<%BANNER%>

Putative Role for RNA Splicing in Maize Endosperm-Embryo Developmental Interactions Is Revealed by the Rough Endosperm 3...

Permanent Link: http://ufdc.ufl.edu/UFE0022022/00001

Material Information

Title: Putative Role for RNA Splicing in Maize Endosperm-Embryo Developmental Interactions Is Revealed by the Rough Endosperm 3 (rgh3) Seed Mutant
Physical Description: 1 online resource (84 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: Plant Molecular and Cellular Biology -- Dissertations, Academic -- UF
Genre: Plant Molecular and Cellular Biology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Endosperm-embryo interactions are an important but poorly understood aspect of seed development. The Rough endosperm3 (Rgh3) locus is involved in these interactions at a developmental level. The rgh3 seed mosaics marked with the pr1 anthocyanin gene indicate that Rgh3 is required in the endosperm for the normal development of the embryo. The Rgh3 locus also has an autonomous function in the embryo and is required for seedling viability. The Rgh3 locus maps to the long arm of chromosome 5. Complementation tests with other seed mutants mapped to the same region indicated that Rgh3 is a novel locus. We identified a tightly-linked transposon-tag from the rgh3-70 allele. This transposon insertion disrupts a predicted splicing factor with a U2 snRNP auxiliary factor homology motif (UHM) that is most similar to the U2AF35-Related Protein (URP). The UHMs are RNA recognition motif (RRM)-like domains that function in protein-protein interactions. Analysis of Rgh3/ZmURP cDNAs and genomic sequences suggests ZmURP produces alternatively spliced transcripts. Maize ESTs indicated that ZmURP is expressed in endosperm tissue. Consistent with this expression pattern, the rgh3 phenotype is characterized by an overproliferation of the aleurone cells and aberrant development of the basal endosperm transfer cell layer. An analysis with endosperm cell type specific markers in mutant rgh3 seeds suggests that the basal endosperm transfer cell defect occurs after cell type specification. Tissue culture of rgh3 endosperm cells showed robust growth and formed homozygous mutant callus. These data indicate that Rgh3 is not an essential gene for gene expression. Instead the data presented argue that the putative RNA splicing function of Rgh3 has an important role in cell differentiation, which then impacts endosperm-embryo developmental interactions.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Settles, Andrew M.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-05-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022022:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022022/00001

Material Information

Title: Putative Role for RNA Splicing in Maize Endosperm-Embryo Developmental Interactions Is Revealed by the Rough Endosperm 3 (rgh3) Seed Mutant
Physical Description: 1 online resource (84 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: Plant Molecular and Cellular Biology -- Dissertations, Academic -- UF
Genre: Plant Molecular and Cellular Biology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Endosperm-embryo interactions are an important but poorly understood aspect of seed development. The Rough endosperm3 (Rgh3) locus is involved in these interactions at a developmental level. The rgh3 seed mosaics marked with the pr1 anthocyanin gene indicate that Rgh3 is required in the endosperm for the normal development of the embryo. The Rgh3 locus also has an autonomous function in the embryo and is required for seedling viability. The Rgh3 locus maps to the long arm of chromosome 5. Complementation tests with other seed mutants mapped to the same region indicated that Rgh3 is a novel locus. We identified a tightly-linked transposon-tag from the rgh3-70 allele. This transposon insertion disrupts a predicted splicing factor with a U2 snRNP auxiliary factor homology motif (UHM) that is most similar to the U2AF35-Related Protein (URP). The UHMs are RNA recognition motif (RRM)-like domains that function in protein-protein interactions. Analysis of Rgh3/ZmURP cDNAs and genomic sequences suggests ZmURP produces alternatively spliced transcripts. Maize ESTs indicated that ZmURP is expressed in endosperm tissue. Consistent with this expression pattern, the rgh3 phenotype is characterized by an overproliferation of the aleurone cells and aberrant development of the basal endosperm transfer cell layer. An analysis with endosperm cell type specific markers in mutant rgh3 seeds suggests that the basal endosperm transfer cell defect occurs after cell type specification. Tissue culture of rgh3 endosperm cells showed robust growth and formed homozygous mutant callus. These data indicate that Rgh3 is not an essential gene for gene expression. Instead the data presented argue that the putative RNA splicing function of Rgh3 has an important role in cell differentiation, which then impacts endosperm-embryo developmental interactions.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Settles, Andrew M.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-05-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022022:00001


This item has the following downloads:


Full Text

PAGE 1

1 PUTATIVE ROLE FOR RNA SPLICI NG IN MAIZE ENDOSPERM-EMBRYO DEVELOPMENTAL INTERACTIONS IS REVEALED BY THE rough endosperm 3 ( rgh3) SEED MUTANT By DIEGO S. FAJARDO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Diego S. Fajardo

PAGE 3

3 To my wife Silvana

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank m y wife, Silvana, for her support, pa tience and love. I dont think I would have been able to go through this without her. I thank my major advisor, Dr. Andrew M. Settles, for his advice, patience and for helpi ng me become a better sc ientist. I thank my committee (Dr. Curt Hannah, Dr Kenneth Cline, Dr. Jon Stewart) for their advice and suggestions. I thank our collaborators Dr. Gre gorio Hueros, Elisa Gomez and Joaquin Royo at the Universidad de Alcala in Spain. I would like to thank the people involve in The Maize Endosperm Development Project especially; Dr. Don McCarty, Dr. Curt Hannah, Dr. Karen Koch, Dr. Mark Settles, Dr. Phil Becraft, Dr. Brian Larkins, Dr. Jo Messing, Sue Latshaw, Dr. Masaharu Suzuki, Dr. Bao-Cai Tan, John Baier. I extend my gratitude to the entire PMCB faculty and staff who have taught me and helped me through my years at UF. I thank the members of the Settles laborator y for their help and friendship, especially to Dr. Chi-Wah Tseung, Dr. Tim Porch, Ewa Wroclawska, Federico Martin, Dr. Gertraud Spielbauer and Dr. Romain Fouquet. I thank my fellow students, Dr. Li-Fen Huang, Dr. Rocio Diaz de la Garza, Dr. John Mayfield for their friendship and for all the great memories. Finally, I thank my family (my parents Mart ha and Adrian, my grandparents Graciela and Alberto, my sister Nuria and my uncle Juan) for the support they have given me through the years.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES.........................................................................................................................8 ABSTRACT.....................................................................................................................................9 CHAP TER 1 INTRODUCTION..................................................................................................................11 The Maize Seed as a Model System for Developmental Biology.......................................... 11 Maize Endosperm Development............................................................................................. 12 Mutants with Aleurone Diffe rentiation Defects are Likely to be Involved in Cell Signaling F unctions............................................................................................................ .14 Transposon Tagging............................................................................................................. ..16 UniformMu.............................................................................................................................17 The Rough Endosperm3 Locus is Required for E ndosperm Development............................ 17 U2AF35-Related Splicing Factor...........................................................................................18 2 GENETICS OF rgh3 ..............................................................................................................21 Introduction................................................................................................................... ..........21 Results.....................................................................................................................................23 Identification of rgh3 .......................................................................................................23 Mosaics Kernels Illustrate the Non-Autonomy of rgh3 ..................................................24 The rgh3 Mutant Endosperm Disrupts Embryo Development........................................ 24 Inheritance of rgh3 ..........................................................................................................25 Genetic Mapping of rgh3 ................................................................................................25 Screen for rgh3 Alleles by Translocation Mapping ........................................................ 26 Introgression of rgh3 into W 22 and B73 Inbred............................................................. 26 Discussion...............................................................................................................................27 Materials and Methods...........................................................................................................27 Plants Growth Conditions................................................................................................ 28 Translocation Stock for the Long Arm of Chromosom e 5.............................................. 28 Allelism Test...................................................................................................................28 3 PHENOTYPE OF rgh3 ..........................................................................................................33 Introduction................................................................................................................... ..........33 Results.....................................................................................................................................34 Seed Phenotype Analysis of rgh3 ....................................................................................34 The rgh3 Mutation is Seedling Lethal .............................................................................35

PAGE 6

6 Causes of the rgh3 Endosperm Phenotype......................................................................36 The rgh3 Mutant Show Defects in Basal Endospe rm Transfer Cell Layer (BETL)....... 37 The Rgh3 Locus is Not Essential for Endosperm Cell Viability..................................... 38 Discussion...............................................................................................................................39 Materials and Methods...........................................................................................................41 Light Microscopy............................................................................................................41 Germination and Seedling Phenotype Analysis.............................................................. 42 Immunolocalization Analysis.......................................................................................... 42 Endosperm Tissue Culture.............................................................................................. 42 4 MOLECULAR ANALYSIS OF rgh3 ....................................................................................54 Introduction................................................................................................................... ..........54 Results.....................................................................................................................................55 Mu1 Insertion is Linked to the rgh3 Phenotype .............................................................. 56 Mu1Tagged Locus Shows Alternative Initiati on of Transcription and Alternative Splicing ....................................................................................................................... .58 Mu1Tagged Gene Encodes a Predicted Splicing Factor ................................................ 59 Molecular Lesion of the z murp Locus .............................................................................60 Discussion...............................................................................................................................61 Materials and Methods...........................................................................................................62 Genomic DNA and Total RNA Extraction..................................................................... 62 Southern Blot Analysis.................................................................................................... 63 MuTAIL-PCR..................................................................................................................63 Polymerase Chain Reaction Linkage Analysis................................................................ 64 Amplification of cDNA Ends .......................................................................................... 64 Molecular Analysis of Normal and Mutant ZmURP transcripts..................................... 65 5 CONCLUSIONS.................................................................................................................... 74 LIST OF REFERENCES...............................................................................................................78 BIOGRAPHICAL SKETCH.........................................................................................................84

PAGE 7

7 LIST OF TABLES Table page 2-1 Chromosome location of five rgh m utants........................................................................ 29 2-2 Inheritance of rgh3. ............................................................................................................32 2-3 Allelism tests of rgh3. ........................................................................................................32 3-1 A high percentage of rg h3/ rgh3 endosperm explants were able to grow compared to normal siblings and A636 inbred....................................................................................... 53 4-1 Nucleotide BLAST search of the ri ce Os02g35150 predicted gene identified two m aize ESTs and one maize truncated cDNA.....................................................................68 4-2 Protein BLAST searches results of the Mu1 -tagged predicted protein.............................. 71

