissors-like” peptide (VINSON et al. 1989). The parallel, dimerized

leucine zipper regions form the stem of the Y, while the basic-DNA binding regions extend

off as arms in a symmetric fashion. These arms are kinked, so that they can follow the major

groove of the DNA and make sequence-specific amino acid side chain to DNA base pair contacts (ELLENBERGER et al. 1992).

Reverse Genetics Historically, geneticists have uncovered new genes and gene functions by a “forward

genetics” process, where one would first find an altered gene product and then try to identify the particular gene that controls the altered phenotype. However, with the advent of molecular genetics, a “reverse genetics” approach has become more prevalent. Genes are now identified purely on the basis of their position or sequence in the genome with no knowledge whatsoever of their related phenotype or gene product. The phenotype is then 11

determined by altering the sequence of the gene and analyzing the resulting changes in the organism.

Reverse genetics approaches can take two forms: those using specific targeting mechanisms, such as RNAi and homologous recombination; and those using random, non- targeted disruptions, such as chemical mutagenesis and transposon insertions (TIERNEY and

LAMOUR 2005). Non-targeted disruption methods require selection of the individuals or lines

carrying a mutation in the desired gene. In this research we have focused on two particular

methods in the “random, non-targeted disruption” group, TILLING chemical mutagenesis

and transposon insertions using the maize Mutator transposon line.

As is true with all genetic approaches, reverse genetics does have some drawbacks.

Not every reverse genetics technique can be applied to every organism since there is a

varying amount of both genetic information and efficient transformation systems available

for different organisms (TIERNEY and LAMOUR 2005). Further, the mutational load created by the reverse genetics scheme must not mask the recovery of mutants from the screening process (TILL et al. 2003). The genome must not become so mutation-riddled that a mutant

phenotype cannot be recognized. Another consideration is the effect that mutagenesis may

have on the fertility of the organism, especially in diploid organisms. If the reproductive

machinery is not working properly, it will become impossible to create the homozygous

mutant organism necessary to study the mutant phenotype (PERRY et al. 2003).

TILLING TILLING (Targeting Induced Local Lesions IN Genomes) is a high-throughput system that creates point mutations in selected genes with the hope of disrupting gene or protein function. TILLING combines chemical mutagenesis with PCR-based screening 12

methods in order to create an allelic series of mutations in genes of interest. These mutations

can then be used in a reverse genetics analysis in order to determine the function of the gene

product of interest (MCCALLUM et al. 2000a; MCCALLUM et al. 2000b; TILL et al. 2003).

EMS (ethylmethanesulfonate) mutagenesis is the most common method of creating

TILLING mutants. EMS is chemical mutagen that creates a G/C to A/T transition mutation

by alkylating guanine bases. These modified guanines will then pair with a thymine residue

instead of the correct cytosine base during DNA replication (TIERNEY and LAMOUR 2005).

In Arabidopsis, five percent of the EMS induced mutations resulted in premature termination

of the targeted gene product, while 50% result in missense mutations that alter the target

protein’s amino acid sequence (GREENE et al. 2003).

In order to create TILLING stocks, seeds are mutagenized with EMS, and the

resulting M1 generation of plants are self fertilized to create a pool of M2 individuals that are

used for both DNA-based screening and seed inventory. The DNA samples are then pooled

and ~1000 bp segments are amplified using end-labeled, gene-specific primers that are

typically designed by the end-user. The amplification products are incubated with Cel I, an

endonuclease that cleaves at the 3’ end of single base pair mismatches (OLEYKOWSKI et al.

1998). Amplification products are then electrophoresed to separate the cleaved, mutant

products from the wild-type ones. When a mutant is identified in the pooled results, the

procedure is repeated with the unpooled samples to identify the specific M2 individual in

which the mutation has occurred and the approximate location of the mutation in the gene

(GILCHRIST and HAUGHN 2005; MCCALLUM et al. 2000b; TILL et al. 2003).

TILLING is a valuable functional genomics tool for several reasons. As is true with other reverse genetics approaches, it allows investigation of a gene without prior knowledge 13

of its function. However, unlike other techniques, TILLING can provide an investigator with

a wide range of allelic diversity in their mutants, including null, partial loss of function, and

neomorphic alleles through missense and mutations. This allows for a more in depth analysis

of the function of the resultant protein and its potential genetic interactions (WEIL and

MONDE 2007). The ability to create non-null mutations also allows the study of otherwise lethal mutations that are often missed using other techniques (WEIL and MONDE 2007).

Further, with the current concerns surrounding genetically modified plants, TILLING may be

a valuable way to improve agronomic species through genetic modification without introducing foreign DNA (TIERNEY and LAMOUR 2005).

Although it was first used in Arabidopsis, TILLING has become available for both

plant and animal species. TILLING stocks are now in production for many plant species

including maize (Zea mays), rice (Oryza sativa L.), barley (Hordeum vulgare L.), soybean

(Glycine max L. Merr.), lotus (Lotus japonicus), tomato (Solanum lycopersicum L.), and wheat (Triticum aestivum L.) (CALDWELL et al. 2004; PERRY et al. 2003; SLADE et al. 2005;

TILL et al. 2004). TILLING can also be accomplished in animal species, and methods have

been shown to be viable in C. elegans, D. melanogaster, zebrafish (Danio rerio), and rats

(Rattus norvegicus) (BENTLEY et al. 2000; SMITS et al. 2004; WEINHOLDS et al. 2003).

There is a severe limitation to TILLING in animal species, however because of the inability

to store germplasm. Either an efficient strategy for germplasm recovery must be established or live animals must be used and screening must often be completed in a single generation to

avoid germplasm loss (HENIKOFF et al. 2004).

Despite a genome ~20 times larger than Arabidopsis, TILLING in maize has yielded

results similar to those in Arabidopsis, although the rate of recovery of mutations has been 14

lower, necessitating a screening population that is twice as large as that used for Arabidopsis

(4000 individuals versus 2300) to recover the same number of mutations per DNA segment

(GILCHRIST and HAUGHN 2005). Following these positive initial tests, the Maize TILLING

project was opened to the public in 2005 at Purdue University

(http://genome.purdue.edu/maizetilling). As Jan. 2007, 319 mutations have been identified in

62 genes, covering ~76 kb of genomic sequence. Further, 17 genes are in the process of

being TILLed and initial primer testing is underway for 25 more genes (WEIL and MONDE

2007).

Transposon-Mediated Mutagenesis Transposon-mediated, or insertional, mutagenesis is a method in which T-DNA or

other transposons are activated in a genome and allowed to proliferate broadly. The

transposon activity is then squelched by using either mutational “killer” lines (SLOTKIN et al.

2003) or by separating the members of a two element system so that there is no longer any

active transposase present (ALTMANN et al. 1995). After the transposon activity is stabilized

in the genome, these mutations are then screened for a gene of interest and the phenotype of

the organism and the resultant gene functions are analyzed (BALLINGER and BENZER 1989;

KAISER and GOODWIN 1990). This method of mutagenesis is valuable for many reasons: it is

a viable alternative to site-specific methods of mutagenesis (for instance, homologous

recombination) that have been difficult to implement for many plants (MAY et al. 2003); the

large size of the transposon insert is more likely to create a knock-out mutant than a point

mutation or other small change would; the known sequence of the transposon facilitates

quick screening procedures via PCR to determine the sites of mutagenesis and these

sequences can also be used to determine the sequence of the region flanking the insertion 15

(PARINOV and SUNDARESAN 2000). However, insertional mutagenesis has several drawbacks: first, it is possible to get somatic mutations rather than germline mutations due to movement of the transposon during development (CHOMET et al. 1991). These somatic

mutations can cause false positives in a screen (MAY et al. 2003). Further, there is a low

frequency of mutations compared to other reverse genetic methods, which necessitates large

populations to find a mutation in a given gene (GILCHRIST and HAUGHN 2005). Finally,

because of the large size of transposons, genic insertions can cause lethality in essential

genes, which limits the genes available for study (TILL et al. 2003).

In maize, members of the Robertson’s Mutator (Mu) element family are most commonly used to generate transposon-insertion stocks (MAY et al. 2003). This element has

several advantages: Mu has many copies present in the maize genome, and these copies

move often, which creates a high mutation rate (1 new allele/1000-10000 plants)

(ROBERTSON 1978). Mu elements also share a highly conserved 0.2 kb terminal inverted

repeat section, so a single primer can be used to amplify many transposons (MAY et al.

2003). Finally, Mu elements tend to insert in genic regions, which makes them more likely

to disrupt gene function, and the transposon does not seem to have a specific insertion

sequence, which allows for more random insertions into the genome (HANLEY et al. 2000).

In order to create a maize insertional mutagenesis library, Mu active lines were

created and crossed to a Mu inhibitor line. The allele for Mu inhibition is dominant to the active allele; therefore the transposon activity was halted in the progeny of the cross. Tissue samples were then taken from the 43,776 F1 progeny that resulted and pooled to create a library for PCR screening. The plants were also self-pollinated to create an F2 seed stock in order to facilitate phenotypic screening. In order to probe this library for a gene of interest, 16

four Mu-specific primers were created. These are used, along with a gene-specific primer, to

PCR amplify the pooled library in order to detect any transposon insertions into the gene of

interest. A positive result occurs only when the two primers can amplify a product, which is

detected by electrophoresis, and also suggests how close the mutation is to the gene of

interest (SPEULMAN et al. 1999). When there is a positive result to the pooled search, the

process is then repeated with the individual samples to determine which of the F2 progeny

from the seed stock will contain the insertional mutation. These plants are then grown to determine any phenotypic change and to deduce gene function (MAY et al. 2003).

Prior Work on ZmbZIP1 (Jia, 2004 – unpublished) The opaque-2 regulatory network has not been well elucidated beyond its function in

controlling zein deposition. Interestingly, the opaque-2 mutation has been shown to have a

highly pleiotrophic effect on gene interactions (HUNTER et al. 2002). Many ESTs have been

shown to be differentially regulated by opaque-2 protein expression have been found,

including the ribosome-related protein b32 (LOHMER et al. 1991), the amino acid synthesis

and modification proteins cytosolic PPDK (MADDALONI et al. 1996), aspartate

aminotrasferase (SUKALOVIC 1986), acetohydroxy acid synthase (DAMERVAL and LE

GUILLOUX 1998), lysine-ketoglutarate reductase (KEMPER et al. 1999), and aspartate kinase

(WANG and LARKINS 2001), the stress-induced protein glyceraldehyde-3P-dehydrogenase

(DAMERVAL and LE GUILLOUX 1998), the ER-related protein elongation factor 1α (WANG et al. 2001) and the aspartic proteinase precursor (DAMERVAL and LE GUILLOUX 1998) whose

function has not been definitively determined. More recently, a microarray profile of near-

isogenic W46A/W46Ao2 lines showed that there was a significant change in transcript

expression between the wildtype and opaque-2 lines, with 60 genes upregulated more than 17

threefold and 66 genes reduced more than threefold in the o2 line compared to the wildtype.

These genes seemed to cluster into several pathways; upregulated genes belonged to stress

response, molecular chaperone and protein turnover pathways, while downregulated genes

tended to be seen in carbon, amino acid and carbohydrate metabolism regulatory pathways

(HUNTER et al. 2002). However, it is often unclear whether there is a direct effect from

interaction with opaque 2 or a more indirect effect at work (MOTTO et al. 2003). It is also

still unclear what role these gene targets and pathways play in the overarching maize architecture (HUNTER et al. 2002).

In order to learn more about the opaque 2 regulatory network, an experiment was

undertaken to compare the transcript profiles of the endosperm in the inbred lines B45 and

B45o2. To do this, endosperm samples were harvested from each line at 11, 14, 18, and 25

days post-pollination and RNA was used to perform a microarray experiment. The 938 genes with differential expression at least one time point in the microarray analysis were further examined to determine their promoter components. From both of these analyses, one

EST (BE130026), which we have termed ZmbZIP1, was found to be particularly interesting: it was expressed at a level at least two-fold higher in B45 than in B45o2, it is a basic-leucine

zipper transcription factor that has high homology to the Opaque-2 family of proteins, and

promoter analysis showed that this gene encoded by this EST had two opaque-2 binding

sites, as well as several other cis-binding elements, in its 5’ promoter region. This profile suggested that it was possible for this gene to be both regulated by and be a binding partner for the O2 protein, and that it could therefore be part of the opaque-2 pathway, and perhaps even play a part in the endosperm-specific expression of zeins. The expression levels of this gene were then verified using semi-quantitative RT-PCR on RNA derived from the inbred 18

lines B45, B46 and M14, as well as the o2 homologues of each line. The results showed that

ZmbZIP1 was expressed at significantly higher levels in the wild-type samples when

compared to their opaque counterparts (JIA 2004).

Methionine

Methionine is an Essential Amino Acid for Humans and Animals Methionine is an essential amino acid and thus vital for human growth and

development; most sulfur containing compounds, including cysteine, homocysteine,

cystathione, taurine, methanethiol, thiamin, biotin, alpha-lipoic acid (ALA), coenzyme A,

glutathione (GSH), chondroitin sulfate, glucosamine sulfate, fibrinogen, heparin,

metallothionein, and the inorganic sulfate found in biological systems are all synthesized

from this parent amino acid (BAKER 1986). Methionine is activated by ATP to become S- adenosylmethionine (SAM), and this precursor is used to produce several neurotransmitting hormones, including creatine, dopamine and serotonin. Because of its neurotransmitting properties, SAM has been implicated as a potential regulator of Alzheimer’s and Parkinson’s diseases, with low SAM levels being positively correlated with the disease (BOTTIGLIERI et al. 1990; MORRISON et al. 1996). This suggests that an increase in the methionine, and thus

the SAM, pool has the potential to work as a therapeutic agent for affected patients

(BOTTIGLIERI 1997; MEININGER et al. 1982; SMYTHIES and HALSLEY 1984; TCHANTCHOU et

al. 2006). Proper levels of methionine (in the form of SAM) have also been shown to

improve immune function in response to the HIV virus (VAN BRUMMELEN and DUTOIT

2006), reduce triglyceride levels, and protect the liver and bladder from bacterial buildup

(FUNFSTUCK et al. 1997; NEUVONEN et al. 1985). The recommended daily value for SAA

(methionine and cysteine) is at least 13 mg/kg/day for an adult; a relatively high level, since 19

the extracellular sulfate pool in humans is the small compared to other animal species

(BAKER 1986). It is thus vital to obtain an adequate amount of methionine in the diet,

especially for vegetarians and vegans who must rely on relatively methionine-poor plant

proteins rather than SAA-rich animal protein sources.

Maize is a necessary component of animal diets, comprising up to 30% of protein and

90% of starch in dairy cow diets (DADO 1999). Maize-soybean rations are used primarily as high protein feed in livestock and poultry production; however young animals, especially swine, have a higher dietary requirement for essential sulfur amino acids (methionine and

cysteine) than this feed can provide. In order to compensate, the livestock industry spends an

estimated $100 million per year to add synthetic methionine to feed in order to maintain

animal growth at optimal levels (ISMANDE 2001). Thus, due to both health and economic

concern, there is much interest in scientific endeavors to improve the methionine content of

maize.

Importance of Sulfur Sulfur has numerous vital functions in the plant, almost all of which involve the

catalytic or electrochemical function of molecules, as opposed to the structural functions of

other macronutrients, such as carbon or nitrogen. One of the main functions of sulfur

containing molecules is to control the redox potential of intracellular spaces. In a self

regulatory capacity, protein thiols help protect against an excess of metal ions, xenobiotics

(REA et al. 1998), or poisons being accumulated in the cellular milieu by acting as a metal

ligand and inactivating enzymes requiring metal co-factors (LEUSTEK et al. 2000). Sulfur, in

the form of glutathione (a tripeptide of Glu-Cys-Gly), creates a reducing environment in the

cell that facilitates the reduction of other compounds (LEUSTEK et al. 2000). Glutathione has 20

also been shown to be an important signaling molecule in chilling, oxidative, hypoxic and

other environmental stresses (JERCA et al. 1996; KOPRIVA et al. 2001; MULLER and GEBEL

1998; WILHELM et al. 1997). It also has been shown to play a role in several plant

developmental pathways. Glutathione has been implicated as an activator of cell division in

the root apical meristem after germination in Arabidopsis (VERNOUX et al. 2000), and as a

modifier of embryo development, carbon metabolism, protein synthesis, hormone synthesis,

nucleotide metabolism and meristem formation in white spruce (P. glauca) (STASOLLA et al.

2004).

Sulfur is also assimilated into many secondary compounds that impact their own

health and the environment around them. Sulfur containing metabolites, such as camalexin

and other phytoalexins, have been shown to have a role in combating plant pathogens (ZHAO

et al. 1998). Isothiocyanates, such as alliinase in Allium species and glucoraphanin and

sulforaphane in Brassica species, have been shown to reduce levels of cardiovascular

diseases (CHAN et al. 2002; LIU and YEH 2002), inflammation (ALI et al. 2000), and several

cancers (FAHEY et al. 1997; KWEON et al. 2003) in humans. Recently, sulfur-containing secondary metabolites have also been implicated as a potential climate-moderating factor.

This is most commonly seen in photosynthetic algae that produce prodigious amounts of the sulfur-containing compound Dimethylsulfoniopropionate, which, when degraded, volatilizes into the atmosphere and promotes cloud formation in the atmosphere, potentially changing the global climate (COOPER and HANSON 1998).