PAGE 8

8 LIST OF FIGURES Figure page 1-1 UniformMu mutagenesis scheme...................................................................................... 20 2-1 Mosaic kernels illu strate non-autonomy of rgh3 ...............................................................30 2-2 The rgh3 mutant endosperm disrupts embryo development.............................................. 31 3-1 Mature rgh3 m utant kernels show a range of phenotypes................................................. 44 3-2 Seedling phenotypes of rgh3 m utants................................................................................ 45 3-3 Root defects of rgh3 m utant seedlings............................................................................... 46 3-4 The rgh3 mutant seedlings show no apparent depletion of the seed endosperm reserve................................................................................................................................47 3-5 Aleurone defects in rgh3 m utant kernels........................................................................... 48 3-6 Starch accumulation is not affected in rgh3 m utants......................................................... 49 3-7 Delayed expression of an al euron e specific protease in rgh3 ............................................50 3-8 Basal endosperm transfer cell layer (BETL) defects in rgh3 mutant kernels .................... 51 3-9 The rgh3 locus is not essential fo r endosperm cell viability............................................. 52 4-1 Cosegregation analysis was used to identify and clone a Mu1 inser tion linked to rgh3 ...66 4-2 Expanded linkage analysis between the Mu1 -tag and the rgh3 phenotype ....................... 67 4-3 Alternative splicing of the ZmURP transcript ...................................................................69 4-4 Structure of the ZmUrp gene..............................................................................................70 4-5 Sequence alignment of UHM domains of ZmURP, hom ologous URPs and U2AF35 protein................................................................................................................................72 4-6 Molecular lesion of the zmurp locus ..................................................................................73

PAGE 9

9 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PUTATIVE ROLE FOR RNA SPLICI NG IN MAIZE ENDOSPERM-EMBRYO DEVELOPMENTAL INTERACTIONS IS REVEALED BY THE rough endosperm 3 ( rgh3) SEED MUTANT By Diego S. Fajardo May 2008 Chair: Andrew M. Settles Major: Plant Molecular and Cellular Biology Endosperm-embryo interactions are an impor tant but poorly understood aspect of seed development. The Rough endosperm3 ( Rgh3) locus is involved in these interactions at a developmental level. The rgh3 seed mosaics marked with the pr1 anthocyanin gene indicate that Rgh3 is required in the endosperm for the normal development of the embryo. The Rgh3 locus also has an autonomous function in the embryo and is required for seedling viability. The Rgh3 locus maps to the long arm of chromosome 5. Co mplementation tests with other seed mutants mapped to the same region indicated that Rgh3 is a novel locus. We identified a tightly-linked transposon-tag from the rgh3-70 allele. This transposon inser tion disrupts a predicted splicing factor with a U2 snRNP auxili ary factor homology motif (UHM) that is most similar to the U2AF35-Related Protein (URP). The UHMs ar e RNA recognition motif (RRM)-like domains that function in protein-prot ein interactions. Analysis of Rgh3/ZmURP cDNAs and genomic sequences suggests ZmURP produces alternatively spliced transcripts Maize ESTs indicated that ZmURP is expressed in endosperm tissue. Consis tent with this expr ession pattern, the rgh3 phenotype is characterized by an overprolifera tion of the aleurone cells and aberrant development of the basal endosperm transfer cell layer. An analysis with endosperm cell type

PAGE 10

10 specific markers in mutant rgh3 seeds suggests that the basal endosperm transfer cell defect occurs after cell type spec ification. Tissu e culture of rgh3 endosperm cells showed robust growth and formed homozygous mutant callus. These data indicate that Rgh3 is not an essential gene for gene expression. Instead the data presented argue that the putative R NA splicing function of Rgh3 has an important role in cell differentiation, which then impacts endosperm-embryo developmental interactions.

PAGE 11

11 CHAPTER 1 INTRODUCTION The m aize seed constitutes one of the worlds most important renewable sources of food and industrial raw material. For economical reas ons, great effort and resources have been invested to identify and study seed mutants a ffecting starch, protein and lipid composition. Analysis of seed mutants such as the opaque floury shrunken brittle and sugary classes have supplied a wealth of information on endosperm genetics, gene-enzyme relationships, and biosynthesis pathways of storage products in the kernel (reviewed in Boyer and Hannah, 2001). The Maize Seed as a Model Sy stem for Developmental Biology In addition to its economic im portance, the ma ize seed is also an attractive system for developmental biology. The maize seed consists of an embryo and a persistent endosperm. The embryo and endosperm are the results of double fe rtilization. Each pollen grain contains two sperm cells, upon fertilization, one ha ploid sperm (1n) nucleus fuses wi th the egg cell (1n) of the female gametophyte to form a diploid embryo (2n) The second sperm (1n) fuses to the diploid central cell (2n) to give rise to the triploid endosperm (3n). The resulting embryo and endosperm are genetically identical except for their ploidy level. The endosperm and embryo develop coordi nately. Crosstalk between them during fertilization appears to ensure a synchronized development. Support for this early signaling interaction can be found in the Arabidopsis cdka;1 mutant (Nowack et al., 2006). CDKA;1 is a class of cyclin-dependent kinase (CDK) which is paternally imprinted. The cdka;1 mutant contains only one sperm cell instead of two in the pollen grain. The mutant pollen is viable but it only fertilizes the egg cell and not the central cell. However, the unfertilized central cell develops. Endosperm development in the absence of the second fertiliza tion event is abnormal and leads to seed abortion. Nowack and co-workers (2007) conclude that the fertilized egg cell

PAGE 12

12 may provide a signal that can promote initial endo sperm growth. This initial signal is modulated by maternal epigenetic changes to the central cell. The maternal state represses central cell divisions through DNA methylation and histone modification (reviewed in Huh et al., 2007). Maternal repression of the central cell is mediated by the Polycomb group (PcG) genes Medea ( MEA ), Fertilization-Independent Seed2 ( FIS2) and Fertilization-Independent endosperm ( FIE ). mea, fis2 and fie mutations lead to the abnor mal proliferation of the centr al cell in the absence of fertilization, producing a non-viab le seed. Fertilization of a fis2 fie and mea mutants with cdka;1 pollen produces viable seeds with homodiploid endosperm (2n) in the absence of paternal genome contribution (Nowack et al ., 2007). This suggests that an unfertilized dipl oid central cell in the female gametophyte has the full potenti al to develop functional endosperm without paternal contribution, requiring a signal from the embryo (Nowack et al., 2007; reviewed in Ohad 2007). The maize endosperm is more sensitive to genomic perturbation than the Arabidopsis endosperm. Maize requires a ratio of 2 ma ternal to 1 paternal genome for normal endosperm development, rather than a specific ploidy le vel (Lin, 1984). Currently, most molecular development research on the endosperm comes from the model plants Arabidopsis and maize. However, differences in early (reviewed in Scott and Spielman, 2006) and late endosperm development (reviewed in Olsen, 2004a) between th ese two plant models, makes it difficult to use Arabidopsis as the only model for identifying genes involved in endosperm-embryo interactions in cereals. Maize Endosperm Development The m aize endosperm is structurally a simple tissue consisting of four major cell types: the starchy endosperm, the aleurone, the embryosurrounding region (ESR) and the basal endosperm transfer layer (BETL) (reviewed in Olsen, 2001). Following fertilization the maize endosperm

PAGE 13

13 goes through four main stages of development: syncytial, cellulari zation, cell fate specification and differentiation (reviewed in Becraft et al., 2001; Olsen, 2004). Initially, the triploid nucleus undergoes a series of free nuclear divisions to form a single-cel led syncytium. After several rounds of mitotic divisions, the nuclei migrate to the peri phery of the central vacuole (Kiesselbach, 1949). Cell wall form ation between non-daugh ter nuclei initiates at the periphery of the endosperm and progresse s centripetally (Kie sselbach, 1949; reviewed by Olsen, 2004). The cellularized peripheral layer then goes throug h a series of anticlinal divisions to form a single uniform layer. At the same time, the endospe rm internal layers begin to cellularize. Soon after endosperm cellularization ends, the peripheral cell layer differentiates into the aleurone and a series of random cell divisions resumes in the inner cell files to form the starchy endosperm (reviewed by Olsen 2001). These inner cells und ergo a nuclear endoduplica tion resulting in an increase of copies of their nuclear DNA and cell size (reviewed in La rkins et al., 2001). The mature aleurone is a desiccation tolerant single cell layer of densely cy toplasmic cells between the starchy endosperm and the maternal pericarp tissue. During germination, the aleurone cells secrete hydrolytic enzymes that break down th e storage compounds in the starchy endosperm, making free amino acids and simple sugars av ailable for uptake by the embryo (Becraft and Asuncion-Crabb, 2000; Lopes and Larkins, 1993; reviewed by Olsen 2001). The embryo-surrounding region (ESR) and the basal endosperm transfer layer (BETL) are two other specialized cells in the endosperm. The ESR is thought to participate in embryo nutrition and embryo-endosperm communication (reviewed by Olsen 2004). Similarly, the BETL participates in carrying nutrients from the mate rnal plant to the endosperm (Thompson et al., 2001). The BETL consists of two to three layers of ce lls over maternal vascular tissue at the basal tip of the developing seed (Kiesselbac h, 1949). These cells are characterized by the

PAGE 14

14 formation of wall ingrowths, producing up to 20-fold increased surface area (Hueros et al., 1995; Opsahl-Ferstad et al., 19 97; Gomez et al., 2002). Despite the detailed morphological descripti on of endosperm development, the molecular basis of endosperm cell specification and di fferentiation is largel y unknown (Becraft and Asuncion-Crabb, 2000). However, a large number of seed mutants involved in endosperm development have been identified (Neuffer and Sh eridan, 1980; Neuffer et al., 1986; Scanlon et al., 1994). A detailed molecular analysis of thes e mutants has been hindered by the number of loci that give rise to seed phe notypes and the lethal nature of most of these mutations (Neuffer and Sheridan, 1980). In spite of these genetic challe nges, in recent years, a series of seed mutants involved in endosperm cellular development have been molecularly characterized (Becraft et al., 1996; Lid et al., 2002; Shen et al., 2003; Lid et al., 2004; da Costa e Silva et al., 2004; Stiefel et al., 1999; Scanlon and Meyers, 1998; Costa et al., 2003). Mutants with Aleurone Differentiation Defects are Likely to be Involved in Cell S ignaling Functions crinkly4 ( cr4 ), defective kernel1 (dek1), and supernumerary aleurone cell layers1 ( sal1 ) are three mutations that stochastically disrupt no rmal development of the aleurone cell layer. These genes encode products consistent with cellto-cell signaling function s (Becraft et al., 1996; Lid et al., 2002; Shen et al., 2003; Lid et al., 2004). Crinkly4 encodes a transmembrane receptor kinase with an extracellular domain similar to the tumor necrosis factor receptor (TNFR) (Becraft et al., 1996; reviewed in Becraft, 2002). TNFR is a member of a large family of cell surface receptors involved in lymphocyte developm ent (Chan et al., 2000). It is proposed that Cr4 works as a receptor for an unknown ligand that exists in the periphery of the endosperm (Olsen et al., 1998). The cr4 mutation inhibits aleurone formation on some parts of the abgerminal face of the endosperm (Becraft et al., 1996). Similar to cr4 dek1 shows a loss of

PAGE 15

15 aleurone cells. Dek1 encodes a transmembrane protein with a membrane targeting signal in its Nterminus and a calpain-like Cys proteinase domain in its cytosolic C-term inus (Lid et al., 2002; Wang et al., 2003). Becraft and Asunc ion-Crabb (2000) suggest that Dek1 is constantly required for aleurone differentiation. In contrast to cr4 and dek1 sal1 mutants form multiple aleurone cell layers. Sal1 functions to limit the differentiation of aleurone cells to the outer cell layer. Sal1 encodes a predicted vacuolar so rting protein, a member of the Chmp1 gene family (Shen et al., 2003). etched1 ( et1 ) is a dek mutant that alters endosperm deve lopment giving rise to etched and pitted seed phenotypes (da Costa e Silva et al., 2004). The et1 seed phenotype is caused by a loss of plastid function in endosperm cells and not by an over-proliferation or loss of aleurone cells. This loss of plastid function affects the normal accumulation of starch in the endosperm cells. These starchless cells create depressions and crevices on the endosperm surface, hence the etched seed surface phenotype. The Et1 gene encodes a chloroplast localized protein similar to the eukaryotic transcription elongation factor TFIIS (da Costa e Silva et al., 2004). Similar to cr4 the effect of et1 mutation is variable throughout the endosperm (da Costa e Silva et al., 2004). An integrated model of aleurone differentiation has not been co mpletely elucidated yet. It has been proposed that aleurone cell identity is specified excl usively in response to surface position, a signal within th e endosperm activates Cr4 while Dek1 is required to maintain the aleurone identity throughout development Sal1 limits the aleurone to a single cell layer (Gruis et al., 2006; Tian et al., 2007). The predicted roles of Cr4 Sal1 and Dek1 suggest that mutations affecting aleurone differentiati on are likely to be involved in cell-to-cell signaling functions. These genes are involved in regulatory func tions required for the normal development of

PAGE 16

16 multiple tissues like leaves (e.g. Cr4, Et1) and embryos (e.g. Dek1, Sal1 ), suggesting that cereal endosperm development uses common mechanisms with other plant developmental processes. Based only on the pred icted functions of et1 dek1 sal1 and cr4 mutants, it is not obvious that they would have important roles in seed development. This gives a strong argument for the power of forward genetic analysis of mutant phe notypes to define gene functions required for seed development. All four of these genes were isolated from transposon-tagged lines (Becraft et al., 1996; Lid et al., 2002; Shen et al., 2003; Lid et al., 2004; da Costa e Silva et al., 2004). Transposon Tagging Transposons are the prim ary tool for taggi ng and cloning maize genes. Three endogenous transposable element families are typically utilized for genetic studies in maize: Activator ( Ac/Ds), Enhancer / Supressor Mutator ( En / Spm ) and Mutator ( MuDR/Mu) (reviewed in Walbot, 2000). Mutator is a two-component system of autonomous ( MuDR ) and a family of nonautonomous ( Mu ) transposable elements that share th e terminal inverted repeat sequences (reviewed in Walbot and Rudenko, 2002; Li sch, 2002). In the presence of an active MuDR Mu elements undergo replicative tran sposition in pre-meiotic cells. Mu elements typically have 50200 copies/genome and are characterized for a hi gh rate of forward mutagenesis (Robertson, 1978; Walbot and Warren, 1988). These Mu insertions occur preferen tially in gene-rich regions (Cresse et al., 1995; Lunde et al., 2003; Settle s et al., 2004; McCarty et al., 2005). Due to the high forward mutation state, Mu stocks have been used to generate multiple transposon tagging populations for forward and reverse genetics st udies (reviewed in Brutnell 2002; Settles, 2005). These populations include the Tr ait Utility System for Corn (TUSC), The Maize Targeted Mutagenesis database (MTMdB), RescueMu and UniformMu populations (Bensen et al., 1995; May et al., 2003; Lunde et al., 2003; McCarty et al., 2005). Some of these Mu populations

PAGE 17

17 segregate for background deks in the M2 families making it difficult to identify independent isolates of seed mutants (reviewed in Settles, 2005). UniformMu The m aize Endosperm Development Project (McC arty et al., 2005) has developed a unique mutagenic population, named UniformMu, as a tool for efficient isolation of transposed tagged mutants that interfere with endosperm and seed development. UniformMu is a Mu transposonactive population, containing a bz1mum9 somatic activity marker gene (Chomet, 1996), which is introgressed into the W22 inbr ed (figure 1-1). UniformMu addresses most of the drawbacks of other mutagenic populations by having a high mutagenic activity but maintaining a steady-state copy number of Mu elements through continuous backcro ssing into the W22 inbred. This also provides a uniform genetic background, thus redu cing phenotypic variation due to heterogeneity (McCarty et al., 2005). Importan tly, parental seed mutations were removed by screening Mu active parents for plants carryi ng pre-existing mutations. This en sured that all seed mutants identified in UniformMu derived from independent mutagenic even ts. Finally, the transposons in UniformMu lines can be stabilized readily through the use of the bz1-mum9 marker. Crossing these stabilized lines to W22 further rem oves non-phenotypic transposons through genetic segregation and generates material for molecular analysis. The UniformMu population has been screened for mutants showing vi sible grain-fill and seed abortion phenotypes. Based on pedigree analys is and general phenotypi c classes of the seed phenotypes, ~2,150 independent mutants were identified (McCarty et al., 2005). The Rough Endosperm3 Locus is Required for Endosperm Develop ment Settles and co-workers completed a secondary screen of UniformMu to identify mutants in a sub-class of the maize defective kernel ( dek ) mutants characterized by a rough, etched, or pitted seed surface phenotype. This cl ass of mutants was termed rough endosperm ( rgh) mutants. rgh

PAGE 18

18 mutants frequently show disrupt ions in aleurone cell different iation and embryo developmental defects analogous to dek1 cr4 and sal1 defects. Based on these phe notypic similarities, it is possible that the rgh mutants might be enriched for genes needed for cell-to-cell signaling or cellular differentiation in the endosperm. To enrich for rgh mutants that are more likely to have endosperm-embryo signaling functions, a small collection of rgh seed mosaics were examined. B-A translocations were used to determine chromosome position and generate seed mosaics to identify mutations involved in endosperm-embryo interactions (Chang and Neu ffer 1994; Neuffer and Sheridan, 1980; Neuffer et al., 1986). In the following chapters, the Rough endosperm3 ( Rgh3) locus is shown to be required for endosperm cellular differentiati on, endosperm to embryo interactions, embryo development, and seedling development/viability. U2AF35-Related Splicing Factor A tightly lin ked rgh3 transposon-tag indicates that Rgh3 is likely to encode an ortholog to the human splicing factor U2AF35-related protei n (URP). The URP protein is characterized by an acidic N-terminal domain, two highly conser ved zinc fingers flanking a central U2AF homology motif (UHM) and an ar ginine/serine rich (RS) C-terminal domain. The URP protein has been shown to interact in vitro with the splicing factor U2AF65 (Tronchere et al., 1997), to form a heterodimer required for recognizing the 3 intron-exon boundary as part of the spliceosome complex in pre-mRNA splicing. Similarly to URP, the U2AF35 splicing factor has also been shown to interact with U2AF65 (reviewed in Ki elkopf et al. 2004). However, Tronchere and co-workers (1997) showed URP to be functionally distinct from U2AF35 because U2AF35 cannot complement URP-depleted extr acts. Additionally, phy logenetic analysis indicates that homologs of U2AF 35 are found in protozoa, yeast, worms, insects and chordates, while homologs of URP exist in insects and chorda tes (Mollet et al., 2006). This further suggests

PAGE 19

19 distinct functions for URP and U2AF35 splicing factors, with more specialized functions expected for the URP homologs. This prediction is supported by endosperm cell culture assays. rgh3 mutants readily develop immortalized cell lin es suggesting that URP is not required for general cell viability.

PAGE 20

20 Figure 1-1. UniformMu mutagenesis scheme. Mu -active stocks carrying the bz1-mum9 marker were backcrossed into the W22 inbred (M0). Each Mu -active parent was selfpollinated, screened for seed phenotypes and scored for somatic Mutator activity (dotted arrow). Only Mu -active parents without seed mutant phenotypes (normal ears) were used for the next round of backcrosses (BC2 to BC6) (M1). Seed mutants were collected from M2 families in the BC4 to BC7 generations (dashed arrow). Each novel mutant was tested for genetic heritability by self pollinating normal Mu inactive (bz1/bz1 kernels with no spots) from a segregating M2 ear (gray boxes). Inset picture shows transposon somatic activ ity markers in UniformMu population. Mu somatic reversion of the bz1-mum9 allele produces densely spotted kernels (black arrow). Open arrow shows bz1/bz1 Mu -inactive kernel on th e right. White arrow points a full color kernel.

PAGE 21

21 CHAPTER 2 GENETICS OF rgh3 Introduction The m aize seed is composed of an embryo and a persistent endosperm. The triploid endosperm is a simple structure that originates as the result of the fusion of the diploid central cell nucleus with one of the two sperm cell nuclei. The diploid embryo originates from the fusion of the haploid egg cell nucleus with the other sperm cell nuclei (reviewed in Gehring et al., 2004). The maize seed provides a good system to study the genetic regulation of endospermembryo interactions during seed development, becau se the endosperm is a large persistent tissue at seed maturity. In addition, maize has good ge netic resources to generate seed mosaics. B-A translocations are a unique genetic tool in maize that allows recessive genes to be located to a chromosome arm in the F1 progeny (Neuffer and Sheridan, 1980). B-A translocations generate mosa ic kernels that can be used to identify embryo-endosperm interactions (Neuffer and Sheridan, 1980; Neuf fer et al., 1986; Cha ng and Neuffer, 1994; Scanlon et al., 1994). B-A translocat ions are reciprocal translocatio ns between the basic, or A, set of chromosomes and the supernumerary B chromosome (reviewed in Beckett 1996). B chromosomes are extra chromosomes that serv e no vital function in the organism and are unrelated to the normal complement. The reciprocal translocation produces a BA chromosome, bearing the B centromere, which carries a factor for non-disjunction, and an AB chromosome with the A centromere. At the second pollen mitosis, the BA chromosome undergoes nondisjunction at a frequency rangi ng from 50-95% of divisions (reviewed in Carlson, 1978). The non-disjunction of the BA chromosome produces gametes with one sperm cell carrying the ABBABA chromosomes and the other sperm cell the AB chromosome. The AB sperm is deficient for all genes located in the translocated region. Thus, when the AB sperm fuses with either the