Sulfur Assimilation 2- Sulfur uptake occurs mainly through soil as anionic sulfate (SO4 ); however gaseous

sulfur dioxide (SO2) has been shown to be absorbed and assimilated through the leaves at a 21

significant level in areas of high air pollution. (LEUSTEK and SAITO 1999). Sulfate is actively

transported into the xylem from the roots, and is carried in the transpiration stream to the

leaves. It enters mesophyll cells and is translocated to the chloroplast, while excess sulfate is

stored the vacuole. Sulfate assimilation occurs primarily in chloroplasts as part of photosynthesis, although a small amount may be incorporated in root plastids. (HELDT 2005).

Methionine Synthesis Through the Sulfur Assimilation Pathway Sulfur enters into the metabolic stream as a relatively inactive molecule, due to its

2- high redox potential. SO4 is adenylated to increase reactivity via the addition of an AMP

molecule by the enzyme ATP sulfurylase to create 5’ adenylsulfate (APS). APS is a branch

point intermediate in the sulfur assimilation pathway, as it can be either directly assimilated

2- into metabolites or further reduced at this stage. In the direct assimilation process, SO4

from APS is directly incorporated into metabolites via an ATP kinase and an enzyme specific

sulfotransferase. Some characterized sulfotransferase products include: flavonol,

desulfoglucosinolate, choline and gallic acid glucoside (VARIN et al. 1997). Plants have been

shown to use the sulfotransferase reaction in several ways, such as creating odiferous secondary metabolites or triggering plant movement (VARIN et al. 1997). Most recently, the

sulfotransferase reaction has been recognized as a mechanism to control bioactivity of

hormones, as is the case with tuberonic acid (12-hydroxyjasmonate) (GIDDA et al. 2003) and

the brassinosteroids of B. rapa (ROULEAU et al. 1999).

Alternately, APS is an intermediate in the sulfur assimilation pathway and is reduced

to hydrogen sulfide (H2S) in a two step process. First, APS reductase catalyzes the reduction

2- APS to SO3 using glutathione (GSH), a cysteine containing reducing agent, to contribute 22

two electrons to the molecule. The sulfide is then further reduced via Photosystem I. The

enzyme sulfide reductase uses six reduced ferredoxin molecules to create H2S.

2- The last step in SO4 assimilation is the incorporation of reduced sulfur (from H2S) into the amino acid cysteine. This reaction is catalyzed using the enzyme o- acetylserine(thiol)lyase (OAS-(thiol)lyase), which is alternately known as cysteine synthase when coupled with the enzyme serine acetyltransferase. These two enzymes use serine, acetylCoA, and the reduced sulfur substrates to produce a cysteine molecule. This complex is often seen as a rate limiting step in cysteine (and by extension, methionine) synthesis, as its activity is regulated by its substrates; o-actetylserine (an intermediate step in the reaction) disrupts the stability of the enzyme complex and reduced sulfur stabilizes it (BOGDANOVA and HELL 1997).

Methionine is made from cysteine using a three step process. The first step involves a synthesis reaction in which cysteine and o-phosphohomoserine are the substrates and a

cystathione molecule is created. Cystathione is then degraded by cystathione β-lyase into pyruvate and homocysteine, to which a methyl group is added to create methionine.

Regulation of the Sulfate Assimilation Pathway Sulfate assimilation is a highly regulated process. Since it is coupled with

photosynthesis, assimilation is dependent on light and carbon compounds; the reduction of

2- 2- SO4 to SO3 is decreased in the absence of CO2, and increased by the addition of sugars and

o-acetylsulfate (NEUENSCHWANDER et al. 1991). Sulfate assimilation is also positively

correlated with the nitrogen assimilation pathway. Nitrogen deficient conditions lead to a

decrease in ATP sulfurylase and APS reductase at the mRNA level and thus the amount of

assimilated sulfur decreases (KOPRIVOVA et al. 2000). APS reductase steady state mRNA 23

level is also altered in the presence of many environmental and cellular stresses (LEUSTEK and SAITO 1999), suggesting that it is a key regulation point in sulfate assimilation (BICK and

LEUSTEK 1998).

At the protein level, sulfur assimilation is regulated at two control points. The first

potential level of regulation is at ATP sulfurylase enzyme; however, it is not clear if this

regulation occurs in vivo. Two studies, both of which used over expression of Arabidopsis

ATP sulfurylase in a heterologous system, have reported opposite results on the issue:

tobacco cell cultures showed no increase in sulfur assimilation (HATZFELD et al. 1998) while

2- Indian mustard plants showed an increase in glutathione production and resistance to SeO4 ,

a toxic sulfur analog that must be reduced via the sulfur assimilation pathway to become non-

toxic (PILON-SMITS et al. 1999). These opposing results may be due to the transgenic

systems used or differences between the species (LEUSTEK and SAITO 1999). The second,

well researched, level of protein regulation occurs at cysteine synthase (OAS-(thiol)lyase),

which regulates the incorporation of S2- and o-acetylserine (OAS). When the level of S2- is

low, the level of OAS increases, which causes the cysteine synthase complex to dissociate

and no cysteine to be produced. In contrast, an overabundance of S2- and low OAS leads to the complex being stabilized in order to produce cysteine (DROUX et al. 1998).

Cysteine Synthase Cysteine synthase (CSase), or o-acetyl-L-serine (thiol)lyase, is responsible for the

formation of L-cysteine from free sulfur (S2-) and o-acetyl-L-serine, as well as the

biosynthesis of heterocyclic β-substituted alanine secondary metabolites (SAITO et al. 1992).

The cysteine synthase gene was first characterized in spinach (Spinacia oleracea L.), with two isoforms present in the leaves, each with a different subcellular localization (SAITO et al. 24

1992). CysA (NCBI: Q00834) is the cytosolic form and Cys B (NCBI: D14722) is the

plastid form. CysA was found to be composed of a single ORF encoding a polypeptide of 35

kDa (SAITO et al. 1992). Multiple compartmentalized isoenzymes have also been purified in

Arabidopsis (A. thaliana), which has cytosolic (NCBI NP_193224), chloroplastic (NCBI:

S48695), and mitochondrial (X81973) enzymes (HELL et al. 1994; HESSE and ALTMANN

1995); Solanum tuberosum, which has cytosolic (NCBI: BAB20861) and chloroplastic isoenzymes (CHRONIS and KRISHNAN 2003); and the flower Datura innoxia, which has

cytosolic, chloroplastic, and mitochondrial forms (KUSKE et al. 1996). This redundancy in

enzymes is consistent with the idea that each cell compartment is independently able to form

cysteine because of its essential nature in the cell (HESSE and ALTMANN 1995). Single

isoforms have also been found in many other species, with cytosolic isoforms being present

in Glycine max (NCBI: AF452451), Brassica juncea (NCBI: O23733) Citrulus lanatus

(NCBI: Q43317), Oryza sativa (NCBI: Q9XEA8), Allium tuberosum (NCBI: BAA93051), and Triticum aestivum (NCBI: P38076). Chloroplastic isoforms are present in Capsicum annuum (NCBI: P31300), and Nicotiana tabacum (NCBI: AJ299249) (CHRONIS and

KRISHNAN 2003). These species may, in fact, have multiple isoforms that are either too similar, or too divergent, to have been recognized. In maize, two isoforms of cysteine synthase have been found, Cys1 and Cys2 (www.maizeGDB.org) (Brander, 1995). However, only Cys2 has been mapped, to chromosome 1, ctg 36 (www.MaizeSequence.org), and there

is some debate as to whether there are actually isoforms present or if there is only a single

locus. 25

10kDa Delta Zein The10kDa delta zein is extremely rich in methionine, making it the primary source

for sulfur in the growing seedling. The mature protein contains 29 Met residues, making up

20% of the amino acids present in the protein (CHUI and FALCO 1995; KIRIHARA 1988).

These amino acids are typically present as Met-Met doublets, separated by a proline

containing spacer of 2-3 aa. The mature 10kDa delta-zein protein is translocated through the

ER/Golgi system (LARKINS and HURKMAN 1978) and accumulates in protein bodies that are

localized primarily in the subaleurone and first few layers of the endosperm, making it

available for sulfur and energy utilization (KIM and KRISHNAN 2003). Different alleles of the

10kDa delta zein have been shown to alter the levels of methionine present in the maize seed, increasing the nutritional quality of the seed. Examples of this phenomenon abound in maize. For instance, the dzs10 gene, which encodes the10kD delta-zein protein, has shown

variation that increases the level of whole kernel methionine by 34-49% through EMS

mutagenesis (OLSEN et al. 1999). A chromosome 9 region containing the dzs10 allele from

B37LTI increases whole kernel methionine level in the inbred line A679 (MNL 73(72):84-

85), and an 6 aa (CRMLLP) insertion in the gene creates a novel 11kDa methionine rich delta-zein allele that is composed of nearly 25% sulfur amino acids, raising the level of sulfur available to the seed (KIM and KRISHNAN 2003).

Beyond differences in primary sequence, transcriptional and post-transcriptional regulation also seem to alter the genotype specific levels of 10 kDa delta-zein. Several inbred maize lines have reduced or undetectable levels of 10kDa delta-zein mRNA, including

A654, C103, Co64, GE37, GT119, Mo17, Mo20, and SD purple (KIM and KRISHNAN 2003;

SCHICKLER et al. 1993). This phenomenon was further investigated using the Mo17 inbred, 26

which was found to accumulate only low levels of the 10kDa delta-zein mRNA transcript for only a small portion of the typical developmental time frame of 10-25 dap (SCHICKLER et al.

1993). Post-transcriptional differences, such as mRNA stability, have also been shown to differ between genotypes. In a study by Cruz-Alvarez et al, inbred line BSSS-53, an overexpressor of the 10 kDa protein in mature kernels, was shown to have higher mRNA levels in developing endosperm, as compared to inbred lines W23 and W64A, despite having no differential expression at the protein level and similar transcription rates (CRUZ ALVAREZ

et al. 1991).

References AGRE, P., P. F. JOHNSON and S. L. MCKNIGHT, 1989 Cognate DNA binding specificity retained after leucine zipper exchange between GCN4 and C/EBP. Science 246: 922- 926. ALI, M., M. THOMSON and M. AFZAL, 2000 Garlic and onions: their effect on eicosanoid metabolism and its clinical relevance. Prostaglandins, Leukotrienes and Essential Fatty Acids 62: 55-73. ALTMANN, T., G. FELIX, A. JESSOP, A. KAUSCHMANN, U. UWER et al., 1995 Ac/Ds Transposon Mutagenesis in Arabidopsis thaliana: Mutant Spectrum and Frequency of s Insertion Mutants. Molecular and General Genetics 247: 646-652. ARGOS, P., K. PEDERSEN, M. D. MARKS and B. A. LARKINS, 1982 A structural model for maize zein proteins. J Biol Chem 257: 9984-9990. BAGGA, S., H. ADAMS and F. RODRIGUEZ, KEMP, JD., SENGUPTA-GOPALAN, C., 1997 Coexpression of the maize delta-zein and beta-zein genes results in stable accumulation of delta-zein in endoplasmic reticulum-derived protein bodies formed by beta-zein. Plant Cell 9: 1683-1696. BAKER, D. H., 1986 Utilization of isomers and analogs of amino acids and other sulfur- containing compounds. Prog Food Nutr Sci 10: 133-178. BALLINGER, D. G., and S. BENZER, 1989 Targeted Gene Mutations in Drosophila. PNAS 86: 9402-9406. BENTLEY, A., B. MACLENNAN, J. CALVO and C. R. DEAROLF, 2000 Targeted recovery of mutations in Drosophila. Genetics 156: 1169-1173. BICK, J. A., and T. LEUSTEK, 1998 Plant sulfur metabolism: the reduction of sulfate to sulfite. Current Opinion in Plant Biology 1: 240-244. BJARNASON, M., and S. K. VASAL, 1992 Breeding of Quality Protein Maize (QPM). Plant Breeding Reviews 9: 181-216. BOGDANOVA, N., and R. HELL, 1997 Cysteine synthesis in plants: protein-protein interactions of serine acetyltransferase from Arabidopsis thaliana. Plant Journal 11: 251-262. 27

BOTTIGLIERI, T., 1997 Ademethionine (S-adenylmethionine) neuropharmacology: implications for drug therapies in psychiatric and neurological disorders. Expert Opinion on Investigational Drugs 6: 417-426. BOTTIGLIERI, T., P. GODFREY, T. FLYNN, M. W. CARNEY, B. K. TOONE et al., 1990 Cerebrospinal fluid S-adenosylmethionine in depression and dementia: effects of treatment with parenteral and oral S-adenosylmethionine. Journal of Neurology, Neurosurgery and Psychiatry 53: 1096-1098. CALDWELL, D. J., P. MCCALLUM, G. J. SHAW, G. J. MUEHLBAUER, D. F. MARSHALL et al., 2004 A Structured Mutant Population for Forward and Reverse Genetics in Barley (Hordeum vulgare L.). Plant Journal 40: 143-150. CHAN, K.-C., C.-C. HSU and M.-C. YIN, 2002 Protective effect of three diallyl sulphides against glucose-induced erythrocyte and platelet oxidation, and ADP-induced platelet aggregation. Thrombosis Research 108: 317-322. CHOMET, P., D. LISCH, K. J. HARDEMAN, V. L. CHANDLER and M. FREELING, 1991 Identification of a Regulatory Transposon That Controls the Mutator Transposable Element System in Maize. Genetics 129: 261-270. CHRONIS, D., and H. B. KRISHNAN, 2003 Sulfur assimilation in soybean: molecular cloning and characterization of o-acetylserine (thiol) lyase (cysteine synthase). Crop Science 43: 1819-1827. CHUI, C. F., and S. C. FALCO, 1995 A new methionine-rich seed storage protein from maize. Plant physiol 107: 291. COLEMAN, C. E., A. M. CLORE, J. P. RANCH, R. HIGGINS, M. A. LOPES et al., 1997 Expression of a mutant alpha-zein creates the floury 2 phenotype in transgenic maize. Proc Natl Acad Sci U S A 94: 7094-7097. COOPER, A. J. L., and A. D. HANSON, 1998 Advances in enzymology of the biogeochemical sulfur cycle. Biochemistry and Molecular Biology 11: 729-747. CRUZ ALVAREZ, M., J. A. KIRIHARA and J. MESSING, 1991 Post-transcriptional regulation of methionine content in maize kernels. M G G Mol Gen Genet 225: 331-339. CURRAN, T., and B. R. FRANZA, 1988 Fos and jun: The AP-1 connection. Cell 55: 395-397. DADO, R. G., 1999 Nutritional Benefits of Specialty Corn Grain Hybrids in Dairy Diets. Journal of Animal Science 77: 197-207. DALBY, A., and I. DAVIES, 1967 Ribonuclease activity in the developing seeds of normal and opaque-2 maize. Science 155: 1573-1575. DAMERVAL, C., and M. LE GUILLOUX, 1998 Characterization of novel proteins affected by the o2 mutation and expressed during maize endosperm development. Mol. Gen. Genet. 257: 354-361. DANNENHOFFER, J., D. BOSTWICK, E. OR and B. LARKINS, 1995 opaque-15, a Maize Mutation with Properties of a Defective opaque-2 Modifier. PNAS 92: 1931-1935. DE PATER, S., F. KATAGIRI, J. KIJNE and N.-H. CHUA, 1994 bZIP proteins bind to a palindromic sequence without an ACGT core located in a seed-specific element of the pea lectin promoter. Plant J 6: 133-140. DELGADO, C., M. ROSEGRANT, H. STEINFELD, S. EHUI and C. COURBOIS, 1999 Livestock to 2020: The Next Food Revolution, pp. in 2020 Vision, edited by IFPRI. DROUX, M., M. L. RUFFET, R. DOUCE and D. JOB, 1998 Interactions between serine acetyltransferase and O-acetylserine(thiol)lyase in higher plants: structural and kinetic 28

properties of the free and bound enzymes. European Journal of Biochemistry 255: 235-245. ELLENBERGER, T., C. BRANDL, K. STRUHL and S. HARRISON, 1992 The GCN4 basic region leucine zipper binds DNA as a dimer of uninterrupted α Helices: Crystal structure of the protein-DNA complex. Cell 71: 1223-1237. FAHEY, J. W., Y. ZHANG and P. TALALAY, 1997 Broccoli sprouts: An exceptionally rich source of inducers of enzymes that protect against chemical carcinogens. Proc. Natl. Acad. Sci. U. S. A. 94: 10367-10372. FAO, 1992 Maize in Human Nutrition. FAO, Rome. FAO, 1999 FAO Yearbook Production. FAO/WHO/UNU, 1985 Energy and protein requirements. Report of a joint FAO/WHO/UNU Expert Consultation. WHO, Geneva. FAOSTAT, 2001 FAOSTAT data - Maize Food Grains Consumption, pp. in http://faostat.fao.org/faostat/form?collection=FBS&Domain=FBS&servlet=1&hasb ulk=0&version=ext&language=EN FONTES, E. B. P., B. B. SHANK and R. L. WROBEL, MOSSE, S.P., O'BRIAN, G.R., WURTZEL, E.T., BOSTON, R.S., 1991 Characterization of an immunoglobulin binding protein homolog in the maize floury-2 endosperm mutant. Plant cell 3: 483-496. FOSTER, R., T. IZAWA and N.-H. CHUA, 1994 Plant bZiP proteins gather at ACGT elements. FASEB J 8: 192-200. FOULKES, N. S., E. BORRELLI and P. SASSONE-CORSI, 1991 CREM gene: Use of alternative DNA-binding domains generates multiple antagonists of cAMP-induced transcription. Cell 64: 739-749. FUNFSTUCK, R., E. STRAUBE, O. SCHILDBACH and U. TIETZ, 1997 Prevention of reinfection by L-methionine in patients with recurrent urinary tract infection. Med Klin 92: 574- 581. GEETHA, K. B., C. R. LENDING, M. A. LOPES, J. C. WALLACE and B. A. LARKINS, 1991 opaque-2 modifiers increase gamma-zein synthesis and alter its spatial distribution in maize endosperm. Plant Cell 3: 1207-1219. GEVERS, H. O., and J. K. LAKE, 1992 Development of modified opaque2 maize in South Africa, pp. 49-78 in Quality Protein Maize, edited by E. T. MERTZ. Amer Assoc Cereal Chem, St. Paul, MN. GIDDA, S. K., O. MIERSCH, A. LEVITIN, J. SCHMIDT, C. WASTERNACK et al., 2003 Biochemical and molecular characterization of a hydroxyjasmonate sulfotransferase from Arabidopsis thaliana. The Journal of Biological Chemistry 278: 17895-17900. GILCHRIST, E. J., and G. W. HAUGHN, 2005 TILLING without a plough: a new method with applications for reverse genetics. Current Opinion in Plant Biology 8: 1-5. GIROUX, M. J., C. BOYER, G. FEIX and L. C. HANNAH, 1994 Coordinated transcriptional regulation of storage product genes in the maize endosperm. Plant Physiol 106: 713- 722. GREENE, E. A., C. A. CODOMO, N. E. TAYLOR, J. G. HENIKOFF, B. J. TILL et al., 2003 Spectrum of chemically induced mutations from a large-scale reverse-genetics screen in Arabidopsis. Genetics 164: 731-740. GUPTA, D., 1979 Shortening of protandry by opaque-2 mutant gene. Indian Journal of Genetics and Plant Breeding 39: 230-233. 29