PAGE 22

22 polar nuclei or the egg th e recessive genes from the female pa rent are phenotypically expressed. Hence, B-A translocations uncover recessive mutations that map to the translocated chromosome arm. The second sperm cell carries the ABBABA chromosomes, which will complement recessive mutations. Thus, double fert ilization with a B-A tr anslocation generates an embryo-endosperm mosaic. B-A translocation stocks are available for 19 of the 20 maize chromosome arms (reviewed in Beckett, 1996). The role of the developing endosperm on the immature embryo during seed development has been studied using B-A translocations. These studies have shown that in the majority of defective kernel ( dek ) mutants the endosperm and the embryo develop in an autonomous manner (Neuffer and Sheridan, 1980; Neuffer et al., 1986; Chang and Neuffer, 1994; Scanlon et al., 1994). The genotype of the endosperm or embryo determines the phenotype of these tissues directly. However, mutants like globby1 ( glo1) (Costa et al., 2003) and discolored-1 ( dsc1 ) (Scanlon et al., 1998) become of sp ecial interest due to an appa rent non-autonomous role in the endosperm. Costa et al. (2003) and Scanlon et al. (1998) showed that wild-type embryos in glo1 and dsc1 B-A mosaics were unable to develop in the presence of defective endosperms, suggesting that the endosperm f unction of these genes influence embryo development. However, the glo1 mutant has not been cloned yet and no mo lecular function has been assigned to the dsc1 mutation. The rough endosperm3 ( rgh3) mutant was identified in a B-A translocation mosaic screen of 5 rgh mutants for non-autonomous mutations. All 5 rgh mutants were initially isolated from the UniformMu population. Similar to the glo1 and dsc1 mutations, rgh3 embryo development is influenced by the endosperm. Se gregation ratios suggest that rgh3 is a recessive allele with no gametophyte effects. These data indicate that the product of the Rgh3 locus is required in the

PAGE 23

23 endosperm for the development of genetically normal embryos. The Rgh3 locus has additional functions in embryo development inde pendent of the endosperm function. Results Identification of rgh3 The rough endosperm seed m utants ( rgh) are a sub-class of defective endosperm mutants characterized by rough, etched or pitted seed surface defects. An initial phenotypic characterization of the rgh mutants showed defects such as: i rregular aleurone development with multiple cell layers, aberrant aleurone cell growth and altered cell shapes of the aleurone (data not shown). These defects are sim ilar to previously characterized etched1 ( et1) Supernumerary aleurone layers1 ( sal1 ) and disorganized alerone layer ( dil1 and dil2) seed mutants (Costa e Silva et al., 2004; Shen et al., 2003 ; Lid et al., 2004; da). These phenotypic similarities suggested that rgh mutations could be involved in plastid sp ecific functions, cell-type differentiation or cell-to-cell signaling. Two hundred and forty three independent rgh seed mutants were identified from the UniformMu transposontagging population. Because most rgh mutants are homozygous lethal, mutants were maintained in a heterozygous state. The rgh3 mutant was identified in an initial screen of 8 rgh isolates tested with 19 B-A transl ocations that uncover all chromosome arms except the small arm of chromosome 8 (8S). rgh3-70 showed distinct uncoverings in multiple F1 crosses for the long arm of chromoso me 5 (5L) (table 2-1). This mapped rgh3-70 to the long arm of chromosome 5. Simple sequence repeat (SSR) markers placed rgh3-70 near bins 5.04/5.05 of chromosome 5 (Carson et al, 2004). Four other rgh isolates showed unique uncoverings for individual chromosome arms (t able 2-1) and were also confirmed by SSR markers (Carson et al, 2004).

PAGE 24

24 Mosaics Kernels Illustrate the Non-Autonomy of rgh3 B-A m osaics for all 5 rgh isolates were further screen ed for phenotypes that showed endosperm-embryo interactions. Longitudinal hand sections of mature seeds from B-A uncovering crosses for four of the five rgh mutants showed mosaic seeds with mutant endosperms and normal embryos (figure 2-1). Thes e data indicate that the defective endosperm has no effect on embryo development for these mutants. However, rgh3-70 B-A mosaics showed a defective endosperm and defective embryo phenot ype (figure 2-1). This result suggests that rgh3-70 has a non-autonomous phenotype and a func tional gene is required in either the endosperm or the embryo for normal seed development. The rgh3 Mutant Endosperm Disrupts Embryo Development To determ ine the direction of the non-autonomous phenotype, rgh3-70 was introgressed into the red aleurone (pr1 ) color background. The B-A translocation for the long arm of chromosome 5 (TB-5La) carries a full color (purple) allele of Pr1 (Birchler and Alfenito, 1993). The presence of the B-A translocation can then be determined by scoring the endosperm or embryo for the presence of purple color. To determine the frequency of non-disjunction of the TB-5La males were crossed to pr1/pr1 female plants. Results from these crosses indicate that approximate 1/2 of the TB-5La males (21 out of 39 total crosses) showed non-disjunction. Introgressed rgh3-70 pr1/Rgh3 pr1 plants were crossed as females to TB-5La stocks. Two classes of defective mosaic kernels were observed as a result of this cross. The first class of mosaic kernels contained a red mutant endosperm and a purple defectiv e embryo. These results suggest that Rgh3 is required in the endosperm fo r the genetically normal embryo to development. The second class of mosaic kern els contained defective red embryos and normal purple endosperm. These data suggest an additional autonomous function of Rgh3 in the embryo. However, it is possible to obtain the same cl asses of defective mosaic kernel if the rgh3 locus has

PAGE 25

25 a maternal-effect specific to the embryo. This ki nd of gametophytic effect has been observed in the Arabidopsis prolifera ( prl ) mutant (Springer et al., 2000). The prl mutant seeds show a continuous range in seed phenotype due to differing levels of the protein at fertilization. A similar continuous range in seed mutant phenotype is observed in the rgh3 mutant (figure 3-1CF). If rgh3 transmission behaves similarly to prl non-Mendelian segregation ratios should be expected for self-pollinations and female crosses. Inheritance of rgh3 The rgh3-70 allele was tested for fem ale transm ission and pollen transmission, to confirm that rgh3-70 B-A mosaics phenotypes are not due to a gametophytic effect, by crossing rgh3-70 as a female or male with the W22 inbred. The rgh3 mutant was tested for altered segregation ratios, by counting mutant and normal kernels from selfpollinated ears of rgh3-70/+ plants from multiple seasons. A Mendelian 3:1 ratio of normal to rgh3 endosperm phenotypes was observed (table 2-2). To test for maternal effects, rgh3 /+ was crossed as a female to W22. F1 progeny were self-pollinated and scored for ears segregating for rgh3. A 1.05:1 ratio of normal to segregating ears was observed (table 2-2). In addition, rgh3 transmission through pollen was measured. F1 progeny from W22 crossed by rgh3/+ were self-pollinated and scored for ears segregating for rgh3. A 1.06:1 ratio of normal to segregating ears was observed (table 2-2). These results indicate that rgh3 behaves as a recessive locus a nd supports the interpretation that Rgh3 is a nonautonomous locus during seed development. Genetic Mapping of rgh3 rgh3 m utant was further mapped using two-point linkage analysis with pr1 rgh3-70 Pr1/Rgh3 Pr1 plants were crossed to Rgh3 pr1/Rgh3 pr1. Normal red kernels were selected from the progeny and self-pollinated to scor e single recombinants events. The rgh3 mutant mapped 6.7 cM from pr1 (7 single recombinants /104 meiotic products).

PAGE 26

26 Two point linkage data positioned rgh3 at 6.7 cM from pr1. Two seed mutants, dek33 and the pitted rough germless1 ( prg1), are known to map near pr1. Complementation tests were carried out between rgh3-70 dek33 and prg1. Two more seed mutants, dek9 and dek26 were identified to map to chromosome 5L by B-A tran slocation. These mutants were also tested for complementation with rgh3-70. Results from these tests suggest that all four loci are not alleles of rgh3 (table 2-3). Screen for rgh3 Alleles by Translocation Mapping An approach to identify a possible allelic group within a large set of m utants is by map location. By mapping seed mutants to a chromoso me arm, it greatly simp lifies alleli c testing, because only mutants that map to the same chromosome arm need to be tested for allelism. Scanlon et al. (1994) identified 20 cases of allelism from a se t of 63 seed mutants by mapping them to chromosome arms using B-A translocati ons and directing complementation tests, based on the map locations of the mutants. An initial subset of 122 rgh mutants was tested for the long arm of chromosome 5 with the goal of identifying possible alleles of rgh3. The TB-5La crosses uncovered the rgh phenotype in 11 of the 122 rgh tested (data not shown). However, complementation tests with rgh3-70 indicated that all 11 rgh mutants were not allelic to rgh3 (data not shown). In this pool of rgh mutants mapped to 5L, 2 mutants gave false positives for TB-5La uncovering due to a female gametophytic effect of the mutation. Introgression of rgh3 into W22 and B73 Inbred In order to reduce the number of non-segregating Mu insertions of the rgh3 m utant, rgh370 heterozygous plants were backcrossed to W2 2 for multiple generations. Advanced backcross material was generated to reduce the number of Mu insertions. W22 contains approximately 21 Mu insertions in comparison with 57 Mu insertions of typical Uniform Mu Mu -inactive mutants

PAGE 27

27 (Settles et al., 2004; Mc Carty et al., 2005). The rgh3 mutant was also cro ssed to the B73 inbred background to generate a F2 mapping population. Discussion The rgh3 mutant is a novel recessive locus th at shows a non-autonom ous function during seed development. Data from marked rgh3 B-A mosaic experiments suggest that the rgh3 endosperm influences embryo development and rgh3 has an additional autonomous function in the embryo. The B-A mosaic data is further supported by the Mendelian segregation ratios observed in rgh3. The rgh3 mutant behaves as a recessive lo cus within the seed and was mapped 6.7 cM from pr1. The rgh3 mutant was also tested for allelism with dek9 dek26 dek33 and prg1 seed mutants on chromosome 5L. These complementation tests indicated that rgh3 is novel locus near the centromere of chromosome 5. Previous genetic studies with B-A mosaics on a large number of dek mutants showed that in the majority of mutants, the endosperm and embryo develop in an autonomous manner (Neuffer and Sheridan, 1980; Neuffer et al., 1986; Chang and Neuffer, 1994; Scanlon et al., 1994). In a few cases, the mutant embryos benefited from the presence of the genetically normal endosperm. These dek mutants were classified as nutritional mutants. These mutants support the idea of the nurturing role of the endosperm during seed development and germination. rgh3, glo1 and dsc1 represent a minority of seed mutants by showing developmental endosperm to embryo interactions. Materials and Methods Due to the lethal n ature of most rgh mutants, the rgh lines were kept in a heterozygous condition. To generate rgh heterozygous plants, normal seeds were collected from rgh segregating ears and planted. I expect 2/3 of the collected seeds to be rgh heterozygous.