HAI, T., and T. CURRAN, 1991 Cross-family dimerization of transcription factors Fos/Jun and ATF/CREB alters DNA binding specificity. PNAS 88: 3720-3724. HANLEY, S., D. EDWARDS, D. STEVENSON, S. HAINES, M. HEGARTY et al., 2000 Identification of Transposon-Tagged Genes by the Sequencing of Mutator-Tagged DNA Fragments from Zea Mays. The Plant Journal 23: 557-566. HARTINGS, H., M. MADDALONI, N. LAZZARONI, N. DIFONZO, M. MOTTO et al., 1989 The O2 gene which regulates zein deposition in maize endosperm encodes a protein with structural homologies to transcriptional activators. EMBO 8: 2795-2801. HATZFELD, Y., N. CATHALA, C. GRINGNON and J. C. DAVIDIAN, 1998 Effect of ATP sulfurylase overexpression in bright yellow 2 tobacco cells. Regulation of ATP 2- sulfurylase and SO4 transport activities. Plant Physiology 116: 1307-1313. HEIDECKER, G., and J. MESSING, 1986 Structural analysis of plant genes. Annu Rev Plant Physiol 37: 439-466. HELDT, H. W., 2005 Plant Biochemistry. Elsevier Academic Press, Burlington, MA. HELL, R., C. BORK, N. BOGDANA, I. FROLOV and R. HAUSCHILD, 1994 Isolation and characterization of two cDNAs encoding for compartment specific isoforms of O- acetylserine (thiol) lyase from Arabidopsis thaliana. FEBS Letters 351: 257-262. HENIKOFF, S., B. J. TILL and L. COMAI, 2004 TILLING. Traditional mutagenesis meets functional genomics. Plant Physiology 13: 630-636. HESSE, H., and T. ALTMANN, 1995 Molecular cloning of a cysteine synthase cDNA from Arabidopsis thaliana. Plant Physiology 108: 851-852. HILL, D. E., I. A. HOPE, J. P. MACKE and K. STRUHL, 1986 Saturation mutagenesis of the yeast his3 regulatory site: requirements for transcriptional induction and for binding by GCN4 activator protein. Science 234: 451-457. HUNTER, B. G., M. K. BEATTY, G. W. SINGLETARY, B. R. HAMAKER, B. P. DILKES et al., 2002 Maize Opaque Endosperm Mutations Create Extensive Changes in Patterns of Gene Expression. Plant Cell 14: 2591-2612. ISMANDE, J., 2001 Selection of soybean mutants with increased concentrations of seed methionine and cysteine. Crop Science 41: 510-515. JERCA, L., A. BUSUIOC, G. NASTASIA and I. HAULICA, 1996 Glutathione, a biochemical indicator for different types of cellular stress: the change in the GSH level in the brain of rats with hypoxia. Rev Med Chir Soc Med Nat Iasi 100: 143-147. JIA, H., 2004 Expression analysis of maize opaque2 mutants and Glu-1Dx5 transgenic corn, pp. 147 in Interdepartmental Genetics. Iowa State University, Ames, IA. JOHNSON, P. F., and S. L. MCKNIGHT, 1989 Eukaryotic Transcriptional Regulatory Proteins. Annu. Rev. Biochem. 58: 799-839. KAISER, K., and S. F. GOODWIN, 1990 Site-selected Transposon Mutagenesis of Drosophila. PNAS 87: 1686-1690. KEMPER, E. L., G. C. NETO, F. PAPES, K. C. M. MORAES, A. LEITE et al., 1999 The Role of Opaque2 in the Control of Lysine-Degrading Activities in Developing Maize Endosperm. Plant Cell 11: 1981-1994. KERPPOLA, T. K., and T. CURRAN, 1991 Transcription factor interactions: basics on zippers. Current Opinion in Structural Biology 1: 71-79. KIM, W. S., and H. B. KRISHNAN, 2003 Allelic variation and differential expression of methionine-rich delta-zeins in maize inbred lines B73 and W23a1. Planta 217: 66-74. 30

KIRIHARA, J. A., PETRI, J.B., AND MESSING, J.W., 1988 Isolation and sequence of a gene encoding a methionine-rich 10-kDa zein protein from maize. Gene 71: 359–370. KOBAYASHI, H., 1998 Demand for Livestock Products and Feed in Southeast Asian Countries: Preliminary Projections, pp. in International Contribution of Japan and ILRI to World Animal Production Research, edited by J. I. R. C. A. SCIENCES, Tsukuba, Japan. KODRZYCKI, R., R. S. BOSTON and B. A. LARKINS, 1989 The opaque-2 mutation of maize differentially reduces zein gene transcription. Plant Cell 1: 105-114. KOPRIVA, S., S. JONES, A. KOPRIVOVA, M. SUTER, P. V. BALLMOOS et al., 2001 Influence of chilling stress on the intercellular distribution of assimilatory sulfate reduction and thiols in Zea mays. Plant Biology 3: 24-31. KOPRIVOVA, A., M. SUTER, R. OPDEN CAMP, C. BRUNOLD and S. KOPRIVA, 2000 Regulation of sulfate assimilation by nitrogen in Arabidopsis. Plant Physiology 122: 737-746. KREIS, M., B. G. FORDE, S. RAHMAN, B. J. MIFLIN and P. R. SHEWY, 1985 Molecular evolution of the seed storage proteins of barley, rye and wheat. Journal of Molecular Biology 183. KRISHNAN, H. B., 2005 Engineering soybean for enhanced sulfur amino acid content. Crop Science 45: 454-461. KUSKE, C. R., K. K. HILL, E. GUZMAN and P. J. JACKSON, 1996 Subcellular location of o- acetylserine sulfhydrylase isoenzymes in cell cultures and plant tissues of Datura innoxia Mill. Plant Physiology 112: 659-667. KWEON, S., K.-A. PARK and H. CHOI, 2003 Chemopreventive effect of garlic powder diet in diethylnitrosamine-induced rat hepatocarcinogenesis. Life Sciences 73: 2515-2526. LANDSCHULZ, W. H., P. F. JOHNSON, E. Y. ADASHI, B. J. GRAVES and S. L. MCKNIGHT, 1988 Isolation of a recombinant copy of the gene encoding C/EBP. Genes and Developement 2: 786-800. LARKINS, B. A., 1986 Modification of proteins encoded by seed storage protein genes. Tailoring genes for crop improvement : an agricultural perspective: p. 163-167. LARKINS, B. A., and W. J. HURKMAN, 1978 Synthesis and deposition of zein in protein bodies of maize endosperm. Plant Physiology 62: 256-263. LARKINS, B. A., C. R. LENDING and J. C. WALLACE, 1993 Modification of maize seed protein quality. Am J Clin Nutr 58: 264S-269S. LARKINS, B. A., K. PEDERSEN, M. D. MARKS and D. R. WILSON, 1984 The zein proteins of maize endosperm. Trends Biochem Sci 9: 306-308. LEUSTEK, T., M. N. MARTIN, J.-A. BICK and J. P. DAVIES, 2000 Pathways and regulation of sulfur metabolism revealed through molecular and genetic studies. Annual Review of Plant Physiology and Plant Molecular Biology 51: 141-165. LEUSTEK, T., and K. SAITO, 1999 Sulfate transport and assimilation in plants. Plant Physiology 120: 637-643. LIU, L., and Y.-Y. YEH, 2002 S-alk(en)yl cysteines of garlic inhibit cholesterol synthesis by deactivating HMG-CoA reductase in cultured rat hepatocytes. The Journal of Nutrition 132: 1129-1134. LOESCH, J., D. C. FOLEY and D. F. COX, 1976 Comparative resistance of Opaque-2 an Normal inbred lines of maize to ear-rotting pathogens. Crop Science 16: 841-842. 31

LOHMER, S., M. MADDALONI, M. MOTTO, N. DI FONZO, H. HARTINGS et al., 1991 The maize regulatory locus Opaque-2 encodes a DNA-binding protein which activates the transcription of the b-32 gene. EMBO J. 10: 617-624. MADDALONI, M., G. DONINI, C. BALCONI, E. RIZZI, P. GALLUSCI et al., 1996 The transcriptional activator Opaque-2 controls the expression of a cytosolic form of pyruvate orthophosphate dikinase-1 in maize endosperms. Mol Gen Genet 250: 647- 654. MAY, B. P., H. LIU, E. VOLLBRECHT, L. SENIOR, P. D. RABINOWICZ et al., 2003 Maize- targeted Mutagenesis: A Knockout Resource for Maize. PNAS 100: 11541-11546. MCCALLUM, C. M., L. COMAI, E. A. GREENE and S. HENIKOFF, 2000a Targeted screening for induced mutations. Nat. Biotechnol. 18: 455-457. MCCALLUM, C. M., L. COMAI, E. A. GREENE and S. HENIKOFF, 2000b Targeting Induced Local Lesions IN Genomes (TILLING) for Plant Functional Genomics. Plant Physiol. 123: 439-442. MEININGER, V., A. FLAMIER, T. PHAN, O. FERRIS, A. UZAN et al., 1982 L-methionine treatment of Parkinson's disease: preliminary results. Rev Neurol (Paris) 138: 297- 303. MERTZ, E., L. BATES and O. E. NELSON, JR., 1964 Mutant Gene that Changes Protein Composition and Increases Lysine Content of Maize Endosperm. Science 16: 279- 280. MOROT-GAUDRY, J.-F., J. FARINEAU and E. JOLIVET, 1979 Effect of leaf position and plant age on photosynthetic carbon metabolism in leaves of 8 and 16 day-old maize seedlings (W64A) with and without the gene opaque 2. Photosynthetica 13: 365-375. MORRISON, L. D., D. D. SMITH and S. J. KISH, 1996 Brain S-adenosylmethionine levels are severely decreased in Alzheimer's disease. Journal of Neurochemistry 67: 1328-1331. MOTTO, M., H. HARTINGS and V. ROSSI, 2003 Gene discovery to improve the maize grain cell factory. AGRO Food industry hi-tech. MULLER, M., J. R. MUTH, P. GALLUSCI, S. KNUDSEN, M. MADDALONI et al., 1995 Regulation of storage protein synthesis in cereal seeds: developmental and nutritional aspects. J. Plant Physiol 145: 606-613. MULLER, T., and S. GEBEL, 1998 The cellular stress response induced by aqueous extracts of cigarette smoke is critically dependent on the intracellular glutathione concentration. Carcinogenesis 19: 797-801. NELSON, O. E., 1969 Genetic modification of protein quality in plants. Adv Agron 21: 171- 194. NEUENSCHWANDER, U., M. SUTER and C. BRUNOLD, 1991 Regulation of sulfate assimilation by light and O-acetyl-L-serine in Lemna minor L. Plant Physiology 97: 253-258. NEUVONEN, P. J., O. TOKOLA, M. L. TOIVONEN and O. SIMELL, 1985 Methionine in paracetamol tablets, a tool to reduce paracetamol toxicity. International Journal of Clinical Pharmacology, Therapy, and Toxicology 23: 497-500. OLEYKOWSKI, C. A., C. R. BRONSON MULLINS, A. K. GODWIN and A. T. YEUNG, 1998 Mutation detection using a novel plant endonuclease. Nucleic Acids Res 26: 4597- 4602. 32

OLSEN, M. S., G. I. GRAHAM and R. L. PHILLIPS, 1999 A chromosome 9 region containing the dzs10 allele from B37LTI increases whole kernel methionine level in the inbred line A679. Maize Genetics Cooperation Newsletter. PABO, C. O., and R. T. SAUER, 1992 Transcription factors: Structural familes and principles of DNA recognition. Annu. Rev. Biochem. 61: 1053-1095. PARINOV, S., and V. SUNDARESAN, 2000 Functional Genomics in Arabidopsis: Large-Scale Insertional Mutagenesis Coomplements the Gemone Sequencing Project. Curr. Opin. Biotechnol. 11: 157-161. PEDERSEN, K., P. ARGOS, S. NARAVANA and B. LARKINS, 1986 Sequence analysis and characterization of a maize gene encoding a high- sulfur zein protein of Mr 15,000. J. Biol. Chem. 261: 6279-6284. PERRY, J. A., T. L. WANG, T. J. WELHAM, S. GARDNER, J. M. PIKE et al., 2003 A TILLING Reverse Genetics Tool and a Web-Accessible Collection of Mutants of the Legume Lotus japonicus. Plant Physiology 131. PILON-SMITS, E. A. H., S. B. HWANG, C. M. LYTLE, Y. L. ZHU, J. C. TAI et al., 1999 Overexpression of ATP sulfurylase in Indian mustard leads to increased selenate uptake, reduction and tolerance. Plant Physiology 119: 123-132. PRAT, S., PEREZ-GRAU, L., AND PUIGDOMENECH, P., 1987 Multiple variability in the sequence of a family of maize endosperm proteins. Gene 52: 41–49. PYSH, L. D., M. J. AUKERMAN and R. J. SCHMIDT, 1993 OHP1: a maize basic domain/leucine zipper protein that interacts with opaque2. Plant Cell 5: 227-236. REA, P. A., Z. S. LI, Y. P. LU, Y. M. DROZDOWICZ and E. MARTINOIA, 1998 From vacuolar GS-X pumps multispecific ABC transporters. Annual Review of Plant Physiology and Plant Molecular Biology 49: 727-760. ROBERTSON, D. S., 1978 Characterization of a Mutator System in Maize. Mutation Research 51: 21-28. ROULEAU, M., F. MARSOLAIS, M. RICHARD, L. NICOLLE, B. VOIGT et al., 1999 Inactivation of brassionsteroid biological activity by a salicylate-inducible steroid sulfotransferase from Brassica napus. Journal of Biological Chemistry 274: 20925-20930. SAITO, K., N. MIURA, M. YAMAZAKI, H. HIRANO and I. MURAKOSHI, 1992 Molecular cloning and bacterial expression of cDNA encoding a plant cysteine synthase. Proc. Natl. Acad. Sci. U. S. A. 89: 8078-8082. SCHICKLER, H., M. S. BENNER and J. MESSING, 1993 Repression of the high-methionine zein gene in the maize inbred line Mo17. Plant J 3: 221-229. SCHMIDT, R., F. BURR, M. AUKERMAN and B. BURR, 1990 Maize Regulatory Gene Opaque-2 Encodes a Protein with a "Leucine-Zipper" Motif that Binds to Zein DNA. PNAS 87: 46-50. SCHMIDT, R. J., M. KETUDAT, M. J. AUKERMAN and G. HOSCHEK, 1992 Opaque-2 Is a Transcriptional Activator That Recognizes a Specific Target Site in 22-kD Zein Genes. Plant Cell 4: 689-700. SLADE, A. J., S. I. FUERSTENBERG, D. LOEFFLER, M. N. STEINE and D. FACCIOTTI, 2005 A Reverse Genetic, Nontransgenic Approach to Wheat Crop Improvement by TILLING. Nat. Biotechnol. 23: 75-81. 33

SLOTKIN, R. K., M. FREELING and D. LISCH, 2003 Mu killer Causes the Heritable Inactivation of the Mutator Family of Transposable Elements in Zea mays. Genetics 165: 781- 797. SMITS, B. M. G., J. MUDDE, R. H. PLASTERK and E. CUPPEN, 2004 Target-selected mutagenesis of the rat. Genomics 83: 332-334. SMYTHIES, J. R., and J. H. HALSLEY, 1984 Treatment of Parkinson's disease with L- methionine. Southern Medical Journal 77: 1577. SONG, R., V. LLACA, E. LINTON and J. MESSING, 2001 Sequence, Regulation, and Evolution of the Maize 22-kD alpha Zein Gene Family. Genome Res. 11: 1817-1825. SPEULMAN, E., P. L. J. METZ, G. VANARKEL, B. T. L. HEKKERT, W. J. STIEKEMA et al., 1999 A Two-Component Enhancer-Inhibitor Transposon Mutagenesis System for Functional Analysis of the Arabidopsis Genome. The Plant Cell 11: 1853-1866. STASOLLA, C., M. F. BELMONTE, L. VAN ZYL, D. L. CRAIG, W. LIU et al., 2004 The effect of reduced glutathione on morphology and gene expression of white spruce (Picea glauca) somatic embryos. Journal of Experimental Botany 55: 695-709. SUKALOVIC, V. H.-K., 1986 Activity and Distribution of Nitrogen-Metabolism Enzymes in the Desveloping Maize Kernel. Physiologia Plantarum 67: 247-252. TAKAIWA, F., U. YAMANOUCHI, T. YOSHIHARA, H. WASHIDA, F. TANABE et al., 1996 Characterization of common cis-regulatory elements responsible for the endosperm- specific expression of members of the rice glutelin multigene family. Plant Mol Biol. 30: 1207-1221. TCHANTCHOU, F., M. GRAVES, D. ORTIZ, A. CHAN, E. ROGERS et al., 2006 S-adenosyl methionine: A connection between nutritional and genetic risk factors for neurodegeneration in Alzheimer's disease. The Journal of Nutrition, Health and Aging 10: 541-544. TIERNEY, M. B., and K. H. LAMOUR, 2005 An introduction to reverse genetic tools for investigating gene function. The Plant Health Instructor DOI: 10.1094/PHI-A-2005- 1025-01. TILL, B. J., S. H. REYNOLDS, E. A. GREENE, C. A. CODOMO, L. C. ENNS et al., 2003 Large- Scale Discovery of Induced Point Mutations With High-Throughput TILLING. Genome Research 13: 524-530. TILL, B. J., S. H. REYNOLDS, C. F. WEIL, N. M. SPRINGER, C. BURTNER et al., 2004 Discovery of Induced Point Mutations in Maize Genes by TILLING. BMC Plant Biology 4: 12. VAN BRUMMELEN, R., and D. DUTOIT, 2006 L-methionine as immune supportive element: a clinical evaluation. Amino Acids 33: 157-163. VARIN, L., F. MARSOLAIS, M. RICHARD and M. ROULEAU, 1997 Biochemistry and molecular biology of plant sulfotransferases. The FASEB Journal 11: 517-525. VERNOUX, T., R. C. WILSON, K. A. SEELEY, J. P. REICHHELD, S. MUROY et al., 2000 The ROOT MERISTEMLESS/CADMIUM SENSITIVE2 gene defines a glutathione- dependent pathway involved in initiation and maintenance of cell division during postembryonic root development. Plant Cell 12: 97-100. VICENTE-CARBAJOSA, J., S. P. MOOSE, R. L. PARSONS and R. J. SCHMIDT, 1997 A maize zinc- finger protein binds the prolamin box in zein gene promoters and interacts with the basic leucine zipper transcriptional activator Opaque2. Proc Natl Acad Sci 94: 7685- 7690. 34

VINCENTZ, M., A. LEITE, G. NESHICH, G. VRIEND, C. MATTAR et al., 1997 ACGT and vicilin core sequences in a promoter domain required for seed-specific expression of a 2S storage protein gene are recognized by the opaque-2 regulatory protein. Plant Mol Biol. 34: 879-889. VINSON, C., P. SIGLER and S. MCKNIGHT, 1989 Scissors-grip model for DNA recognition by a family of leucine zipper proteins. Science 246: 911-916. VOGT, P. K., T. J. BOS and R. F. DOOLITTLE, 1987 Homology between the DNA-binding domain of the GCN4 regulatory protein of yeast and the carboxyl-terminal region of a protein coded for by the oncogene jun. PNAS 84: 3316-3319. WANG, X., and B. A. LARKINS, 2001 Genetic analysis of amino acid accumulation in opaque- 2 maize endosperm. Plant physiol 125: 1766-1777. WANG, X., Y.-M. WOO, C. S. KIM and B. A. LARKINS, 2001 Quantitative trait locus mapping of loci influencing elongation factor 1α content in maize endosperm. Plant Physiology 125: 1271-1282. WEIL, C. F., and R.-A. MONDE, 2007 Getting the point-mutations in Maize. Crop Science 47: S60-S67. WEINHOLDS, E., F. J. VAN EEDEN, M. KOSTERS, J. MUDDE, R. H. PLASTERK et al., 2003 Efficient target-selected mutagenesis in zebrafish. Genome Res. 13: 2700-2707. WEISS, M. A., T. ELLENBERGER, C. R. WOBBE, J. P. LEE, S. C. HARRISON et al., 1990 Folding transition in the DNA-binding domain of GCN4 on specific binding to DNA. Nature 247: 575-587. WILHELM, D., K. BENDER, A. KNEBEL and P. ANGEL, 1997 The level of intracellular glutathione is a key regulator for the induction of stress-activated signal transduction pathways including Jun N-terminal protein kinases and p38 kinase by alkylating agents. Molecular and Cellular Biology 17: 4792-4800. WILSON, C. M., 1991 Multiple Zeins from Maize Endosperms Characterized by Reversed- Phase High Performance Liquid Chromatography. Plant Physiol 95: 777-786. WOO, Y. M., D. W. N. HU, B. A. LARKINS and R. JUNG, 2001 Genomics analysis of genes expressed in maize endosperm identifies novel seed proteins and clarifies patterns of zein gene expression. Plant Cell 13: 2297-2317. WU, C.-Y., H. WASHIDA, Y. ONODERA, K. HARADA and F. TAKAIWA, 2000 Quantitative nature of the Prolamin-box, ACGT and AACA motifs in a rice glutelin gene promoter: minimal "cis"-element requirements for endosperm-specific gene expression. Plant J 23: 415-421. YUNES, J. A., G. CORD NETO, M. J. DA SILVA, A. LEITE, L. M. M. OTTOBONI et al., 1994a The transcriptional activator Opaque2 recognizes two different target sequences in the 22- kD-like α-prolamin genes. Plant Cell 6: 237-249. YUNES, J. A., G. CORD NETO, A. LEITE, L. M. M. OTTOBONI and P. ARRUDA, 1994b The role of the Opaque2 transcriptional factor in the regulation of protein accumulation and amino acid metabolism in maize seeds. An Acad Bras Cienc 66: 227-238. ZHAO, J. M., C. C. WILLIAMS and R. L. LAST, 1998 Induction of Arabidopsis tryptophan pathwayenzymes and camalexin by amino acid starvation, oxidative stress, and an abiotic elicitor. Plant Cell 10: 359-370.

CHAPTER 2: MOLECULAR CHARACTERIZATION OF ZMBZIP1, A DOWNSTREAM TARGET OF THE OPAQUE-2 TRANSCRIPTION FACTOR

Abstract Opaque-2 is a transcription factor that is responsible for regulating the deposition of

zein storage proteins into maize endosperm tissue. However, opaque-2 knock-out plants

have pleiotrophic phenotypes that negate their efficacy for commercial grain production. The

maize bZIP transcription factor ZmbZIP1 was identified in a transcript profiling experiment

as being differentially expressed when comparing the B45 and B45opaque-2 genotypes. The

ZmbZIP1 promoter region contained an opaque-2 binding sequence, suggesting that it was

regulated directly by opaque-2 (JIA 2004). ZmbZIP1 was physically mapped to maize

chromosome 9, contig 380, and a rice orthologue, RISBZ5 was found through synteny. To

further characterize ZmbZIP1, mRNA levels were compared in the genotypes B46 and b46o2 and were found to be not significantly different (p= 0.2188). Thus it appears that the opaque-

2 versus wildtype differential of ZmbZIP1 is a genotype specific phenomenon, or the degree of differential expression is too small to be detected. A tissue specific expression profile was also created, and ZmbZIP1 was found to be expressed in all tested maize tissues at differing levels. An attempt to make a ZmbZIP1 knock-down/knock-out mutant did not result in any phenotypic changes in either the maize plant as a whole or in alpha-zein endosperm deposition, suggesting either a functionally redundant protein exists, the knock-out alleles retained sufficient activity to be functional, or phenotypes produced by disruption of

ZmbZIP1 not detectable by the methods we used. 36

Materials and Methods

ZmbZIP1 Expression Levels

Tissue Collection and RNA Extraction B73 plants were grown during the summer of 2004, B46 plants were grown during

the summer of 2006 and B46o2 plants were grown during the summer of 2007, all at the

Iowa State University Agronomy Farm (Boone, IA). 14 days after pollination, a single plant

was used to harvest samples from six tissues: kernel, embryo, endosperm, pericarp, cob, and

mature leaf. Samples were immediately frozen in liquid N2 and stored at -80°C. To obtain immature leaf and root tissues, B73 plants were grown during winter 2005 and B46o2 plants were grown during summer 2006, both under greenhouse conditions in the Iowa State USDA greenhouse (Ames, IA). These tissues were collected 14 days after germination and were immediately frozen in liquid N2 and stored at -80°C. For both genotypes, endosperm polyadenylated RNA extraction was performed using the Promega PolyATract 1000

(Promega, Madison, WI) isolation kit without modification. For all other tissues, total RNA extraction was performed using the Ambion ToTally RNA kit (Ambion, Austin, TX), also without modification. The isolated RNA was then stored at -80°C until amplification reactions were performed.

Quantitative Real Time RT-PCR to Determine bZIP mRNA Levels in Eight Maize Tissues DNA was removed from approximately 1µg RNA with RQ1 RNAse free DNAse

(Promega, Madison WI). The sample was then incubated at 37°C for 30 min, after which 1µl

of RQ1 Stop Solution (Promega, Madison, WI) was added to the reaction, which was then

heat inactivated at 65°C for 10 min. Primers were designed to flank an intron in order to

differentiate DNA from RNA based upon the MAGI 6167 sequence from the Maize 37

Assembled Genomic Islands project (http://magi.plantgenomics.iastate.edu/) (FU et al. 2006).

Primer sequences are as follows: ActinF2 (CCT GAA GAT CAC CCT GTG CT), Actin R

(TGT ACG TGG CCT CAT GGA C), bZIP F3 (GCG GTC TAG GAG GAG GAA AC),

bZIP R2 (GGC CAC ACG ATA CAG CAC TTC TC). An actin standard curve consisting

of 4x – 1024x dilutions of the DNAse-free RNA was created for each tissue; each reaction of the dilution curve was assembled using the Stratagene Brilliant SYBR® Green QRT-PCR

Master Mix Kit (La Jolla, CA) as described in the manual with 200 nM of each primer of the

Actin F2/Actin R primer combination. Experimental reactions were created in triplicate

using the Stratagene Brilliant SYBR® Green QRT-PCR Master Mix Kit (La Jolla, CA) as

described in the manual with 200 nM each of the bZIP F3/bZIP R2 primer combination, with

a dilution of RNA (either 4x, 16x, or 64x) that was sufficient for the experimental reaction Ct

to fall within the range of Cts of the actin standard curve.

The above reactions were run in a Stratagene Mx3000P (Stratagene, La Jolla, CA)

thermocycler. Cycling protocols consisted of a first strand synthesis of 45°C for 45 min,

followed by 95°C for 10 min, followed by 40 cycles of 95°C for 30 sec, 59°C for 1 min,

72°C for 45 sec, followed by a dissociation curve, taking a fluorescent reading at every degree between 55°C and 95°C to ensure only one product of the correct Tm was being amplified. Using the average of cycles 4-14, the Stratagene analysis software established a fluorescence threshold below which amplification fluorescence levels were not statistically different than the background fluorescence levels. The cycle at which any amplicon crosses this threshold level is referred to as its Ct value. From the actin dilution curve, a standard curve of Ct values (y-axis) versus concentration logarithmic values (x-axis) was plotted and fit to a linear equation, and the relative concentration of the experimental ZmbZIP1 sample 38

and actin control sample were determined from the curve. The resulting actin concentration

was then subtracted from the ZmbZIP1 concentration in order to normalize results for

comparison within each genotype.

Quantitative Real-Time RT-PCR to Determine Endosperm bZIP mRNA Concentrations in Two Inbred Lines, B46 andB46o2 For the Real-time RT-PCR experiments, 1 ng of either B46 or B46o2 endosperm

mRNA was used as an initial template for each sample. An actin standard curve consisting

of serial 4x dilutions ranging from 200ug – 0.39ug of B46 leaf total RNA was created in

order to compare between the two tissues; each reaction of the dilution curve was assembled

using the Stratagene Brilliant II SYBR® Green QRT-PCR Master Mix Kit (La Jolla, CA) as

described in the manual with 200 nM of each primer of the Actin F2/Actin R primer

combination. Triplicate technical replicates for each experimental reaction were created

using the Stratagene Brilliant SYBR® Green QRT-PCR Master Mix Kit (La Jolla, CA) as

described in the manual with 200 nM each of the bZIP F3/bZIP R2 primer combination.

Triplicate control reactions of B46 and B46o2 were also created using the same protocol and

200 nM each of the Actin F2/Actin R primer combination in order to compare their relative concentrations; the B46 and B46o2 Actin Ct was required to fall within the range of Cts of the Actin standard curve in order for the experiment to be valid. All reactions were run in a

Stratagene Mx3000P (Stratagene, La Jolla, CA) thermocycler as noted above, and the results were analyzed using the delta-delta Ct method (LIVAK and SCHMITTGEN 2001) which adjusts

the relative concentration by normalizing the experimental ZmbZIP1 values based on actin

internal standard values in the same RNA sample in order to more accurately compare

concentrations across tissues and to ensure that differences in expression levels were not due 39 to varying amounts of RNA template being added to each set of reactions. mRNA copy number was determined using the formula:

-∆Ct -∆Ct Copy number = 2 exp – 2 actin and was averaged over three biological replicactes. Statisical significance was determined using the Student’s t-test with a threshold value of p<0.05.

“Virtual Northerns” to Evaluate ZmbZIP1 Expression in Maize Tissues The cDNA sequence that corresponds with the MAGI 6167 sequence was used to

BLAST the MaizeSeq.org EST database (www.maizeseq.org). The BLAST returned contigs made up of aligned ESTs out of a specific library. Contigs with an e-value less than 1x10-12 were considered to be actual representatives of the ZmbZIP1 or Actin expressed sequence tags present in the database. For each sequence, the EST name, the genotype, library name, and the tissue represented by that library were recorded and used to determine the percentage of ZmbZIP1 sequences present in the library. The Zea mays Actin1 sequence (NCBI:

J01238) was also mined, and the same data as above was collected in order to determine the percentage of Actin present in the library and to get a measure of the relative ZmbZIP1:Actin

ESTs present in the library so that a direct comparison could be made to the RT-PCR data.

Physical Mapping of the ZmbZIP1 Gene

BAC DNA Isolation All BAC clones corresponding to the overgo hybridization anchor CL7173_1 (NCBI accession: AY110737) were chosen as candidates to hold the ZmbZIP1 gene. When BAC clones were redundant (mapped to the same location on www.maizesequence.org), a representative BAC was chosen. The 13 BAC clones chosen (Table 6) were grown on duplicate LB plates inoculated with 12.5 µg/ml chloramphenicol and allowed to grow for 16 40 hours at 37°C. Two single BAC colonies were chosen from each set of duplicate plates and each was inoculated into separate flasks of 250 ml LB broth containing 12.5 µg/ml

Chloramphenicol and allowed to grow with shaking for 16 hours at 37°C. BAC DNA was isolated from these cultures as described by Sambrook and Russell (Sambrook and Russell

2001), without modification.

BAC DNA Amplification 200 ng of DNA from each BAC served as an initial PCR template. Primers were designed to amplify the 3’ region of the ZmbZIP1 gene and to produce approximately 1 kb amplicons. Reactions were performed in a 20µl volume containing 10x PCR buffer, minus

Mg2+ (200 mM Tris-HCl (pH 8.4), 500 mM KCl) (Invitrogen, Carlsbad, CA), 1.5 mM

MgCl2, 0.2 mM each dNTP, 0.5 µM each 5’ and 3’ primer (Table 1), and 0.625 U Platinum®

Taq DNA Polymerase (Invitrogen, Carlsbad, CA).

Amplification was carried out under the following conditions: an initial denaturation at 95°C for 2min, followed by 35 cycles of 95°C for 45sec, Tm°C (Table 1) for 45s, 72°C for

2m, and a final extension of 72°C for 10m. Reactions were electrophoresed on a 1% agarose gel to determine whether there was amplification with each primer set from the BAC DNA template.