PAGE 28

28 Plants Growth Conditions Plants were grown in fields at University of Florida Plant Science Research and Education Unit (Citra, Fl). Greenhouse grown plants were planted in 3 gallon pots with Metro-Mix 300 (Scotts-Sierra. Marysville, OH). Greenhouse plan ts were grown under the following regim en: 14 hours day length at 28C during the day, and at 20 C at night. Humidity levels were at 70%. Translocation Stock for the Long Arm of Chromosome 5 The TB-5La stock uses easily identifiable Pr (purple) pr (red) anthocyanin m arkers to track the translocation (Birchler and Alfenito, 1993). For TB-5La, the normal Pr gene is located on the B-A translocation and confers a purple pi gment to the seed tissues carrying the B-A chromosome. Using this marker system, seeds ca rrying the B-A translocation were selected and crossed to rgh/+ plants. To determine the frequency of non-disjunction, TB-5La plants were crossed as males to pr/pr female plants. If a rgh mutation maps to the chromosome 5L, we expect 1/3 of the crosses to uncover the rgh phenotype (2/3 rgh/+ X 1/2 hyperploid male = 1/3). A minimum of 6 crosses were required to be with in a 92% confidence interval that at least one uncovering will be observed if the rgh isolate maps to chromosome 5L. Allelism Test Twenty norm al kernels from segregating ears were planted per rgh mutant. Whenever possible, 10 plants from each rgh mutant were self-pollinated and crossed onto 10 plants of a rgh3 mutant, the remaining 10 plants were used as females for the reverse cross, so each mutant had 10 crosses as a male and 10 as a female for each complementation test. If the mutant crossed is an allele of rgh3, 4/9 of the resulting ears were e xpected to segregate 3:1 for the rgh3 seed phenotype. Self-pollinations confir med the genotype of the male parent. When available, second ears were selfed on females.

PAGE 29

29 Table 2-1. Chromosome location of five rgh mutants. Mutant Isolate Chromosome location SSR Mo17 F2 SSR B73 F2 rgh-00S-005-14 4S 4.03 4.03 rgh3-070 5L 5.04-5.05 5.04-5.05 rgh-99F-253-18 6L 6.04 6.04-6.05 rgh-99F-249-02 6L 6.05 6.04-6.05 rgh-99S-116-06 10L 10.05 10.04-10.07 Five rgh mutants were selected for an initial scre en to uncover mutation involve in endospermembryo interactions. Chromosome arm loca tion was determined by B-A translocation uncovering. Arm locations were confirmed by SSR markers in two different genetic backgrounds. SSR results show the bin where th e mutations were mappe d. Short chromosome arm = S. Long chromosome arm = L.

PAGE 30

30 Figure 2-1. Mosaic kernels i llustrate non-autonomy of rgh3. A) Shows longitudinal hand sections of recessive mutant kernels from 5 rgh loci. B) Shows longitudinal hand sections of corresponding BA mosaic kernels that are likely to have hyperploid embryos and hypoploid endosperms. All of the B-A mosaics except rgh3-70 showed well-defined embryonic leaves a nd roots (black arrows). The rgh3-70 B-A mosaics show a defective embryo with no defined em bryonic leaves and r oots (white arrows). Scale bars = 2mm. rgh3-70 rgh-00S-005-14 rgh-99F-249-02 rgh-99F-253-18 rgh-99S-116-06 A B

PAGE 31

31 Figure 2-2. The rgh3 mutant endosperm disrupts embryo development. Marked mosaics were developed to determine the direction of the rgh3 non-autonomous phenotype. A-D) The presence of the B-A translocation can be tracked by scoring the endosperm or embryo for the presence of the purple Pr1 marker color (white arrow). The red marker shows the uncovering of the recessive pr1 allele (black arrows). B and D) Show enlarged areas of A and C respec tively. E-K) Two classes of B-A mosaic kernels are found in the rgh3 uncovering in the pr1 background. E and I) Shows the first class of B-A mosaic with a hypoploid red mutant endosperm (black arrow) and a hyperploid purple (white arrow) defective embryo. F and J) Show enlarges areas of E and I respectively. I and K) The second class of mosaic kernels showed a hypoploid mutant red (black arrow) embryos with a hyperploid normal purple (white arrow) endosperm. Scale bars = 2mm. A C D E F G H B I J K M

PAGE 32

32 Table 2-2. Inheritance of rgh3. Self pollination # of normals # of mutants ratio P( X2) *rgh3/+ 664 (seeds) 207 (seeds) 3.2:1 0.4 **rgh3/+ x W22 83 79 1.05:1 0.75 ** W22 x rgh3/+ 76 (ears) 72 (ears) 1.06:1 0.74 Expected model for X2 test was 3 normal to 1 rgh3 seeds ** Expected model for X2 test was 1 normal to 1 rgh3 ear We tested rgh3-70 for altered segregation ratios to determine if the B-A mosaic phenotypes were due to maternal-effects. These results suggest that rgh3-70 behaves as a recessive locus. Table 2-3. Allelism tests of rgh3. Isolate Isolate dek phenotype WT phenotype P( X2) rgh3-70 dek9 0 8 6x10-3 rgh3-70 dek26 0 12 1x10-8 rgh3-70 dek33 0 7 2x10-4 rgh3-70 prg1 0 5 2x10-3 Expected model for X2 test was 2/3 dek to 1/3 normal The rgh3 mutant was mapped by B-A translocation and SSR markers to the long arm of chromosome 5. A twopoint linkage mapped rgh3-70 6.7 cM from pr1. The rgh3-70 mutant was crossed to dek33, prg1 (both mapped near pr1), dek9 and dek26 (which are mapped on chromosome 5L). The table shows crosses with confirmed rgh /+ and dek /+ plants used as males.

PAGE 33

33 CHAPTER 3 PHENOTYPE OF rgh3 Introduction The m ature maize endosperm consists of four major cell types: starchy endosperm, aleurone, the embryo-surrounding region (ESR), and basal endosperm transfer layer (BETL) (reviewed by Olsen, 2001). Despite the apparent structural simplicity, endosperm development is a complex process (Becraft and Asuncion-Cr abb, 2000; Costa et al., 2003; Young and Gallie, 2000; reviewed by Becraft et al., 2001). A detaile d cytological descriptio n of maize endosperm development has been reported (Kiesselbach, 1949) A large number of seed mutants involved in endosperm development have been identified (N euffer and Sheridan, 1980; Neuffer et al., 1986; Scanlon et al., 1994), but only a fraction of th ese mutants have been phenotypically and molecularly characterized. Thus, fundamental que stions about the mechanisms of early seed development are largely unanswered. However, a small number of mutations, which specifically affect the epidermal aleurone layer, have been characterized in detail These mutants provide insight into the aleurone cell fa te specification pathway. Genes involved in aleurone cell fate specification Crinkly4 ( Cr4 ), Defective kernet1 ( Dek1 ) and Supernumerary aleurone layers1 ( Sal1 ), encode proteins consiste nt with cell-to-cell signaling functions (Becraft et al., 1996; Lid et al., 2002; Shen et al., 2003; Li d et al., 2004). An integrated m odel of aleurone differentiation has not been completely elucidated yet. Howe ver, in a recent study it has been proposed that aleurone cell fate specificati on in maize endosperm occurs autonomously in response to endosperm surface position (Gruis et al., 2006; Tian et al., 2007). A critical process in seed development is the uptake of assimilates by the growing endosperm. Efficient assimilation is facilitate d by the development of the basal endosperm transfer layer (BETL) at the base of the endosperm over the main vascular tissue of the maternal

PAGE 34

34 plant. Several groups of transcripts have been identified to be expresse d in the BETL like the Basal endosperm transfe r cell layer1 to 4 ( Betl1-4 ) (Hueros et al., 1995; Hueros et al., 1999). Many of these proteins are simila r to antimicrobial proteins, sugge sting a role in defense against pathogens. Two seed mutants, reduced grain-filling1 ( rgf1 ) and globby1 ( glo1), have been identified to have BETL defects. The rgf1 mutant shows a significantly reduced expression of BETL1 and BETL2 proteins, however, it s hows normal BETL and embryo morphology. The rgf1 mutant exhibits a reduced accumulation of star ch but unaltered free sugar levels in the endosperm (Maitz et al., 2000). In contrast, the glo1 mutant shows defective BETL and embryo morphology and ectopic differentiation of th e aleurone (Costa et al., 2003). The rgf1 and glo1 seed mutants suggest that the nutrient conducting tissue has a direct role in determining final seed size, while glo1 suggests that the BETL may have a ro le in embryo development. However, neither of these mutants is cloned. Thus, the pa rticular mechanisms that distinguish between BETL roles in growth versus development are not known. The rgh3 mutants have a characteristic endosp erm morphology with an apparent loss of BETL cell identity late in seed development. More over, cells with aleurone characteristics form ectopically in the BETL regi on suggesting a role for Rgh3 in cell differentiation. Thus, the rgh3 mutant reveals the presence of a novel developmental cue re quired late in endosperm development to maintain BETL identity. Results Seed Phenotype Analysis of rgh3 Mature rgh3 m utant seeds are characterized by an etched or pitted surface and a reduced seed size compared to normal siblings seeds (Figure 3-1A). The rgh3 seed phenotype is readily visible as early as 15 DAP. Within a segregating ear, rgh3 mutants show variable seed phenotype severity, ranging from almost normal grain-fill to a nearly empty peri carp phenotype (Figure 3-

PAGE 35

35 1C to F). Longitudinal hand sections of muta nt seeds showed a corr elation between embryo defects and seed weight. The heaviest 10% of the mutant seeds developed near normal embryo morphology (figure 3-1C). Seeds heavier than the median produce embryonic roots (figure 31D). Seeds with weights below the median a nd the lightest 10% showed embryos with an abnormal development with no embryonic shoot or root-like structures (figure 3-2E to F). The rgh3 Mutation is S eedling Lethal Germination tests in soil were conducted for rgh3-70 mutant seeds to determine if they developed viable embryos. Seed selection was biased towards the less severe (heaviest 10%) mutant seeds to enrich for possible viable muta nt embryos. A fraction of the heaviest class of mutant seeds was able to germinate. Mutant seedlings showed stunted growth and died approximate 22 days after planti ng (figure 3-2A). These seedlings all showed aberrant narrow leaves. Germination tests of rgh3-70 mutant seeds were repeated in a B73/W22 hybrid F2. Mutant seedlings in this F2 background also showed stunted growth (figure 3-2B). In contrast to mutant seedlings in the W22 gene tic background, seedlings from F2 develop an open short round leaf (figure 3-2A and 3-2B). No differences in leaf coloration were detected between mutant and normal seedlings. Mutant seedlings in both genetic backgrounds s howed similar root defects. Mutant seedlings showed a short primary root with few or no secondary roots and some below ground short adventitious roots (figure 3-3). To determine if the early death of the s eedling was due to a depletion of the seed endosperm reserve, seed remnants were sectioned from necrotic seedlings at 18 and 20 days after emergence. All necrotic seedli ngs showed remaining endosperm tissue (figure 3-4B to D). Normal seedlings showed a collapsed pericarp w ith little endosperm tissue left (figure 3-4A). These results suggest that rgh3 lethality occurs prior to the depletion of the seed reserves.