Primer Primer sequence Tm (°C) Name ZmbZIP1 3F 5’ ACC AAC CTT TCC ACG TAC CA 3’ 60.0 ZmbZIP1 3R 5’ AGA GCA TGA TTG GCT GAG GT 3’ 59.9 ZmbZIP1 4F 5' TCA AGT ATT GCT CAC CTG CCG TAT TGC 3' 69.7 ZmbZIP1 4R 5' ATG GTT CAA GAG AAC TCG TGC CAG AAG 3' 68.6 ZmbZIP1 5F 5’ CTG ATT CAA AGC TTG CTC CA 3’ 60.0 ZmbZIP1 5R 5’ GCC TAC CCA AAA TTT GCT TC 3’ 60.0 Table 1. Primer sequences used to amplify the ZmbZIP1 gene in BAC genomic DNA.

PCR amplicon bands from BAC ZMMBBb0173M13 were excised from the gel and

purified using the Montage DNA Gel Extraction kit (Millipore Corp, Bedford, MA) and 3µl

of each resulting DNA sample were cloned into the TOPO-TA 2.1® cloning vector

(Invitrogen, Carlsbad, CA) and transformed into Stratagene XL-1 Blue competent E. coli

cells (Stratagene, La Jolla, CA), each as described by the manufacturer. The cells were then

grown for 16 hours on LB plates containing 50 µM Ampicillin on which 10µl of 100mM

IPTG and 40ul of 20mM X-Gal had been spread. Single colonies were selected, and plasmid

DNA was purified using the QIAprep Spin Miniprep Kit ™ (Qiagen, Valencia, CA). DNA

was then sequenced in the forward and reverse directions at the Iowa State University DNA

Facility using the Applied Biosystems 3730xl DNA Analyzer (Applied Biosystems, Foster

City, CA) and aligned to MAGI 6167 sequence (http://magi.plantgenomics.iastate.edu/) (FU

et al. 2006) using VectorNTI AlignX (Invitrogen, Carlsbad, CA) with optimal settings.

Creation of a ZmbZIP1 TILLING Mutant to Attempt to Alter ZmbZIP1 Protein Expression

DNA Isolation To prepare W22 and B73 DNA, plants were grown under greenhouse conditions and

100 mg of leaf tissue was harvested at 14 days post germination. Duplicate samples were

prepared using the MasterPure™ Plant Leaf DNA Purification Kit (Epicenter

Biotechnologies, Madison, WI) with no modification and were quantified using a NanoDrop

ND-1000 Spectrophotometer (NanoDrop Technologies, Wilmington, DE).

Primer Creation and Specificity Testing Maize mutants were ordered from the Maize TILLING project

(http://genome.purdue.edu/maizetilling/). The CODDLe program

(http://www.proweb.org/input/) was used determine output primers adequate for use in the 42

TILLING procedure as described in Till et al (TILL et al. 2004). Primer sequences are as follows: MAGI6167 3-64 F (TCA ACG ATT GCT CAC CTG CCG TAT TCG) and MAGI

6167 3-64 R (ATG GTT CAA GAG AAC TCG TGC CAG AAG). Primers were tested for specificity in the maize inbred lines W22 and B73.

PCR reactions were performed as described at the Maize TILLNG project website

(http://genome.purdue.edu/maizetilling) with the following modifications. For each sample,

5 µl of master mix was made consisting of 10x ExTaq buffer minus Mg2+ (20 mM Tris-HCl

(pH8.0), 100 mM KCl, 0.1 mM EDTA, 1 mM DTT, 0.5% Tween 20, 0.5% Nonidet P-40,

50% Glycerol) (TaKaRa Bio USA, Madison, WI), 2.75 mM MgCl2, 0.375 mM each dNTP,

0.3 µM each primer, 0.25 U ExTaq Hot Start Polymerase (TaKaRa Bio USA, Madison, WI).

To this mixture, 5 µl of either W22 or B73 DNA was added to a concentration of 0.9 ng/µl.

Amplification was carried out under the following conditions as suggested at the

Maize TILLING Project website (http://genome.purdue.edu/maizetilling). Amplification

conditions were as follows: denaturation at 95°C for 2min, followed by an initial

amplification step of 8 cycles of 94°C for 20sec, 73°C (-1°C/cycle) for 30 sec, ramp to 72°C

at 0.5°C/sec, 72°C for 1m, followed by amplification for 45 cycles of 94°C for 20 sec, 65°C

for 30 sec, ramp to 72°C at 0.5°C/sec, 72°C for 1min, followed by a final extension step at

72°C for 5 min. The samples were then electrophoresed on a 1% agarose gel. Each

genotype produced a single ~1kb amplicon, each of which were excised from the gel and

purified using the Montage DNA Gel Extraction kit (Millipore Corp, Bedford, MA).

To confirm that the primers were amplifying the correct section of ZmbZIP1, 2 µl of

each resulting DNA sample were cloned into the TOPO-TA 2.1® cloning vector (Invitrogen,

Carlsbad, CA) and transformed into Stratagene XL-1 Blue competent E. coli cells 43

(Stratagene, La Jolla, CA), each as described by the manufacturer. The cells were then

grown for 16 hours on LB plates containing 50 µM Ampicillin on which 10 µl of 100mM

IPTG and 40 µl of 20mM X-Gal had been spread. Single colonies were selected, and

plasmid DNA was purified as in Sambrook and Russell (SAMBROOK and RUSSELL 2001) with

minor modification. The purified samples were sequenced in the forward and reverse

directions at the Iowa State University DNA Facility using the Applied Biosystems 3730xl

DNA Analyzer (Applied Biosystems, Foster City, CA) and analyzed using VectorNTI

AlignX (Invitrogen, Carlsbad, CA) with optimal settings.

Having shown the primers to be suitable for TILLING, the primer sequences were

submitted to the Maize TILLING Project (http://genome.purdue.edu/maizetilling) and

screening commenced as described in Till et al (TILL et al. 2004). Following a successful

screen, seeds from the line NW2622 were ordered from the TILLING stock center.

Extraction of NW2622 DNA for TILLING Genotyping Seeds from the NW2622 Maize TILLING line were planted in the field during the

summer of 2007 at the Agronomy Farm of Iowa State University (Boone, IA). 100mg of leaf tissue was harvested from each plant at 34 days after germination. Genomic DNA was extracted using the MasterPure™ Plant Leaf DNA Purification Kit (Epicenter

Biotechnologies, Madison, WI) without modification and quantified using a NanoDrop ND-

1000 Spectrophotometer (NanoDrop Technologies, Wilmington, DE).

Identification of mutant ZmbZIP1 alleles via SNaPshot Genotyping 150 ng of DNA from each TILLING sample served as the initial template for PCR

reactions. Primers initially designed for the TILLING procedure described above were used to amplify ~1 kb fragments of the ZmbZIP1 gene. PCR reactions were performed in 30 µl 44

volumes consisting of 10x PCR buffer, minus Mg2+ (200 mM Tris-HCl (pH 8.4), 500 mM

KCl) (Invitrogen, Carlsbad, CA), 1.5 mM MgCl2, 0.25 mM each dNTP, 0.5 µM each 5’ and

3’ primer (Table 1), and 0.625 U Platinum® Taq DNA Polymerase (Invitrogen, Carlsbad,

CA). Amplification was carried out under the following conditions: denaturation at 95°C for

2min, followed by an initial amplification step of 8 cycles of 94°C for 20 sec, 73°C (-

1°C/cycle) for 30 sec, ramp to 72°C at 0.5°C/sec, 72°C for 1 min, followed by amplification

for 45 cycles of 94°C for 20 sec, 65°C for 30 sec, ramp to 72°C at 0.5°C/sec, 72°C for 1m,

followed by a final extension step at 72°C for 5min. 15 µl of reaction were then

electrophoresed on a 1% agarose gel to ensure that product was present.

Unincorporated dNTPs and single stranded DNA was removed from the remaining 15

µl of sample using 2U Exonuclease I (Fermentas, Hanover, MD) and 5U Shrimp Alkaline

Phosphatase (SAP) (Fermentas, Hanover, MD). Samples were incubated at 37°C for 1h, followed by heat inactivation at 75°C for 15 min.

The resulting PCR products were re-amplified using the ABI Prism® SNaPshot™

Multiplex Kit (Applied Biosystems, Foster City, CA). Experimental and control reactions were carried out as indicated by the manufacturer, with the exception that a half-reaction size was used. Experimental primer sequence was designed to correspond to the sequence immediately 5’ to a G/A SNP present at basepair 2774 in the MAGI 6167 sequence and is as follows: TILLING genotype (ACA GAA TCC TCA AAT CCG ATG TA). The primer was added to the reaction at a 0.2 µM concentration. Amplification procedures were carried out under the following conditions: 96°C for 1 min, followed by 25 cycles of 96°C for 10 sec,

50°C for 5 sec, 60°C for 30 sec. 45

Following re-amplification, excess [F]ddNTPs were immediately removed from the reaction as follows. 0.5 U of SAP (Fermentas, Hanover, MD) was added to each sample,

after which they were incubated at 37°C for 1 h, and the enzyme was heat inactivated at 75°C

for 15 min. Samples were then genotyped at the Iowa State University DNA Facility using

the Applied Biosystems 3100 DNA Analyzer (Applied Biosystems, Foster City, CA).

Results were analyzed for single nucleotide polymorphism (SNP) mutations using the

SoftGenetics Genemarker™ (SoftGenetics LLC, State College, PA) analysis program.

Zein Measurements using HPLC Pericarp tissue was removed from 40 kernels from the TILLING line NW2622 with

sandpaper and 10mg of powdered endosperm was excised from each kernel using a drill. 10

volumes of extraction buffer (70% (v/v) ethanol, 61 mM (v/v) sodium acetate and 5% (v/v)

2-mercaptoethanol) were added to each sample. The sample was then shaken horizontally at

37°C for 1 hour, followed by centrifugation at 14,000 rpm for 10 min. The supernatant was then placed into a clean tube, and diluted 4:1 with extraction buffer. Each sample was separated on an RP-C18 column. Chromatography employed a gradient elution of solvents A

(99.9% (v/v) acetonitrile, 0.1% (v/v) trifluoroacetic acid) and B (99.9% H20, 0.1% (v/v)

trifluoroacetic acid (PAULIS et al. 1991; WILSON 1991). 25 µl samples were injected into a

Waters 717 Autosampler HPLC system, and the UV absorbance of the column effluent was

monitored at 210 nm. Elution profiles were compared using Igor Pro (WaveMetrics Inc,

Lake Oswego, Oregon) statistical software. 46

Results and Discussion

ZmbZIP1 Has Varying Transcript Levels in Different Tissues

Verification of Endosperm Microarray Results using B46 and B46o2 via QRT-PCR Endosperm mRNA from the maize genotypes B46 and B46o2 were comparatively

analyzed so that we could directly determine whether there was a change in ZmbZIP1

transcript level when the opaque-2 gene was knocked out. Since the Opaque-2 protein is

hypothesized to be a direct regulator of ZmbZIP1 due to the dual Opaque-2 binding sites in the ZmbZIP1 promoter, it was assumed that the ZmbZIP1 transcript levels would be lower in the opaque-2 knockout than in the wildtype endosperm. However, QRT-PCR data showed that there was only a 1.65 fold change between the B46 and B46o2 endosperm ZmbZIP1 levels, with the B46 genotype producing more ZmbZIP1 transcript (Fig. 1). This compares well to previous semi-quantitative RT-PCR results performed in B45, which show a higher level of ZmbZIP1 transcript present in the wildtype genotype when compared to the opaque-

2 genotype (Jia, unpublished data). However, the current expression study comparing B46 to

B46o2 showed less than a two fold differential in mRNA expression, unlike the >2 fold expression change in the original B45/B45o2 microarray results (JIA 2004) from which

ZmbZIP1 was chosen for study. Furthermore, this differential in ZmbZIP1 mRNA

expression in B46 compared to B46o2 is not statistically significant at a 95% level of

confidence using the Student’s t-test (p=0.2188, Table 2). Since the microarray experiment was performed in B45/B45o2, while the current study was undertaken in the B46/B46o2,

there is a strong likelihood that a genotypic effect is causing the differences in expression.

The lack of differential expression is surprising as it suggests that the Opaque-2 transcription

factor may either not be actively regulating ZmbZIP1, or that the action of Opaque-2 may be 47

modified by other transcription factors. There is a precedent for this in the bZIP family; it

has been shown that these transcription factors may act as either homo- or hetero-dimers to

more finely control gene regulation (FOULKES et al. 1991; PABO and SAUER 1992).

ZmbZIP1 expression in B46 and B46o2 Endosperm

2.4

1.6

1.2

0.8 Fold Change (2^delta-delta Ct)

0.4

0 B46 B46o2

Figure 1. ZmbZIP1 expression in B46 and B46o2 endosperm. The 1.65 cycle fold change is not significantly different between the two genotypes; therefore ZmbZIP1 is not differentially expressed between the two tissues. The maroon line denotes the 2 fold level of expression change that has been deemed to be the significant level.

Mean Copy Number Fold Change Genotype (2-∆Ct) (n=3)a Std. Dev (2-∆∆Ct)b p-value B46 2.18 x 10-7 2.39 x 10-7 1.65 0.2188 B46o2 1.72 x 10-8 2.28 x 10-7 1.00 Table 2. Analysis of 2-∆Ct in the B46 and B46o2 genotypes. The mean copy number of the two genotypes are not statistically different than each other by a Student’s t-test. The results are an average of three technical replicates on three biological replicates. a Mean copy number is normalized to actin content. b Fold change is a comparative value between the two genotypes; therefore, in order to express this value in both genotypes, the fold change of the B46o2 genotype has been normalized to 1.00. 48

Transcript Levels in Different Tissues of B73 and B46o2 Measured by QRT-PCR Jia noted that the ZmbZIP1 promoter contained two tandem Opaque-2 binding sites, which suggests that ZmbZIP1 may be directly regulated by the Opaque-2 protein (JIA 2004;

SCHMIDT et al. 1992). Since the Opaque-2 protein is located solely in the endosperm

(SCHMIDT et al. 1990), it became interesting to determine whether ZmbZIP1 is also endosperm specific, or if it has different tissue specificity. In order to do this, total RNA was extracted from seven tissues (the cob, embryo, immature and mature leaf, kernel, pericarp, and root) and polyadenylated mRNA was extracted from endosperm tissue, each from the maize inbred lines B73 and B46o2 in order to determine the level of ZmbZIP1 transcript present in each tissue. The samples were evaluated using quantitative real-time RT-PCR, and compared to an actin internal standard in order to determine the relative levels of

ZmbZIP1 transcript present in both a wild-type and opaque-2 background. ZmbZIP1 transcript was present in all tissues tested in both the wild-type and opaque-2 lines; however, there was more variation in concentration between the tissues of the wild-type than the opaque-2 line (Table 3, Table 4). In both the wild-type and opaque-2 lines, the tissues that contained the lowest and highest level of ZmbZIP1 are the same; the lowest relative level was found in embryonic tissue (4.80% in B73, 27.25% in B46o2) while the highest level was found in immature leaf tissue (41,089% in B73, 1451% in B46o2). The level of ZmbZIP1 transcript present in immature leaf tissue was surprisingly large in comparison to the rest of the samples. Because of this high level of ZmbZIP1, it was necessary to dilute the RNA added to the ZmbZIP1 amplification reactions to 1/2048 its normal concentration in order to get the immature leaf ZmbZIP1 transcript to fall within the confines of the actin standard curve. (If dilutions were necessary in other tissues, they were typically 4 – 64 fold). This has 49 the potential to introduce a large amount of dilution error in both the ZmbZIP1 and actin results. Transcript concentrations in roots seemed particularly high in both the B73 and

B46o2, and were relatively consistent between the two genotypes. This result was interesting since there seems to be wide variation in the percentage of ZmbZIP1 transcript present between the two genotypes in every other tissues, even those that are consistently high or low when being compared within genotypes. The most surprising result was a 10-fold increase in the endosperm ZmbZIP1 level in the opaque-2 genotype when compared to the wild-type

B73 expression level. This result can potentially be explained by the variation in genotype.

Jia (2004) showed that there is a wide variation in repression or activation by opaque-2 in different genotypes, so there may simply be less repression in the B46 genotype or less activation in the B73 genotype. (The question of repression of ZmbZIP1 by Opaque-2 is further addressed below through a comparison of B46 and B46o2 endosperm mRNA levels.)

Overall, this suggests it is dangerous to compare too closely between the wild-type and opaque-2 expression levels, as they were taken from different genotypes, and prior experiments have shown genotypic variation in ZmbZIP1 expression levels (JIA 2004).

ZmbZIP1 Actin ZmbZIP concentration concentration relative concentration Cob 0.05 0.94 5.45% Embryo 0.0030 0.06 4.80% Endosperm 0.27 0.39 68.16% Immature leaf 0.20 0.00049 41089.12% Kernel 1.19 0.61 196.37% Mature leaf 0.18 0.26 68.69% Pericarp 0.24 1.01 24.10% Root 0.35 0.02 1525.75% Table 3. Quantitative real-time RT-PCR expression levels for ZmbZIP1 in several B73 maize tissues. Maize ZmActin1 gene expression levels were used as an internal standard to determine relative levels of ZmbZIP1. 50

ZmbZIP1 Actin ZmbZIP1 concentration concentration relative concentration Cob 3.12 0.86 361.31% Embryo 0.08 0.28 27.75% Endosperm 0.31 0.07 463.96% Immature leaf 0.35 0.02 1451.76% Kernel 0.05 0.04 133.50% Mature leaf 0.73 0.06 1306.68% Pericarp 0.48 0.04 1213.09% Root 0.52 0.04 1300.78% Table 4. Quantitative real-time RT-PCR expression levels for ZmbZIP1 in several B46- opaque2 maize tissues. Maize ZmActin1 gene expression levels were used as an internal standard to determine relative levels of ZmbZIP1.

Tissue Comparisons: Virtual Northern Versus B73 and B46o2 QRT-PCR Virtual northern data mining using the MaizeSeq.org EST database was also

performed in order to determine the tissue specificity of ZmbZIP1 in a broad variety of

genotypes and tissues. It was hoped that these data would reinforce RT-PCR results and give a broader understanding of ZmbZIP1 expression levels in the plant. However, the results showed, in general, a much lower level of ZmbZIP1 present in each tissue than the RT-PCR data (Table 5, Fig. 2). Comparisons are difficult with the EST libraries that comprise the

MaizeSeq.org database because often genotypes, experimental growth conditions, tissue used to make the library, and the type of experiment for which the library was created (microarray,

BAC cloning, etc) were not noted when describing the ESTs present in a library. This lack of data made it difficult to determine the reason for any discrepancies. One factor contributing to lower relative levels of ZmbZIP1 transcript in virtual northern may be that the experimental conditions under which the library was created are unknown, and these conditions may have either upregulated actin expression, or conversely, downregulated

ZmbZIP1 expression. Finally, since maize actin is a relatively highly conserved (MEAGHER 51

1991) gene family comprising at least eight genes and one pseudogene (MONIZ DE SA and

DROUIN 1996), it is likely that when the whole maize Actin1 (MAc1) cDNA sequence

(NCBI: J01238) was used to query the database, other actin sequences were also extracted, despite selecting the stringent E-value threshold of 10-12 This threshold was chosen because of a natural break E-values the of the assembled contig sequences returned in the BLAST.

For purposes of comparison to RT-PCR data, it may then be better to use 1/8 of the amount of actin present in the library as an appropriate normalization tool for bZIP levels in these tissues assuming each actin transcript is transcribed equally. While this transformation does not equalize the virtual northern values to those of the QRT-PCR data, several values

(endosperm, mature leaf/leaf) are within an order of magnitude of each other when this method is used.

Sequences ZmbZIP1 ZmMAc1 ZmbZIP: ZmbZIP: per tissue ZmbZIP1 sequences ZmMAc1 sequences ZmMAc1 1/8 ZmMAc1 (total) sequences (%) sequences (%) (%) (%) Tassel 8503 4 0.05% 27 0.32% 14.81% 118.48% Root 65216 11 0.02% 140 0.21% 7.87% 62.96% Anthers 4899 1 0.02% 5 0.10% 20.00% 160.00% Kernel/seed 26667 4 0.01% 64 0.24% 6.25% 50.00% Cob 3519 1 0.03% 12 0.34% 8.33% 66.64% Callus/cell culture 12753 8 0.06% 20 0.16% 40.00% 320.00% Endosperm 27745 4 0.01% 35 0.13% 11.43% 91.44% Leaf (and sheath) 58846 17 0.03% 191 0.32% 8.90% 71.20% Stalk/ veg. shoot/ internode 19500 5 0.03% 16 0.08% 31.25% 250.00% Meristematic cell 6294 1 0.02% 13 0.21% 7.69% 61.52% Ear 30680 3 0.01% 76 0.25% 3.95% 31.60% Whole plant 102112 8 0.01% 371 0.36% 2.16% 17.28% Total Sequences 366734 67 0.02% 970 0.26% 6.91% 55.28% Table 5. Virtual Northern of ZmbZIP1 levels in several maize tissues. The MaizeSeq.org database was used to compile expression data. The maize Actin1 expression level was used as an internal standard to determine relative ZmbZIP1 levels. 52

Relative tissue specific levels of ZmbZIP:Actin via Electronic Northern

20 bZIP:Actin (%) 15

b oot o Ear R C Tassel heath) Anthers nel/seed s r hole plant e t/ internode K Endosperm o W sho Callus/cell culture Leaf (and Meristematic cell

Stalk/veg. Tissue

Figure 2. Tissue specific levels of ZmbZIP1 relative to Actin via Electronic Northern. The percentage of ZmbZIP1 ESTs present in each tissue specific library was compared to the percentage of an Actin control EST in order to determine a relative percentage of ZmbZIP1.

Physical Mapping of the ZmbZIP1 Locus Another aspect of characterizing ZmbZIP is to determine where in the genome it is

located and if there is any synteny between ZmbZIP1 and genes in other plant species. In

order to accomplish this, the genomic sequence for ZmbZIP1 was compared against the

Maize Genetics and Genomics Database (MaizeGDB) using their BLAST function

(www.maizegdb.org) and was found to correspond to the overgo hybridization anchor

CL7173_1 (NCBI accession: AY110737) (GARDINER et al. 2004). The CL7173_1 overgo

has been mapped to at least 4 places in the maize genome- on chromosomes 2, 4, 5, and 9

(www.maizeGDB.org) – which suggests at least some level of gene similarity, and perhaps duplication, in these regions. Bacterial artificial chromosomes (BACs) hybridizing to the 53

CL7173_1 marker (Table 6) were ordered from the Clemson University Genomic Institute

(CUGI) (www.genome.clemson.edu) and the Children’s Hospital and Research Center at

Oakland (CHRCO) BAC PAC Resources (http://bacpac.chori.org). The first build of the

maize genome sequence (www.maizesequence.org) was used to determine whether any of

the selected clones had been mapped to the maize genome and to determine their respective

positions if they had been. Each clone was PCR amplified using ZmbZIP1 specific primer

sets 3, 4, and 5 (Table 1) to identify BACs containing the ZmbZIP1 gene. The section of

DNA amplified by primer set 4 contained the CL7173_1 overgo fragment and primer sets 3

and 5 flanked that region, so these three primer sets were chosen to PCR amplify portions of the BAC as a screen, since the BACs containing the full-length ZmbZIP1 gene should contain all three regions. Primer set 4 amplified in approximately ¾ of the clones to produce an amplicon of ~1kb (BACs ZMBBb0025A22, ZMBBb0104G08, ZMBBb1073M13,

ZMBBb0183K23, ZMBBb0059O02, ZMBBb0022J17, ZMBBc0182J16, CH201-214C04 and CH201-206I22). Amplification by primer set 4 in the majority of the clones was expected for two reasons: first, because primer set 4 amplifies the part of ZmbZIP1 that corresponds to the most highly conserved region of the bZIP gene family (the basic domain and the leucine zipper region); secondly the CL7173_1 overgo hybridization site that was

used to identify these BACs is contained within this amplicon. However, only clone

ZMBBb0173M13 produced a product when primer set 3 (~1kb) and primer set 5 (900 bp) were used, making it the most likely candidate to contain the full-length ZmbZIP1 gene.

Direct sequencing with primer sets 1-5 have confirmed that the full ZmbZIP1 sequence is present on this BAC. This BAC has been mapped to maize chromosome 9s, contig 380.

Interestingly, when compared to the rice (Oryza sativa) genome using Gramene 54

(www.gramene.org) this portion of chromosome 9s shows synteny to rice chromosome 6. In an equivalent location in this syntenic region is the rice gene risbz5 (NCBI accession

AB053474), a member of the rice Opaque-2 like family (ONODERA et al. 2001) that shows

62% protein identity (Fig. 3) and 68.6% DNA sequence identity (Fig. 4) with ZmbZIP1.

Figure 3. Protein identity between ZmbZIP1 and risbz5. 62% identity was found between these two proteins. Asterisks denote identical amino acids.

1 50 ZmbZIP1 ATGAAGAAGT GCGCGTCGGA GCTGGAGCTG GAGGCGTTCA TCCGGGAG-A risbz5 ATGAAGAAGT GCCCGTCGGA GCTGCAGCTG GAGGCGTTCA TCCGGGAGGA ********** ********** **** ***** ********** ******** *

51 100 ZmbZIP1 G--CGGCGAG GACGCCCGCG ---CCGCCG------CC GGAGGTAGCA risbz5 GGCCGGCGCC GGCGACCGCA AGCCCGGCGT GTTATCTCCC GGCGACGGCG * ***** * ** **** *** ** ** ** * **

101 150 ZmbZIP1 GTCCGGGGTG CGG--TGGA- TCAAGCGATC CCGGAGGGAG CGGCGTCTTC risbz5 CGCGTAAGTC CGGCCTGTTC TCTCCCGGCG ACGGCGAGAT GTCCGTGTTG * ** *** ** ** ** *** * ** *** **

151 200 ZmbZIP1 --TCACCCGG CTTCGGTTTC --GCCGACTC GGACACCATG GATGGAGGCA risbz5 GATCAGAGTA CACTGGACGG AAGCGGCGGC GGCCACCAGC TGTGGTGGCC *** * ** ** * * ** ***** *** *** 201 250 ZmbZIP1 GTTGGTGGTA CGGGAACGTC CGCACGCCGA ACCCA------GTCATGT risbz5 GGAGAGCGTC CGTACGCCGC CGCGCGCCGC CGCCGCCTTC TCGGCCACGG * * ** ** * * *** ***** ** ** *

251 300 ZmbZIP1 CGCAGG------CGGCG TCCATATCCG CTAGCCCCGG GCTAACCACC risbz5 CCGACGAGCG GACGCCGGCG TCCATCTCCG ATGACCCCAA ACCAACCACC * * * ***** ***** **** * **** *******

301 350 ZmbZIP1 TCAGCCAATC ATGCTCTTGA AAGCGAGTCA GACTCCGACA GCGAATCACT risbz5 TCAGCGAACC ACGCGCCTGA AAGCGACTCG GACTCCGATT GCGATTCGCT ***** ** * * ** * *** ****** ** ******** **** ** **

351 400 ZmbZIP1 GTATGAGGTA GAGGGAGTTC CATACGAGCG AGGTAACAGA TCCATTGAGA risbz5 GTTAGAAGCA GAGAGGAGTC CACGCCTGCG TGGCACGAAA TCCACAGAAA ** ** * * *** * ** ** * *** ** * * * **** ** * 401 450 ZmbZIP1 CGAAGCGAAT AAGAAGGATG GTGTCCAATA GGGAGTCTGC GCGGCGGTCT risbz5 CAAAGCGAAT AAGAAGGATG GTGTCCAACA GGGAGTCCGC TCGACGATCC * ******** ********** ******** * ******* ** ** ** ** R M V S N R E S A R R S 451 500 ZmbZIP1 AGGAGGAGGA AACAGGCACA GTTGTCTGAC CTTGAGTCAC AGGTTGAACG risbz5 AGGAGGAGAA AGCAGGCACA GTTATCTGAA CTCGAATCAC AGGTCGAGCA ******** * * ******** *** ***** ** ** **** **** ** * R R R K L 501 550 ZmbZIP1 ACTCAAAGGT GAAAACGCAA CACTGTTCCA GCAACTTTCA GATGCCAACC risbz5 ACTCAAAGGC GAAAACTCAT CCCTCTTCAA GCAGCTCACA GATGCCAGCC ********* ****** ** * ** *** * *** ** ** ******* ** L L A

551 600 ZmbZIP1 AACAGTTCAG TACTGCAGTC ACAGACAACA GAATCCTCAA ATCCGATGTA risbz5 AGCAGTTCAA TACAGCGGTC ACGGACAACA GGATCCTCAA ATCGGATGTA * ******* *** ** *** ** ******* * ******** *** ****** A L 601 650 ZmbZIP1 GAAGCGTTAA GAATTAAGGT AAAGATGGCA GAGGATATGG TAGCGAGAAG risbz5 GAGGCCTTAA GAGTCAAGGT CAAGATGGCT GAAGACATGG TCGCGAGGGC ** ** **** ** * ***** ******** ** ** **** * ***** L 651 700 ZmbZIP1 TGCTGTATCG TGTGGCCTAG GCGACCTTGG CCTGGCACCA TACGTGAACT risbz5 CGCGATGTCG TGTGGCCTGG GCCAGCTCGG GCTGGCGCCA TTGCTCAGCT ** * *** ******** * ** * ** ** ***** *** * * * **

701 750 ZmbZIP1 CAAGGAAGAT GTGCCAAGCT TTGAATGTGC TCACAGGGTT GGATTTACTA risbz5 CCAGGAAGAT GTGCCAAGCT TTGGATATGC TCAG------TTTACCA * ******** ********** *** ** *** *** ***** *

751 800 ZmbZIP1 GGGAGTGATG CGT------T CAGGGGTCCA ACCGCAGGTA CACGAGTACA risbz5 CGGAACGATG CCTGTGGTTT CAAAGGCTTG AACCTGGGTC GACAGGTTCA *** **** * * * ** ** * * *** ** ** ** 56

801 850 ZmbZIP1 GAACTCACCA GTACAGAGCA CTGCAAGCCT AGAGAGTCTG GATAACCGAA risbz5 GAACTCACCG GTTCAAAGCG CTGCAAGCCT AGAGAGCCTG GACAACCGGA ********* ** ** *** ********** ****** *** ** ***** *

851 894 ZmbZIP1 AGTCCAACGA GGTGACCAGT TGCGCGGCGG ACATTTGGCC TTGA risbz5 TATCCAGCGA GGTGACCAGC TGCTCGGCTG ATGTGTGGCC T--- **** *** ********* *** **** * * * ***** * Figure 4. Coding sequence comparison between ZmbZIP1 and risbz5. Sequence comparison shows 68.6% identity between the regions, which is denoted by asterisks. The underlined sequence comprises the conserved basic region of the encoded protein, while double underlines denote the leucine heptad repeats.

Currently, a second BAC clone, ZMMBBb0139N19, (NCBI accession AC188012) is in the process of being sequenced and aligned by the Washington University Genome

Sequencing Center (http://genome.wustl.edu) as part of the Maize Sequencing Consortium.

This clone contains the genomic sequence of ZmbZIP1 in its entirety, and it also maps to

chromosome 9s, contig 380 (www.maizesequence.org), effectively overlapping clone

ZMMBBb1073M13 and further solidifying the position of ZmbZIP1 in the maize genome.

BAC name Maize Chromosome Location Library of Origin CH201-199E08 Chromosome 2, ctg 88 CHRCO CH201-182J16 CHRCO CH201-163-I22 CHRCO CH201-214C04 CHRCO CH201-182K15 CHRCO CH201-206I22 Chromosome 6, ctg 257 CHRCO ZMMBBb0025A22 Chromosome 4, ctg 164 CUGI ZMMBBb0104G08 Chromosome 5, ctg 214 CUGI ZMMBBb0173M13 Chromosome 9, ctg 380 CUGI ZMMBBb0183K23 CUGI ZMMBBb0059O02 CUGI ZMMBBb0022J17 CUGI ZMMBBb0066F06 CUGI Table 6: Maize BACs corresponding to the overgo hybridization anchor CL7173_1 (NCBI accession: AY110737) and their corresponding placement in the maize genome, when known. 57

TILLING Mutants

SNaPshot Genotyping In order to further attempt to characterize the ZmbZIP1 gene, we next identified a mutant for the gene to determine whether there was any phenotype created by altering the gene. To accomplish this, a TILLING mutant was identified for the ZmbZIP1 gene and grown in the field. This line was designated NW2622 and contained a G to A SNP at basepair 2774 in the ZmbZIP1 gene sequence. This SNP alters the protein coding sequence in the alpha-helical dimerization “zipper” motif of ZmbZIP1 from a negatively charged glutamic acid residue to a positively charged lysine residue, which has the potential to alter bonding in the alpha-helix and perhaps disrupt protein structure. Seventeen M3 plants were genotyped for the TILLING polymorphism and a single plant was homozygous for the wild- type allele, while 11 plants were heterozygous and 5 plants were homozygous for the new allele produced by the TILLING procedure (Table 7).

Seed morphology and general plant appearance were investigated for phenotypic

changes corresponding to the genotypic variation seen among the M3 generation of

TILLING mutant plants; however, no visible phenotype co-segregated with the mutant allele.

Genotype (SNP) Number of M3 plants Percentage of M3 plants Wild-type (G/G) 1 5.9% Heterozygote (G/A) 11 64.7% TILLING “mutant” (A/A) 5 29.4% Table 7. SNP genotype of M3 plants received from the Maize TILLING line NW 2622. Each plant was genotyped using the SNaPshot Genotyping platform from Applied Biosystems (Foster City, CA).

HPLC Shows No Change in Zein Deposition in TILLING Mutants Since ZmbZIP1 is downstream of, and potentially controlled by, the Opaque-2

protein that is vital for zein deposition in maize, the TILLING mutants were also screened to 58

see if there was any correlation between the altered ZmbZIP1 gene and a change in the level

of zein deposition in the endosperm. Zein protein was extracted from seeds from the M3

generation of the TILLING mutant line NW2622 and resolved and quantified by high

performance liquid chromatography (HPLC) using a RP-18C column in order to determine if

there were any significant differences between the zein profiles that corresponded to the

genotypic change in ZmbZIP1. However, there was no apparent difference among the zein

levels in the wild-type, heterozygotes or mutant plants when corrected for any differences in

injection volume (Fig. 5).