PAGE 36

36 Causes of the rgh3 Endosperm Phenotype Longitudinal sections of mature rgh3-70 mutant seeds suggest that the etched or pitted seed surface characteristic of rgh3 seeds is due to over-p roliferation of the aleurone layer into the endosperm tissue (figure 3-5). This aleurone def ect was observed in all mutant seed size classes (figure 3-5B to D). The rgh3-70 seed surface phenotype is also reminiscent of the etched1 mutant, which is characterized by depressions and crevices on the endosperm surface. The etched1 seed phenotype is caused by a loss of plastid function in endosperm cells, affecting the normal accumulation of starch in endosperm cells. These starchless cells cr eate the etched seed surface phenotype (da Costa e Silva et al, 2004). To test for reduced starch accumulation in the rgh3 endosperm, longitudinal sections of rgh3-70 mutant seeds were st ained using an iodinepotassium-iodine solution (IKI) and found no sector s of reduced starch accumulation (figure 36B). A vp5 isolate from UniformMu was used as an IKI staining control. The vp5 allele has endosperm sectors of reduced starch similar to etched1 (figure 3-6C). Also, rgh3-70 seeds show an over-proliferation and abnormally shaped and sized aleurone cells (figure 3-5B to D). It has been proposed that aleurone cell identity is specified by an unknown signal that induces cells at the surface of the endosperm to differentiate into aleurone cells (Olsen et al., 1998; Becraft and Asuncion-Crabb, 2000; Shen et al ., 2003; Olsen, 2004; Gruis et al., 2006; Tian et al., 2007). A hypothesis based on this model is that the rgh3 endosperm cracks late in seed development and cells at the surface of the endosperm cr evasses differentiate into aleurone. To test this hypothesis rgh3-70 mutant and normal seeds from segregating and nonsegregating ears were collected at 20 days after pollination (DAP) when the rgh3 phenotype is readily visible. The seeds were sent to Gr egorio Hueros at University of Alcala for immunolocalization assays for an aleurone specific protease on bot h seed classes. This aleurone specific protease is normally expressed betw een 8 to 12 DAP (figure 3-7A inset). In rgh3 mutant

PAGE 37

37 seeds, the aleurone marker is expressed in th e endosperm crevasses at 20 DAP (figure 3-7B to C). Also, seeds from segregating and non-segregati ng ears were collected at 10 DAP, prior to the onset of the mutant seed phenotype, no abnor malities were observed in the endosperm morphology for all the seeds collected at this time point. These results support the hypothesis that the rgh3 defect in aleurone differentiation occurs late in seed development as a consequence of cracks in the endosperm. The rgh3 Mutant Sho w Defects in Basal Endosperm Transfer Cell Layer (BETL) The kernel phenotype of rgh3 seeds (reduced size) suggested that solute transport through the BETL might be affected. This possibility was investigated by analyzing rgh3 mutant and normal seeds for BETL morphology and expression of the BETL2 protein. Betl2 encodes a secreted peptide with similarity to an antimicro bial protein and is exclusively expressed in the BETL cells at approximated 10 DAP (Hueros et al., 1999) (figure 3-8A). Thus, antibodies against BETL2 were used as a marker for BETL cells. The BETL develops in two to three cell layers over the maternal vasc ular tissue. BETL specialized cel ls are characterized by elongated and angular cells with cell walls ingrowths at the BETL-pedicel region (f igure 3-8A and B). No abnormalities were observed in the BETL cell morphology or BETL2 expression for seeds collected from rgh3 segregating and normal ears at 10 DAP. At 20 DAP BETL2 can still be detected in normal seeds (Hueros et al., 1999) (f igure 3-8B). BETL2 was not detected in mutant seeds at 20 DAP (figure 3-8C and D). Interestingl y, the aleurone protease was detected in the BETL region in rgh3 mutant seeds at 20 DAP (figure 3-8F). Moreover, abnormal BETL cell morphology was observed in 20 DAP rgh3 mutant seed. The cells in the BETL position have irregular shapes and less cell wall ingrowths th an normal BETL cells (figure 3-8C, D and F). These data indicate that rgh3 disrupts normal development of the BETL.

PAGE 38

38 The Rgh3 Locus is Not Essential fo r Endosperm Cell Viability The data from the BETL2 immunolocalizat ion experiment indicated a role for Rgh3 in cellular differentiation. However, the reduced grai n-fill and lethal seed ling phenotypes suggest a possible role in general cell viability. Endosperm cell cultures were used to distinguish between an essential role in cel l viability and a role restricted to cellular differentiation functions. If Rgh3 is only required for cellular differentiation then it should be possible to generate mutant endosperm cell culture s. Conversely, if Rgh3 is required for general cell viability then only normal endosperm cultures will grow. Endosperm culture was done following the Shannon and Batey (1973) protocol, which is optim al for the A636 genetic background. rgh3-70 was crossed to A636 and F1 seeds were planted and self-pollinated. Seeds for endosperm cultures were collected from the F2. rgh3-70 heterozygous and homozygous norma l plants were identified by a PCR marker linked to the rgh3-70 allele (Chapter 4). Individual rgh3 mutant and normal endosperm tissues were colle cted at 10 and 11 DAP from rgh3 segregating, normal and A636 self-pollinated ears. The rgh3 homozygous seeds showed an early onset of the seed phenotype. Differentiation of mutants from normal siblings was determined by a reduced seed size and translucent appearance. A la rge percentage (49%) of rgh3 homozygous mutant endosperm cultures were able to grow compared to only 8% and 9% of normal si blings and A636 cultures respectively (table 3-1). No culture growth was observed from normal seeds from nonsegregating ears (table 3-1). The rgh3 mutant endosperm cultures showed a higher growth rate than normal siblings and A636 cultures (figure 3-9). These endosperm cultures were genotyped by PCR and confirmed to consist of rgh3/rgh3 homozygous mutants (data not shown). These results suggest that Rgh3 is not required for gene ral endosperm cell viability.

PAGE 39

39 Discussion One of the main functions of the BETL is to transfer solutes from the plant to the seed (reviewed in Becraft, 2001; Olsen, 2001; Thompson et al. 2001). An impaired BETL would produce significant changes to the flow of nutrients, resulting in pleiotropic effects on seed development (Cheng et al. 1996; Maitz et al., 20 00; Costa et al., 2003; Gu tierrez-Marcos et al., 2007). The rgh3 mutant is associated with specific deve lopmental seed defects, these defects are visible late in seed development at approxi mately 15 DAP. At this developmental stage, rgh3 mutant kernels are identifiable by a reduced si ze grain and defects in the epidermal aleurone layer. It is likely that these la ter seed defects are a consequen ce of the initial loss of the BETL identity some time after 10 DAP. Results from endosperm iodine staining suggest rgh3 starchy endosperm cells store starch properly, while the in vitro endosperm culture indicates that rgh3 is not needed for general cell viability. These results fu rther support the idea of an indirect effect of rgh3 on the starchy endosperm. Sim ilarly, ectopic aleurone in the rgh3 mutant appears to be the result of the starchy endosperm fragmentation later in development and not a direct effect of rgh3. Recent studies have shown that aleurone differe ntiation is induced in cells at the surface of the endosperm (Olsen 2004; Gruis et al., 2006; Ti an et al., 2007). Immunolocalization for the aleurone marker late in seed development (20 DAP) detected aleurone specific protease in cells at the surface of the fragmented areas of the endosperm. However, detection of the aleurone marker was not observed in earlier established aleu rone cells in the same mutant kernel. These observations reinforce the notion of rgh3 having an indirect effect on aleurone differentiation late in the seed development. Conversely, BETL defects appear to be a direct consequence of the rgh3 mutation. Immunolocalization and morphologica l analysis suggest that the B ETL is correctly specified at 10 DAP. The onset of the mutant seed phenotype at 15 DAP and immunoloca lization analyses at

PAGE 40

40 20 DAP indicate that the BETL loos es its cell identity after an initial correct specification. However, due to the difficulty of identifying homozygous rgh3 mutant seeds at 10 DAP, there is a small level of uncertainty of whether the de fects at 20 DAP are due to a failure in cell specification or a later lo ss of cell identity. Support for the la tter model can be found in the fact that 1/4 of the fifteen 10 DAP seeds analyzed ar e expected to be mutant (3-4 seeds), but no obvious mutant class was observed within the 10 DAP sample. Importantly, this sample is within a 98% confidence interval that at least one mutant was observed in 10 DAP sections. The rgh3 loss of cell identity is further supported by the lack of detection of BETL2 in the BETL region in the rgh3 kernels at 20DAP and the detection of the aleurone specific protease in the BETL region late in development. The endospe rm has full flexibility to respond to positional information by converting between starchy endosperm and aleurone cell fate late in endosperm development (Becraft and Asuncion-Crabb, 2000; Costa et al., 2003; Gruis et al., 2006). A similar model for the BETL may be valid since rgh3 mutant BETL cells appear to revert to aleurone cell fate. B-A translocation experiments on rgh3 indicated that under the rgh3 mutation condition, development of the embryo is influenced by the endosperm. This non-autono mous effect of the endosperm can be due to a restricted flow of solutes through the mutant endosperm. A similar non-autonomous effect of the endosperm over the embryo can be observed when examine the glo1 seed mutant B-A mosaics (Cos ta et al., 2003). Similar to rgh3, glo1 mutant seeds have aberrant BETL morphology and a defective embryo. The rgh3 and glo1 phenotypes suggest that the BETL could have an additional role in the normal development of the embryo. The rgf1 is another seed mutant id entified to have a BETL defect (Ma itz et al., 2000). In contrast, the rgf1 mutant fails to express the BETL markers, BETL1 and BETL2, yet it has a normal BETL

PAGE 41

41 morphology and develops a normal viable embryo. The rgf1 mutant seeds deve lop a significantly reduced starchy endosperm; however, rgf1 BETL defects do not disrupt free sugar concentrations in the rgf1 endosperm. Thus, it is possible th at sugar flux and/or free sugar concentrations are reduced in the rgh3 mutant endosperm due to a more severe BETL defect. A specific concentration of free sugars may be required in the endosperm or to be transported to the embryo, possibly as a signal, for th e normal development of the embryo. B-A mosaics also indicate that rgh3 has an autonomous function in the embryo. The rgh3 mutant embryos show a range of phenotypes, frequently failing to form any recognisable embryonic structures. However, a small fraction of rgh3 mutant embryos are able to form embryonic leaves and roots, produci ng a lethal seedling. All mutant seedlings died at the 2 or 3 leaf stage at around 22 days after emergence. At 12 to 14 DAP normal embryos had established a shoot apical meristem (SAM) and initiated embr yonic leaves (Becraft et al., 2001; Elster et al., 2000; reviewed in Bomm ert and Werr, 2001), the rgh3 seed phenotype onset at around 15 DAP likely disrupts embryonic morphogenesis at this time point. This suggests the rgh3 mutation disrupt embryo shoot and root meristem development, affecting seedling morphology and viability. Materials and Methods Light Microscopy Hand sections of m ature seeds we re obtained by imbibing mature dry rgh3 and normal seeds in water under vacuum overnight. Seed sections were cut with a sh arp razor blade cutting through the central longit udinal axis. Seed sections were ex amined either unstained or after staining with 50 mg/L of toludi ne blue or 20 mg/L iodine-potassium-iodine (IKI) solutions. Pictures of the seed section were taken usi ng a Wild Heerbrugs stereomicroscope with an