The lack of visible phenotypes in the ZmbZIP1 mutants may be due to several

factors. The first is that a single nucleotide or amino acid change, even in the most vital

region of the protein, may not be enough to knock-down the function of the ZmbZIP1 protein

enough to affect the visible phenotype. Secondly, there is the possibility of functional

redundancy between ZmbZIP1 and another (or several other) bZIP proteins. The opaque-2

like family of bZIP proteins, to which ZmbZIP1 belongs, has been documented to have

several members present in the maize endosperm, including O2, OHP-1 and OHP-2 (PYSH et al. 1993); a opaque-2 like gene family for with strong similarity to the ZmbZIP1 protein has also been documented in rice (RISBZ1-5) (ONODERA et al. 2001). Zein deposition specifically may not have been altered in the ZmbZIP1 mutant plants because the ZmbZIP1 protein may not be directly involved in zein deposition. The microarray in which ZmbZIP1 was gene was identified was based in differential expression between wild-type and opaque-

2 genotypes. 59

γ-zeins

α-zeins

Figure 5. HPLC comparison of Maize TILLING line NW2622 by genotype. From the HPLC traces, no differences in kernel α- or γ- zeins can be seen, suggesting that the bZIP TILLING mutant has not affected zein deposition.

Conclusion The Opaque-2 transcription factor is responsible for regulating zein storage protein

deposition into maize endosperm tissue to provide required energy and nutrients for the

growing plant. By knocking-out this protein, zein deposition in the kernel is reduced,

increasing the relative concentration of lysine, an amino acid essential for animal growth that

is limiting in typical maize kernels. However, opaque-2 knock-out plants have pleiotrophic

phenotypes that negate their efficacy for commercial grain production. The maize bZIP

transcription factor ZmbZIP1 was identified in a transcript profiling experiment as being

differentially expressed when comparing the B45 and B45opaque-2 genotypes. This study

strove to further characterize ZmbZIP1 to determine whether it could be a more specific

regulator of zein-deposition in the maize kernel. ZmbZIP1 was mapped to chromosome 9, 60

contig 380 in the maize genome, and a rice ortholog (RISBZ5) was identified. Transcript

profiles of ZmbZIP1 in eight tissues in two lines, B73 and B46o2 were created, showing that

ZmbZIP1 mRNA is present in all tissues to differing extents. This data was echoed by a

virtual northern; between these two experiments, it seems that ZmbZIP1 is present in all

maize tissues. To more specifically look at ZmbZIP1 endosperm activity, mRNA levels were

directly compared B46 and B46o2. It was found that while transcript expression was higher

in B46, it was not increased at a statistically-significant level. However, this does not negate

the potential biological relevance of ZmbZIP1, as small amounts of change in mRNA can

produce large phenotypic differences.

Zein levels in a ZmbZIP1 knock-out mutant were shown to be identical to those in a

wild-type plant. This may be due to one of two factors: first, the knock-out mutant was

created using EMS mutagenesis, which creates only single basepair mutations which may

have not been enough to completely cripple the protein. Second, functional redundancy may

exist between ZmbZIP1 and another protein. To further explore which hypothesis is correct,

a more severe knock-out mutant should be created, perhaps using a t-DNA insertion.

Another future direction for research would involve making an opaque-2-zmbZIP1 double

mutant. This may create a phenotype that is severe enough to be detected, and differences

between the opaque2 and zmbZIP1 phenotypes could be elucidated.

References FOULKES, N. S., E. BORRELLI and P. SASSONE-CORSI, 1991 CREM gene: Use of alternative DNA-binding domains generates multiple antagonists of cAMP-induced transcription. Cell 64: 739-749. FU, Y., T.-J. WEN, Y. I. RONIN, H. D. CHEN, L. GUO et al., 2006 Genetic dissection of intermated recombinant inbred lines using a new genetic map of maize. Genetics 174: 1671-1683. 61

GARDINER, J., S. SCHROEDER, M. L. POLACCO, H. SANCHEZ-VILLEDA, Z. FANG et al., 2004 Anchoring 9,371 Maize Expressed Sequence Tagged Unigenes to the Bacterial Artificial Chromosome Contig Map by Two-Dimensional Overgo Hybridization. Plant Physiol. 134: 1317-1326. JIA, H., 2004 Expression analysis of maize opaque2 mutants and Glu-1Dx5 transgenic corn, pp. 147 in Interdepartmental Genetics. Iowa State University, Ames, IA. LEFEVRE, A., L. CONSOLI, S. A. GAZIOLA, A. P. PELLEGRINO, R. A. AZEVEDO et al., 2002 Dissecting the opaque-2 regulatory network using transcriptome and proteome approaches along with enzyme activity measurements. Scientia Agricola 59: 407-414. LIVAK, K. J., and T. D. SCHMITTGEN, 2001 Analysis of relative gene expression data using real-time quantitative PCR and the 2(−Delta Delta C(T)) method. Methods 25: 402- 408. MEAGHER, R. B., 1991 Divergence and differential expression of actin gene families in higher plants. Internal Review of Cytology 125: 663-672. MONIZ DE SA, M., and G. DROUIN, 1996 Phylogeny and substitution rates of angiosperm actin genes. Molecular Biology and Evolution 13: 1198-1212. ONODERA, Y., A. SUZUKI, C.-Y. WU, H. WASHIDA and F. TAKAIWA, 2001 A Rice Functional Transcriptional Activator, RISBZ1, Responsible for Endosperm-specific Expression of Storage Protein Genes through GCN4 Motif. J. Biol. Chem. 276: 14139-14152. PABO, C. O., and R. T. SAUER, 1992 Transcription factors: Structural familes and principles of DNA recognition. Annu. Rev. Biochem. 61: 1053-1095. PYSH, L. D., M. J. AUKERMAN and R. J. SCHMIDT, 1993 OHP1: a maize basic domain/leucine zipper protein that interacts with opaque2. Plant Cell 5: 227-236. SAMBROOK, J., and D. W. RUSSELL, 2001 Molecular Cloning: A Laboratory Manual, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. SCHMIDT, R. J., F. A. BURR, M. J. AUKERMAN and B. BURR, 1990 Maize regulatory gene opaque-2 encodes a protein with a "leucine-zipper" motif that binds to zein DNA. Proc Natl Acad Sci 87: 46-50. SCHMIDT, R. J., M. KETUDAT, M. J. AUKERMAN and G. HOSCHEK, 1992 Opaque-2 Is a Transcriptional Activator That Recognizes a Specific Target Site in 22-kD Zein Genes. Plant Cell 4: 689-700. TILL, B. J., S. H. REYNOLDS, C. F. WEIL, N. M. SPRINGER, C. BURTNER et al., 2004 Discovery of Induced Point Mutations in Maize Genes by TILLING. BMC Plant Biology 4: 12. 62

CHAPTER 3: MOLECULAR MECHANISM OF METHIONINE DIFFERENTIATION IN THE BS11 HIGH AND LOW MET MAIZE LINES

Abstract Since maize is a primary foodstuff for animals and humans, its amino acid balance is

important for proper nutrition. Methionine, an essential amino acid and a primary source of

sulfur, is lacking in the maize endosperm. To attempt to determine biologically relevant

ways to increase methionine content in maize, we set out to molecularly characterize the

mechanism by which methionine levels are increased in a BS11 high met (HM) population over its low met (LM) counterpart. Two candidate genes were characterized: the 10kDa

delta-zein (Dzs10), a methionine-rich storage protein in endosperm, and cysteine synthase

(Cys2), a key regulator in cysteine/methionine biosynthesis. mRNA expression levels and haplotypes were determined for each population, and were correlated to the average methionine content. At the mRNA level, Dzs10 was found to be differentially expressed in maize endosperm, while Cys2 is likely not differentially expressed. Further, a correlation of

0.4535 was found between Dzs10 mRNA and kernel methionine levels, suggesting part of the increase in kernel methionine deposition may be due to a change in Dzs10 expression. In

Cys2, a strong relationship has been noted between the level of kernel methionine deposition and a G/A SNP present in the gene, suggesting that different segregations of the Cys2 alleles

may also alter the kernel methionine levels.

Materials and Methods

Sample Collection BS11HM (high methionine) and BS11LM (low methionine) maize populations were created by selection of the 50 ears with the highest or lowest methionine content from an 63

original BS11 population of about 250 F1 ears (cycle 0). Recurrent selection for grain

methionine content was then done on each resulting population, each year choosing the 5 out of 50 ears with the highest or lowest methionine content for the next generation for a total of six generations (cycle 6). Leaf tissue was collected from seven BS11HM and seven

BS11LM cycle six plants grown at the Iowa State University Agronomy Farm (Boone, IA) during summer 2007. Each plant was then pollinated by a chain sib mating design, and endosperm tissue was collected from the resulting kernels at 14 dap. All tissue was frozen in liquid nitrogen and stored at -80C for further processing.

RNA Extraction Total RNA was isolated from approximately 100 mg of each leaf tissue sample

collected using the Ambion ToTALLY RNA kit (Ambion Inc., Austin TX) with no

modification. Any DNA remaining in the sample after isolation was completely removed

using the Ambion DNA-free™ kit (Ambion Inc., Austin TX) with no modification.

Polyadenylated mRNA was isolated from approximately 100g of each endosperm tissue

collected using the Promega PolyATtract 1000 kit (Promega Corp., Madison, WI), also

without modification. Isolated RNA was quantified using a NanoDrop ND-1000

Spectrophotometer and stored at -80oC until amplification reactions could be performed.

DNA Extraction DNA was isolated from 35-100 mg of leaf tissue collected above. Samples from each plant were prepared using the MasterPure™ Plant Leaf DNA Purification Kit (Epicenter

Biotechnologies, Madison, WI) with no modification and were quantified using a NanoDrop

ND-1000 Spectrophotometer (NanoDrop Technologies, Wilmington, DE). Isolated DNA

was stored at -20oC until amplification reactions could be performed. 64

Differential Expression of Cys2 and Dzs10 via Quantitative Real-Time RT-PCR Real-time QRT-PCR experiments were performed with either 1ng of endosperm mRNA or 200 ng of total leaf RNA as an initial template. Primers were designed based upon the maize Cys2 (NCBI: X85803), Dzs10 (NCBI: AF371266), or MAc1 (Actin) (J01238) cDNA sequence; an intron was spanned in both the Cys2 and MAc1 sequences in order to distinguish a DNA from RNA product, but this was not possible in the Dzs10 gene. Primer sequences are as follows: Cys2: F: AGG CTC TTG CTC TGA AAG AAG GGT, R: CGA

CAA CAA ATA GCT TTC CGG CGT; Dzs10: F: AGA CGA TGC CAG TGA TGA TG, R:

TGC AGC ATA ATC TGC GAG AC; ActinF2: CCT GAA GAT CAC CCT GTG CT, Actin

R: TGT ACG TGG CCT CAT GGA C. Stratagene Brilliant II QRT-PCR kit was used to assemble all reactions in 25ul volumes as described in the manual with 200nM F and R primers. The fluorophore CyBr Green is included in these reactions to quantify the amplified product. To create an external standard RNA, the seven BS11 HM RNAs were pooled and used to create 6- 4x serial dilutions of 200-0.39 ng total RNA (leaf) or 1ng-1pg mRNA

(endosperm). These dilutions were amplified with the actin primers to create a standard curve.

Cycling protocols were performed in a Stratagene MX3000P (Stratagene, La Jolla,

CA) thermocycler and consisted of a first strand synthesis of 45°C for 45 min, followed by

95°C for 10 min, followed by 40 cycles of 95°C for 30 sec, 59°C for 1 min, 72°C for 45 sec, followed by a dissociation curve, that involved taking a fluorescence reading at every degree between 55°C and 95°C to ensure only one product of the correct annealing temperature was being amplified. Using the average fluorescence of cycles 4-14, the Stratagene analysis software established a fluorescence threshold below which fluorescence levels were not 65

statistically different than the background fluorescence levels. The cycle at which any

reaction crosses this threshold level is referred to as its Ct (threshold cycle) value. All

reactions were run in triplicate to assess consistency, and each sample was normalized

against actin (also in triplicate) to ensure that any resulting differential expression was not due to varying amounts of initial RNA template. The delta-delta Ct value was calculated using actin as the internal control and the actin amplified from the pooled RNA sample as the calibrator (LIVAK and SCHMITTGEN 2001). Statistical differences between populations was

calculated by first determining the copy number of each individual plant using the formula

-∆Ct -∆Ct Copy number = 2 exp – 2 actin

then finding the population average and standard deviation, and calculating significance

using the Student’s T-test with a threshold of p< 0.05.

Cys2 and Dzs10 Sequencing and Allele Comparison 150 ng of DNA was used as an initial PCR template and amplification was done

using the Cys2 and Dzs10 primer pairs listed above. Reactions were performed in a 25 uL

volume containing 5x Green GoTaq Reaction buffer®, with 1.5 mM final concentration

MgCl2 (Promega, Madison, WI), 1.0 mM additional MgCl2, 0.2 mM each dNTP, 0.5 µM

each 5’ and 3’ primer, and 0.625 U GoTaq® DNA Polymerase (Promega, Madison, WI).

Amplification was carried out under the following conditions: an initial amplification at 95C

for 2 min, followed by 35 cycles of 95°C for 30 sec, 55°C for 30 sec, 72°C for 1 min, and a

final extension of 72°C for 10 min. Ten uL of each reaction were electrophoresed on a 1%

agarose gel and the resulting bands were excised and purified using the Montage DNA Gel

Extraction kit (Millipore Corp, Bedford, MA). Two uL of the resulting product were cloned

into the TOPO-TA 2.1® cloning vector (Invitrogen, Carlsbad, CA) and transformed into 66

Stratagene XL-1 Blue competent E. coli cells (Stratagene, La Jolla, CA), each as described by the manufacturer. The cells were then grown for 16 hours on LB plates containing 50 µM

Ampicillin on which 10µl of 100mM IPTG and 40ul of 20mM X-Gal had been spread. Four single colonies were selected for each sample, and plasmid DNA was purified using the

QIAprep Spin Miniprep Kit ™ (Qiagen, Valencia, CA). Thus, each sequence should represent one of the two possible haplotypes contained by the plant. DNA was then sequenced at the Iowa State University DNA Facility using the Applied Biosystems 3730xl

DNA Analyzer (Applied Biosystems, Foster City, CA) and the resulting sequences were aligned using the BioEdit Sequence Alignment Editor (HALL 1999). The ClustalW alignment application with optimal settings was used to determine the haplotype present in each plant.

Amino Acid Assay to Determine Kernel Methionine Levels Methionine was quantified using a high-throughput microbiological method described by Scott et al (SCOTT et al. 2004). Ears were harvested from the fourteen BS11HM and

BS11LM plants from which DNA and RNA had been isolated. Forty kernels from each ear were ground using a Stein grinder in order to create a sample representative of each ear. 10 mg of the ground grain was then deposited into a randomly assigned well of a 96-well plate; each sample was replicated five times in the plate for a total of 75 filled wells. Two plates were filled in this manner. The powdered grain powder was then hydrolyzed and protein was extracted using 100 uL of 50mM KCl adjusted to pH 2.0 with HCl plus 0.2 mg of pepsin.

Each plate was then incubated with shaking at 37oC for 16 hours. The plate was then centrifuged at 3000 x g for 20 minutes. 5 uL of supernatant was then transferred to the corresponding well of a second plate, and ten of the remaining wells were used to 67

accommodate 5 uL standards of commercially obtained methionine (Sigma, St. Louis, MO)

in concentrations ranging from 0.1 to 0.8 mM. This plate was inoculated with the P4X E.

coli strain, which is auxotrophic for methionine (JACOB and WOLLMAN 1961) in 100 uL of

M9 minimal media. This plate was then shaken at 37C for 16 hours. Following incubation,

the optical density at 595 nm was read using a microplate reader to give a value that directly reflects that level of amino acid per mass of tissue. The concentration of methionine in each

analysis was calculated by linear regression onto the line created by the standard dilutions.

The predicted value of each sample was calculated using a linear ANOVA model with plate

as a fixed effect, while differences within and among groups were characterized using a one-

way ANOVA with a threshold value of p>0.05. Where variation was significant, groupings

were assigned using the pairwise Students T test with p>0.05.

Results and Discussion

The BS11HM Population Contains Significantly More Methionine than BS11LM At the third cycle of selection, the BS11HM and LM populations were subjected to

direct comparison in order to ensure that the proper directional selection for methionine was

being achieved. The resulting methionine values (0.21 g/100g sample in the BS11HM

population and 0.18 g/100 g sample in the BS11LM population) showed that there was a

14.3% increase in methionine content in the high methionine population; however, that

increase was not statistically significant (results not shown). In the current cycle (cycle 7),

methionine levels have continued to differentiate: BS11HM contained 0.36g/100g sample,

while BS11LM contained 0.24g/100g sample (Table 1), resulting in a 32.5% increase in

methionine production in the HM line. This results in significantly higher values of

methionine content being found on average in the BS11HM population when compared to 68

the BS11LM population, with average relative methionine contents of 0.085 and 0.078 (Fig.

1), respectively (p=0.0169), although the individual plant values did not separate completely

into statistically significant categories (Table 1).

Individual Groupinga g Met/100g sample LM3 A 0.463 HM4 A B 0.419 HM3 A B C 0.412 HM5 A B C 0.358 HM2 A B C 0.344 HM6 A B C 0.338 HM7 A B C 0.332 LM6 A B C D 0.312 HM1 BC D E 0.283 LM4 BC D E 0.256 LM7 CD E 0.245 LM1 D E 0.158 LM2 E 0.138 LM5 E 0.127 Mean HM 0.360 Mean LM 0.243

Table 1. Average methionine content of BS11HM and BS11LM individuals. a Levels not connected by the same letter are statistically different by a Student’s t-test (p< 0.05). 69

Figure 1. BS11HM population has significantly higher levels of methionine than BS11LM population. The BS11HM population has an average relative methionine level of 0.85, while the BS11LM population has an average relative methionine level of 0.78. These values are significantly different at the level of p=0.0168.

Correlation Between Population Methionine Levels and Cys2 or Dzs10 mRNA Levels In order to elucidate the mechanism conferring differing levels of methionine to the

BS11HM and BS11LM maize populations, a candidate gene approach was taken. Cysteine synthase (Cys2) and the 10kDa delta-zein (Dzs10) transcripts were chosen as candidates for

investigation via quantitative real-time RT-PCR analysis because of their respective roles as

a control point in cysteine (methionine) biosynthesis (DROUX et al. 1998) and the main

methionine sink in the kernel (CHUI and FALCO 1995), respectively. Seven plants of each

genotype were chosen as a representative sample of the population and mRNA levels were

compared on the basis of population averages.

Maize cysteine synthase activity is found primarily in leaf mesophyll cells and a

small amount is present in developing endosperm (BURNELL 1984; HELDT 2005), so mRNA

levels were compared in both leaf tissue and endosperm tissue. In neither case did the

expression levels between the overall populations produce over a 2-fold difference (1.10 and 70

-∆∆Ct 1.01 cycle difference, respectively) (Fig. 2), nor were the 2 values (LIVAK and

SCHMITTGEN 2001) statistically different (Table 2). Furthermore, no correlation was found

between Cys2 mRNA levels and kernel methionine levels, with correlation coefficients of -

0.011 (p=0.969) for leaf tissue and 0.102 (p=0.729) for endosperm tissue (Fig. 3). Together, these results suggest that the Cys2 mRNA level is not likely to be responsible for the differences in methionine levels between the populations.

Zein mRNAs are expressed solely in the maize endosperm (BOSTON and LARKINS

1987), so the 10 kDa delta-zein expression levels were measured in that tissue only.

Quantitative real-time RT-PCR results showed greater than 2-fold difference between the

levels of Dzs10 expression in the BS11 HM and LM populations, with the BS11HM plants

producing 2.23 times as much mRNA as the LM population on average (Fig. 2). This

difference between populations is also statistically significant, with a p-value of 0.020 (Table

2). The correlation coefficient between methionine content and 10kDa delta zein content

mRNA levels was 0.4535 (Fig. 3) which was not significant at a 95% confidence level

(p=0.1196), which is not surprising given the large variation in methionine content in

individuals within the two populations.

A two-fold expression threshold is often reported for QRT-PCR data analysis in lieu

of or along with statistical methods. This level, however, is simply a threshold chosen when

it was not possible to include replication in the experiment to allow for statistical results to be

reported. This threshold is simply a guideline to determine which candidate genes may be

worth further study. It may be the case that a two-fold increase in mRNA expression is not a

biologically relevant level, for instance if mRNA levels do not correspond directly to protein

levels, if there is an excess of that protein relative to another protein with which it complexes, 71

or if the protein is has low biological activity. On the other hand, for some proteins, for

instance enzymes that are very sensitive to concentration, have high activity, or have high turnover rates, a two-fold expression change may be far above what is necessary to create a biologically significant response.

Expression fold change between BS11HM and BS11LM maize lines

2.50

2.00

1.50

1.00 Fold change (2 delta Ct) delta (2 change Fold

0.50

0.00 Cys2 Leaf HM Cys2 Leaf LM Cys2 Cys2 Dzs10 Dzs10 Endosperm HM Endosperm LM Endosperm HM Endosperm LM

Figure 2. mRNA expression fold change between BS11HM and BS11LM maize populations. The threshold for differential expression was set at a 2-fold change, which is delineated by the pink, dashed line. As such, there is no differential expression of cysteine synthase in either the leaves or the endosperm; however, in the 10kDa delta-zein, a two fold level of differential expression was seen.

Mean Copy Number Fold Change p- Gene/Tissue Population (2-∆Ct) (n=7)a Std. Dev. (2-∆∆Ct)b value Cys2 Leaf BS11 HM 4.16 x 10-6 1.49 x 10-6 1.10 0.606 BS11 LM 3.79 x 10-6 9.74 x 10-7 1.00 Cys2 BS11 HM 8.96 x 10-7 3.88 x 10-7 1.01 0.181 Endosperm BS11 LM 1.16 x 10-7 3.88 x 10-7 1.00 Dzs10 BS11 HM 5.26 x 10-6 1.26 x 10-6 2.23 0.020 Endosperm BS11 LM 4.74 x 10-7 1.26 x 10-6 1.00 * Table 2. Analysis of 2-∆Ct in the BS11 HM vs. BS11 LM population. An asterisk(*) denotes statistically significant differences in mRNA level between BS11HM and LM populations as determined by a Student’s t-test. These results are comprised of seven biological replicates each of which was technically replicated in triplicate. a Mean copy number is normalized to actin content. b Fold change is a comparative value between the two genotypes; therefore, in order to express this value in both populations, the fold change of the BS11 LM population has been normalized to 1.00. 72

Corr. = -0.011

Corr. = 0.102

Corr. = 0.454

Figure 3. Correlations between mRNA levels and methionine content. Cysteine synthase mRNA levels have little to no correlation with kernel methionine content. However, the 10kDa delta-zein has a comparatively large correlation with kernel methionine content. The red dots denote individuals from the HM population and the blue dots denote individuals from the LM population. 73

Correlation Between Kernel Methionine and Cys2 or Dzs10 Haplotypes One potential explanation for differences in methionine levels between BS11HM and

BS11LM is a difference in haplotypes between the two populations. In order to be

considered a separate haplotype, the sequence must contain two or more SNPs from the

consensus sequence of all BS11 alleles sequenced. When only a single nucleotide change

from the consensus was present, this was assumed to be a technical error (either from PCR

amplification or sequencing) unless it was seen in a second individual in the population.

When a representative 278 bp of the Cys2 gene was sequenced, 9 haplotypes were found in

total. These haplotypes segregate almost completely between the HM and LM populations,

with haplotype 2 and 3 found only in the HM population and haplotypes 4-9 found only in

the LM population. However, only two, haplotypes 1 and 4, were found with regularity.

Haplotype 1 was the most common and was the only haplotype present in both populations,

occurring on at least one chromosome in all BS11 HM plants, as well as at least once in LM3

and LM6 (Table 3). Interestingly, these two LM plants have the two highest values for

methionine content in the LM population (0.463 and 0.312 g Met/100g sample, respectively;

Table 1). Haplotype 4 is present in five of the seven BS11 LM plants; it is notably absent in both LM3 and LM6 (Table 3). All other haplotypes were also seen only in either the HM or

LM populations; haplotypes 2 and 3 are found in the HM population, and haplotypes 5-9 are

found in individuals in the LM population.

The sole difference between haplotypes 1 and 4 is a G to A (respectively) single

nucleotide polymorphism (SNP) at basepair 122 of the sequenced region. While it is likely

that this SNP is not the causative mutation, but instead a linked marker, it is interesting to

note that this basepair is found in an intron close to an intron/exon junction in Cys2 (Table 3), 74

and G/A basepairs are often seen in these border regions, allowing for the remote possibility

of an alteration of the splice junction. More likely, though, is the possibility that the

causative mutation is found outside the sequenced region, which makes up only a small

portion of the Cys2 genomic sequence (21.2% of 1314 bp) and does not include any of the

regulatory domains that may be present in the promoter or terminator regions. This SNP has

the potential to be useful in marker assisted-selection; however, more individuals must be

first tested in order to ensure that the distribution of these markers is correlated with

methionine content in these populations.

When a representative 157 nucleotides of the Dzs10 gene was sequenced, there was a

sharp contrast to the Cys2 situation. There appears to be a highly conserved sequence in

Dzs10, with one major haplotype (as defined above) that encompasses all individuals except

LM1, which contains an 11 nucleotide deletion at basepairs 20-31 of the sequenced region

but did not have a detectable affect on methionine deposition, as LM1 has an intermediate level of methionine deposition at 0.158 g Met/100 g sample (Table 2). As such, there was no

Dzs10 allele favored in either HM or LM genotypes (summarized in Table 4). However,

haplotypes may be more evident in other parts of the gene, since the area sequenced only

represents 25.3% of the 10kDa delta zein cDNA (157 of 620 nucleotides) and includes none

of the promoter region, which is extremely important for gene regulation. The mRNA level

differences seen above could be due to a trans-acting modifier gene rather than something

present in cis. This type of regulation was seen in BSSS53, where a change in expression in

the regulatory gene Zpr10|(22) was shown to be the factor increasing levels of the 10kDa delta-zein, rather than any change in the 10kDa delta zein sequence itself (BENNER et al.

1989). Such a trans-acting factor could be regulated by the availability of reduced sulfur, 75 which could in turn be controlled by the specific alleles of sulfur assimilation genes such as

Cys2.

exon exon exon exon exon exon exon intron intron intron intron intron exon Individual 14nt 35 42 67 71 83 118 122 129 206 212 218 261 Haplotype HM1 G T G T T T T G A T G G G 1 G T G T C T T A A C G G G 2 HM2 G T G T C T T A A C G G G 2 G T G T T T T G A T G G G 1 HM3 G T G T C T T A A T G G G 3 G T G T T T T G A T G G G 1 HM4 G T G T T T T G A T G G G 1 G T G T T T T G A T G G G 1 HM5 G T G T T T T G A T G G G 1 G T G T C T T A A T G G G 3 HM6 G T G T T T T G A T G G G 1 G T A T T T T G A T G G G (1) HM7 G T G T T T T G A T G G G 1 G T G T T T T G A T G G - (1) LM1 G T G T T T T A A T G G G 4 G T G T T C T A A T G G G 5 LM2 G T G T T T T G A T G G G 1 G T G T T T T A A T G G G 4 LM3 G T G T T T T G A T G G G 1 G T G T T T T A A T A G G 6 LM4 G T G T T T T A A T G G G 4 - T G T T T T A T T G G G 7 LM5 G T G T T T T A A T G G G 4 G A G T T T T A A T G G G 8 LM6 G T G T T T T G A T G G G 1 G T G T T T G G A T G G G (1) LM7 G T G T T T T A A T G G G 4 G T G T T T T A A T G A G 9 CONTIG G T G T T T T G A T G G G 1

Table 3: Haplotypes present in the Cys2 gene. There are 9 haplotypes present in the 14 individuals sequenced, a consensus haplotype (hap 1) is present in 10 of the 14 sequences, two additional haplotypes (hap 2, 3) are present solely in the high methionine genotype, and five additional haplotypes (hap 4-9) are present exclusively in the low methionine genotype. The highlighted polymorphisms only appeared in one instance and were excluded from the haplotype analysis. The consensus nucleotide was assumed to be present at these due to the possibility that they were artifacts of technical errors in the sequencing procedure. 76

Individual 21 nt 20-31nt 57 nt 102 nt 105 nt Haplotype HM1δ C - A T C 1 HM2δ C - A T C 1 C - A C C (1) HM3δ C - A T C 1 C - G T C (1) HM4δ C - A T C 1 HM5δ C - A T C 1 HM6δ C - A T C 1 HM7δ C - A T C 1 LM1δ - deletion A T C 2 C - A T C 1 LM2δ C - A T C 1 T - A T C (1) LM3δ C - A T C 1 LM4δ C - A T C 1 LM5δ C - A T C 1 C - A T T (1) LM6δ C - A T C 1 LM7δ C - A T C 1 CONTIG C - A T C 1

Table 4. Haplotypes present in the Dzs10 gene. There are 2 haplotypes present in the 14 individuals sequenced: a consensus haplotype is present in 13 of the 14 individuals and a second haplotype is found only in LM1, and is due to an 11 bp deletion of nucleotides 20-31. Bold letters represent the more prevalent nucleotide present in each SNP. The highlighted polymorphisms were excluded from the haplotype analysis due to the possibility that they were artifacts of technical errors in the sequencing procedure.

Conclusion Methionine is a nutritionally-limiting amino acid in maize kernels. It is an essential amino acid in animals that supplies an endogenous sulfur pool and is necessary for protein formation and hormone synthesis (BAKER 1986). Because of the essential nature of

methionine in the diet, maize-dominated feedstocks must be supplemented with exogenous

methionine at an expense of $100 million/ year (ISMANDE 2001). In this study, two BS11

populations, one with high kernel methionine content (HM) and one with low methionine 77 content (LM) were compared in order to determine whether two candidate genes, the 10kDa delta-zein (Dzs10) and cysteine synthase (Cys2) were, in part, responsible for this difference.

It was found that Dzs10 mRNA was differentially expressed between the two populations, and was correlated with methionine deposition with a coefficient of 0.454. Further, two haplotypes were found in Cys2 that segregate with kernel methionine deposition. The SNPs found in these haplotypes have the potential to be used in breeding for increased directional selection for increased methionine production in a marker-assisted selection program; however, they must first be tested on a larger population scale. These results suggest that on a biological level, sulfur is controlled at both the source and sink levels. At the source, mRNA levels are constant, so different alleles probably encode enzymes with different activities, while at the sink, mRNA levels respond to the availability of sulfur.

This study suggests that both Dzs10 and Cys2 are involved in increasing the levels of kernel methionine deposition, however, they are only two candidate genes; it is likely that other factors are at work as well. In the future, it would be useful to use these BS11 populations to perform either a microarray-type experiment to determine other genes likely involved in methionine deposition, or to choose a larger number of candidate genes from pathways known to be associated with methionine biosynthesis: sulfur uptake and translocation channels, more members of the sulfate assimilation and methionine biosynthesis pathway, and factors related to methionine deposition. 78

References BAKER, D. H., 1986 Utilization of isomers and analogs of amino acids and other sulfur- containing compounds. Prog Food Nutr Sci 10: 133-178. BENNER, M. S., R. L. PHILLIPS, J. A. KIRIHARA and J. W. MESSING, 1989 Genetic analysis of methionine-rich storage protein accumulation in maize. Theoretical and Applied Genetics 78: 761-767. BOSTON, R. S., and B. A. LARKINS, 1987 Control of maize zein gene expression. Genet Eng Princ Methods 9: 61-74. BURNELL, J. N., 1984 Sulfate assimilation in C4 plants. Plant Physiology 75: 873-875. CHUI, C. F., and S. C. FALCO, 1995 A new methionine-rich seed storage protein from maize. Plant physiol 107: 291. DROUX, M., M. L. RUFFET, R. DOUCE and D. JOB, 1998 Interactions between serine acetyltransferase and O-acetylserine(thiol)lyase in higher plants: structural and kinetic properties of the free and bound enzymes. European Journal of Biochemistry 255: 235-245. HALL, T. A., 1999 BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl. Acids. Symp. Ser. 41: 95-98. HELDT, H. W., 2005 Plant Biochemistry. Elsevier Academic Press, Burlington, MA. ISMANDE, J., 2001 Selection of soybean mutants with increased concentrations of seed methionine and cysteine. Crop Science 41: 510-515. JACOB, F., and E. WOLLMAN, 1961 Sexuality and the genetics of bacteria. Academic Press. LIVAK, K. J., and T. D. SCHMITTGEN, 2001 Analysis of relative gene expression data using real-time quantitative PCR and the 2(−Delta Delta C(T)) method. Methods 25: 402- 408. SCOTT, M. P., S. BHATNAGAR and J. BETRAN, 2004 Tryptophan and methionine levels in quality protein maize breeding germplasm. Maydica 49: 303-311. 79

GENERAL CONCLUSIONS

Our results suggest that while it may be possible to increase lysine content in maize

kernels by altering ZmbZIP1, current data is inconclusive as to whether or not modification

would be useful. It was found that while ZmbZIP1 expression was increased in B46

compared to B46o2, it was not statistically differentiated. However, this does not negate the

potential biological relevance of ZmbZIP1, as small amounts of change in gene expression

can produce large phenotypic differences. Zein levels in a ZmbZIP1 knock-out mutant were

shown to be identical to those in a wild-type plant. This may be due either a lack of

mutational severity produced through EMS mutagenesis or because of functional redundancy

that may exist between ZmbZIP1 and another protein. Further research should include a

more severe knock-out mutant, perhaps created using a t-DNA insertion, as well as making

an o2:zmbZIP1 double mutant. This may create a phenotype that is severe enough to be

detected, and differences between the opaque2 and zmbZIP1 phenotypes could be elucidated.

Through comparison of BS11 HM and LM populations, it was concluded that both

the 10kDa delta-zein (Dzs10) and cysteine synthase (Cys2) were, in part, responsible for

methionine deposition into maize endosperm. Results suggest that Dzs10 may act through

differential mRNA expression in the two populations, whereas Cys2 may have different

alleles that have different levels of activity for sulfur incorporation into cysteine. These differences may be useful for increased methionine production in a marker-assisted selection breeding program. These results suggest that on a biological level, sulfur is controlled at

both the source and sink levels. However, it is unlikely that Dzs10 and Cys2 are the only genes involved in this process, suggesting the need for larger scale investigation in the future.