PAGE 42

42 attached digital CCD camera (model DC300, Leica Ba nnockburn, IL). Pictures were stored in a computer (model 8200, Dell). Germination and Seedlin g Phenotype Analysis Mutant and norm al kernels were grown under ty pical greenhouse conditions (20C to 27C, 16 hours of light/8 hours of dark ) using a commercial potting me dium (Metro-Mix 300; ScottsSierra). B73/W22 hybrids plan ts were obtain by crossing rgh3/+ plants in the W22 bacground as males to B73 wildtype females. Male plants were self pollinated to confirm its genotype. Immunolocalization Analysis Mutant and norm al kernels were collected fr om freshly harvested segregating and nonsegregating ears at 10 and 20 DAP. The top half of the cob was harvested and kernels were collected from it. The bottom half was left to mature in the plant and used to confirm the genotype of the plant. Kernels were fixed in a FAA solution containing 3.7% formaldehyde, 5% glacial acetic acid, 45% ethanol a nd 40% distilled water. Kernels were fixed at 4C overnight. Subsequently, kernels were dehydrated with in creasing concentrations of 45 %, 50% and 70% ethanol solutions at 4C for 30 minutes. Kernel samples were kept in 70% ethanol and sent to Gregorio Hueros to the Univ ersidad de Alcal in Spain fo r wax embedding, sectioning and immunolocalization analysis. Endosperm Tissue Culture Individual rgh3 m utant and normal endosperm tissues were collected at 10 and 11 DAP from rgh3 segregating, normal and A636 F2 self-pollinated ears. Following Shannon (1994) endosperm culture protocol, freshly harvested ears were surface sterilized by spraying the outside of the ear with 95% ethanol before removing the hus k. Once the husk was removed, the ear was immersed in a 20% bleach solution for 10 minutes. Bleached ears were rinsed twice with sterile water to eliminate traces of the bleach solution. The ear was place on sterile paper towels

PAGE 43

43 and the top of the kernels removed with a ster ile scalpel. Endosperm tissues from individual kernels were scooped out with a sterile spatula and placed in a tissue culture tube containing 4.43 g/L Murashige and Skoog salts (MS) with vita mins stock solution (no. M519; Phytotechnology laboratories), 30 g/L of sucrose, 10 mL/L of 100X Thiamine and 2 g/L of Asparagine. The pH of the MS solution was adjusted to 5.6 and 2 g/L of phytogel (n o P8169-500G; Sigma) was added. The mixture was heated to a boil to dissolve al l components and 9 mL of the mixture poured in individual culture tubes before being autoclav ed for 40 minutes. Endosperm cultures were grown in darkness at 30C. Endosperm callus was subc ultured to fresh medium every 3-4 weeks as needed. At the time of subculture, a sample of endosperm callus was taken for DNA extraction and genotype analysis.

PAGE 44

44 Figure 3-1. Mature rgh3 mutant kernels show a range of phenotypes. The rgh3 mutants are characterized by an etched, pitted, or crazed kernel surface. A) Shows self-pollinated rgh3/+ ear. White arrows indicated rgh3 seed mutants. B-F) Longitudinal hand sections of normal and mutant kernel s from a single segregating ear. C-F) rgh3 mutant seeds show a range in grain-fill and level of embryo development. Embryo defects correlate with kernel weight. C) The heaviest 10% of the mutant seeds develop near normal embryo morphology. D and E) Average weight rgh3 mutant seeds that can develop embryonic roots do not develop shoots. F) The lightest 10% of the mutant seeds only develop a small mass of embryonic tissue. Scale bars = 2.5mm.

PAGE 45

45 Figure 3-2. Seedling phenotypes of rgh3 mutants. A small fraction of rgh3-70 mutant seeds germinate in soil. The rgh3 seedlings grew more slowly than normal siblings, turning necrotic and died during the third week after germination. A-C) Pictures show rgh3 homozygous mutant and normal seedlings at 10 days after germination. A) rgh3 mutant seedlings in the W22 genetic background. B) rgh3 mutant seedlings in the B73/W22 hybrid genetic background C) normal seedling. Red bar = 2cm. A B C

PAGE 46

46 Figure 3-3. Root defects of rgh3 mutant seedlings. Mutant seedlings showed a short primary root (white arrows) with few or no secondary roots and some below ground short adventitious roots A) rgh3 mutant and normal seedlings in the W22 genetic background. B) rgh3 mutant and normal seedlings in the B73/W22 hybrid genetic background. Seedlings were collected 15 days after planting. Scale bars = 2cm.

PAGE 47

47 Figure 3-4. The rgh3 mutant seedlings show no apparent depletion of the seed endosperm reserve. Seed were sectioned from muta nt and normal seedlings at 18 days after planting. A) Normal seedlings showed a seed with a collapsed pericarp due to very little endosperm tissue left in the seed (wh ite arrow). B-D) Mutant seedlings showed remaining endosperm tissue in the seed (bl ack arrow). Scutellum = sc. Scale bars = 2.5mm.

PAGE 48

48 Figure 3-5. Aleurone defects in rgh3 mutant kernels. A) The al eurone tissue in maize is composed of a single layer of small cubodial cells. B-D) The rgh3 mutants show abnormal aleurone cell shapes, overprolifer ation of aleurone cells, as well as depressions of the aleurone cell layer in to the endosperm tissue. Pericarp = p, aleurone = al, and starchy endosperm = se.

PAGE 49

49 Figure 3-6. Starch accumula tion is not affected in rgh3 mutants. IKI stained normal (A), rgh3 (B), and vp5 (C) kernels. Starch stains black, and vp5 kernels show sectors of starchless cells similar to the et1 Pericarp = p, aleurone = al, and starchy endosperm = se.

PAGE 50

50 Figure 3-7. Delayed expression of an aleurone specific protease in rgh3. A) The aleurone marker shows little or no staining in normal kernels at 20 DAP (white a rrow). Inset shows the marker expression at 10 DAP (black arrow). B-C) The aleurone marker is express in rgh3 mutants at 20 DAP (black arrows). Marker expression is associated with cracks and lobes of the endosperm tissue. Pericarp = p, aleurone = al, and starchy endosperm = se.

PAGE 51

51 Figure 3-8. Basal endosperm transf er cell layer (BETL) defects in rgh3 mutant kernels. A) Normal 10 DAP Kernel showing expressi on of BETL2 protein (dark stain). B) Normal 20 DAP Kernel showing some expression of BETL2 late in seed development (black arrow). C-D) Abnormal BETL cell morphology was observed in rgh3 mutant seed at 20DAP. No BETL2 was detected in 20 DAP rgh3 kernels. E) Normal 20 DAP kernel shows no aleurone marker been detected at this time in development. F) Aleurone marker was dete cted in the transfer cell region of 20 DAP rgh3 kernels (white arrow). Peri carp = p, aleurone = al, ba sal endosperm transfer cell layer = betl, and starchy endosperm = se.

PAGE 52

52 Figure 3-9. The rgh3 locus is not essential for endospe rm cell viability. Data from BETL2 immunolocalization experiment suggests a role for Rgh3 in cellular differentiation. However, we tested a possible role of rgh3 in general cell grow th by in-vitro culture of rgh3 endosperm tissue. The rgh3/rgh3 endosperm cultures showed a higher growth rate than normal siblings and A636 inbred line. Picture shows 57 days old explants.

PAGE 53

53 Table 3-1. A high percentage of rgh3/ rgh3 endosperm explants were able to grow compared to normal siblings and A636 inbred. Phenotype Genetic background Number of endosperm explants Number explants that produce callus % of endosperm culture viability rgh3/rgh3 A636/W22 179 87 49 Normal siblings A636/W22 108 8 8 Homozygous normals A636/W22 36 0 0 A636 A636 144 13 9 Data were collected 57 days after initial culture.

PAGE 54

54 CHAPTER 4 MOLECULAR ANALYSIS OF rgh3 Introduction Mutant phenotypes provide inform ation about the biological function of genes. Mutants like rgh3, with visible phenotypes clearly have non-re dundant roles, but molecular information is essential to understand the mechanisms affect ed. Transposon mutagenesis has been used frequently for maize research to identify and clone genes. Activator ( Ac/Ds ) and Mutator ( MuDR / Mu ) are two transposable element families that are utilized in a large scale for functional genomics studies in maize: (reviewed in Settles, 2005). Mutator elements are characterized by a high rate of forward mutagenesis (Robert son, 1978; Walbot and Warren, 1988). These Mu insertions occur preferen tially in gene-rich regions (Cresse et al., 1995; Lunde et al., 2003). It is well documented that Mu insertion within or near gene s often alters th e expression or transcription of the gene (Lue hersen and Walbot, 1990; Ortiz and Strommer, 1990; Barkan and Martienssen, 1991; reviewed in Chandler and Hardeman 1992). A common approach for cloni ng transposon-tagged genes relies on correlating the inheritance of a plant phenotype with a band on a DNA hybridiz ation blot. To find a linked band, a number of restriction enzymes and Mu -specific hybridization probes are used to resolve the different classes of Mu elements (Mu1 Mu2 Mu3 Mu7, Mu8 and MuDR ). The cosegregating fragment is then clone d in a size-selected phage library and pla que purified with the Mu probe. Alternative methods have been developed ba sed on the polymerase chain reaction (PCR) to identify Mu insertion flanking regions. Inverse PCR, ad apter-mediated PCR and variations of the thermal asymmetric interlaced PCR (TAIL-PC R) are common PCR-based methods to recover transposon flanking sequences (Fret et al, 1998 ; Singh et al., 2003; Se ttles et al., 2004).

PAGE 55

55 The TAIL-PCR method was adapted and optimized by Settles and co-workers (2004) to amplify maize genomic sequences associated to Mu insertions sites (MuTAIL-PCR). MuTAILPCR uses a combination of nested high Tm primers specific for the terminal inverted repeats (TIR) sequences of the Mu transposon and 12 low Tm maize specific arbitrary primers. By interlacing PCR cycles with high and low annealing temperature, amplification of Mu -specific fragments is favored. MuTAIL-PCR is estima ted to amplify a complex pool of products representing ~80% of the transp oson insertions within the geno me (Settles et al., 2004). Based on sequencing studies (McCarty et al., 2005), a library of 384 Mu TAIL clones is estimated to contain ~70% of the total Mu insertions in the sampled genome. Thus, if the mutant is tagged, ~70% of the time the insertion causing the pheno type should be present in a MuTAIL library. These libraries also contain non-segregating Mu insertions common to the genetic background. These background insertions can be removed by in silico subtraction using the UniformMu database (Suzuki et al., 2006; Su zuki et al., 2008) or by sibling s ubtraction hybridi zations (Porch et al., 2006). I used a combination of DNA blot analysis and MuTAIL-PCR to identify and clone a Mu1 transposon insertion tightly linked to the rgh3 phenotype. This Mu1 insertion disrupts a predicted gene homologous to the human splicing factor U2 small nuclear ribonucleoprotein auxiliary factor-related protein (URP). Th e URP protein has been predicte d to be part of the pre-mRNA splicing complex (Tronchre et al., 1997). Pre-mRNA splicing factors have been show to have an important role in plant organ development a nd meristem identity (Veit et al., 1998; Wang and Brendel, 2006). Results The Rgh3 locus was initially m apped by B-A transl ocation to the long arm of chromosome 5. Two point linkage analysis positioned Rgh3 at 6.7 cM from pr1, which maps to the proximal

PAGE 56

56 end of genomic contig240. The rgh3 locus was further placed in the physicalgenetic map by SSR markers mmc0081 and umc1941. The rgh3 locus mapped at 1.4 cM from the mmc0081 marker (1 recombinant observed in a mapping popul ation of 68 individuals), which maps to the distal end of genomic contig238, and at 27 cM from umc1941 (12 recombinants observed in a mapping population of 43 individu als), which maps to the proximal end of contig250. This places rgh3 between the distal end of genomic con tig238 and the proximal end of contig240 in the physical map. Mu1 Insertion is Linked to the rgh 3 Phenotype A conventional Mu cosegregation method by DNA gel bl ot was used to identify a Mu1 insertion linked the rgh3 phenotype. The number of non-segregating Mu insertions was reduced by selecting Mu -inactive rgh3-70/+ individuals from the Un iformMu population and backcrossing (BC) them into the W22 inbred multiple times. W22 contains approximately 21 Mu insertions in comparis on with approximately 57 Mu insertions of typical UniformMu Mu inactive mutants (Settles et al., 2004; McCa rty et al., 2005). Only advanced backcross generations of rgh3-70 (BC2 or BC3) were used for molecular analysis. Southern blots, with 6 rgh3-70 heterozygous mutants and 5 homozygous normal plants digested with PstI, EcoRI, HindIII, SstI and X hoI restriction enzymes, were hybridized with specific Mu1, Mu3, Mu7 Mu8 and MuDR probes. Figure 4-1A shows a ~3.9 kb Mu1 /PstI specific fragment that segregates with the rgh3 mutant phenotype. This fragme nt showed tight linkage in an expanded segregating population of 72 he terozygous mutant and 54 homozygous normal plants from a BC2 population. These data give an upper bound of <0.8 cM for the map distance between rgh3 and the Mu1 element. To clone the linked Mu1 element, a PstI digestion of a rgh3-70 BC3 homozygous mutant seedling was run in a 1% agarose gel. A gel fragment corresponding to 3.5 to 4.2 kb fragments

PAGE 57

57 were excised. The DNA was purified from the gel fragment and used as a template for MuTAILPCR (Settles et al., 2004). Two sequential MuTAIL -PCR reactions were carried out with nested primers specific for the Mu1 internal and terminal invert ed repeat (TIR) sequences in combination with three maize specific arbitrary degenerate primers (fi gure 4-1B). MuTAIL-PCR products were cloned into a pCR4-TOPO vector for nucleotide sequence analysis. A library of 96 MuTAIL-PCR clones was sequenced. 40% of th e sequenced reads grouped in 5 different contigs. A 898 bp contig had the largest number of reads (13) of the 5 contig groups. Sequence analysis of the 898 bp contig confirmed the exp ected 55 bases at one end corresponded to the Mu1 TIR sequence. This Mu1 flanking sequence tag (FST) was used as a probe for the PstI digested rgh3 segregating DNA blots (figure 4-1C). Th e FST probe hybridized to the 3.9 kb cosegregating fragment. To further map and obtain the promoter sequence of the Mu1 -tagged locus, the FST was used as a probe to screen the Childrens Hosp ital Oakland Research Institute (CHORI) maize BAC library (Young-Sun Yim et al., 2002). The FST identifie d four maize BACs (CHORI201351B17, CHORI201-376K17, CHORI201-497J22 and C HORI201-302O16) in the proximal end of genomic contig240 in the maize physical map. S outhern blot analysis of the four maize BACs was used to identify a suitable fragme nt for subcloning and sequencing of the Rgh3-linked locus. All 4 BACs were digested with restriction enzy mes BamHI, BsrI, ClaI, EcoRI, SstII and XbaI and hybridized with the FST probe. A BamHI fragme nt of 9 Kb was identified to subclone. Due to the size of the subclone (9 kb), sequencing of this subclone should identify the promoter region of the locus. To expand the Mu1 rgh3 linkage analysis, two PCR primer s specific for the FST were used in combination with the Mu1 -TIR specific primer for a co-d ominant PCR assay (figure 4-2).

PAGE 58

58 Linkage analysis of the rgh3 Mu1 -tag was expanded to 550 meiotic products with no recombinants observed. These data reduced, the upper bound for the map distance between the Mu1 -tag and rgh3 to <0.18 cM. This result indicates that the Mu1 -tag is tightly linked with the rgh3 seed phenotype and is a strong candidate for the Rgh3 locus. Nucleotide BLAST (BLASTN) searches with the 898 bp contig sequence against the Maize Assembled Genomic Island (MAGI), National Center for Biotechnology Information (NCBI) and Gramene data bases identified tw o maize Genomic Survey Sequences (GSS), accesions BZ738415 and BZ738412, and a predicted rice gene Os02g35150. BLASTN searches with the rice Os 02g35150 sequence identified additional maize expressed sequence tags (ESTs) and a truncated maize cDNA (t able 4-1). PCR primers were designed to close the gap betw een the FST and maize ESTs. Genomic PCR products connected the FST to the maize ESTs, suggesting that th e predicted Os02g35150 rice gene is homologous to the Mu1 -tag. The maize ESTs were isolated fro m 7-23 DAP endosperm tissue (Lai et al., 2004). Thus, the expression of the Mu1 -tagged locus appears to be consistent with the rgh3 phenotype. Mu1Tagged Locus Show s Alternative Initiation of Transcription and Alternative Splicing To obtain the full-length Mu1tagged cDNA, a 5' and 3' Rapid Amplification of cDNA Ends (RACE) PCR were conducted using th e GeneRACER Kit (# L1502-01 Invitrogen, Carlsbad, CA USA) in combination with FST and cDNA BT023976 specific primers. The 5' RACE-PCR identified two transcri ption initiation site s. Alignment of the two 5' RACE products with the genomic sequence indicated that one of the 5' RACE-PCR products does not contain the first predicted exon and the transcription ini tiates 75 bases into th e predicted second exon. Reverse transcriptase (RT) P CR was performed with two primers designed from the 5' and 3' RACE products to obtain the full-length c DNA. Multiple RT-PCR products, ranging from 2.2

PAGE 59

59 to 2.7 kb, were obtained indicating a lternative splice varian ts of the transcript (figure 4-3A). Due to the size of the products, it was difficult to identify the number of alternative transcripts by agarose gel electrophoresis. RT-PCR primers we re designed to amplify 588 bases of the transcript internal region. RT-P CR products visualized on a 1% ag arose gel suggest five possible alternatives splice variants of the gene (figure 43B). Partial sequence analysis of the alternative splice variants indicated that most variants introduce stop codons with in the predicted open reading frame (ORF) (data not shown). A 2,515 base pair (bp) cDNA clone with a predicted 2,268 bp ORF was identified. A 9 exons/8 introns architecture of the maize gene was elucidated by computer alignment of the cDNA containing the ORF sequence and the genomic sequence. Alignment of the cDNA and the FST sequence identified the Mu1 insertion site within the co ding region of the first exon, 342 nucleotides downstream of the 5 end (Figure 4-4). Analysis of the ORF shows that the maize gene encodes a predicted protein of 755 amino acids (figure 4-4B). The 5 untranslated terminal region (UTR) is 17 bp while the 3UTR is 207 bp. Mu1Tagged Gene Encodes a Predicted Splicing Factor Protein BL AST (BLASTP) searches of the predicted maize and rice Os02g35150 proteins against the NCBI protein database revealed a high level of sim ilarity with a region that is conserved with the small subunit of human U2 small ribonucleoprotein aux iliary factor-related protein (HsURP), mouse U2AF1-RS2 and Arabidopsis U2AF35-related protein (AtURP) (table 4-2). A higher degree of sequen ce identity was found between the maize predicted protein and the human URP (49%) than between the maize protein and the closes t human U2AF35 protein (39%). In addition, both the maize and HsURP pr otein have acidic domains in the N-terminus, which is not found in U2AF35 proteins. These sequence similarities suggest that the Rgh3 linked

PAGE 60

60 locus is an ortholog of hu man URP. Consequently, the Mu1 -tagged locus and rice Os02g35150 were designated the ZmUrp and OsUrp respectively. Molecular Lesion of the z murp Lo cus It has been well documente d that the insertion of Mu transposable elements within or near genes often alters the transcrip tion of the target gene (reviewed in Chandler and Hardeman 1992). ZmUrp expression was analyzed by RT-PCR in both normal and rgh3 mutant tissues. Total RNA was collected from normal and muta nt 18 DAP seeds, seedling leaf and seedling root. The RT-PCR was performed with a primer de signed for the 5 UTR in combination with a primer that bridges exons four and five. An expected 859 bp RT-PCR fragment was amplified in all normal tissues, however a 997 bp fragment was amplified in all mutant tissues (figure 4-6A). Sequence comparison of the mutant and normal RT-PCR products revealed that the mutant transcript contains 141 bases of the Mu1 TIR sequence 334 bases downstr eam of the start of the ORF in exon 1 (figure 4-6C). Analysis of the mutant RT-PCR product also revealed a 44 nucleotide deletion of the ORF seque nce flanking the 3 end of the Mu1 element in the first exon. This polmorphism introduces a novel ORF of only 297 bases that does not have sequence identity with ZmURP. Mu1 elements have been shown to alter mRNA processing previously (Ortiz and Strommer, 1990). In the study, they identified a donor splice site in the 5 TIR of Mu1 The zmurp RT-PCR products suggest a similar mechanism for the production of noncoding ZmURP transcripts in rgh3 mutants. The zmurp transcript is produced from splicing a donor site 141 nucleotides downstream of the 5 end of the Mu1 TIR to the zmurp second exon acceptor site, effectively excising most of the Mu1 element plus 44 nucleotides of the 3 end of exon 1 (figure 4-6B).

PAGE 61

61 Discussion The data presented here provide a strong argument that the rgh3 mutation resulted from a Mu1 transposon insertion in the ZmUrp locus. The Mu1 -tagged zmurp allele is tightly linked to the rgh3 mutant phenotype. Cosegreg ation analysis placed the rgh3 mutant <0.18 cM from the zmurp Through nucleotide BLAST searches two ZmUrp ESTs were identified. The ESTs are expressed in the endosperm tissue between 7 and 23 DAP (Lai et al., 2004). This is consistent with the tissue and onset of the rgh3 phenotype at 15 DAP. Also, the rgh3 mutation resulted in altered ZmURP transcription. rgh3 mutants show a chimeric ZmUrp Mu1 transcript due to a donor splice site inside the Mu1 element. The Mu1 element has distinct structural features which includes an unusually long termin al inverted repeats (TIRs) and internal GC-rich regions surrounding a central AT-rich core (reviewed in Chandler and Hardeman, 1992). The GC to ATrich transition region is considered necesary fo r the splicing of introns (Ortiz and Strommer, 1990; Luehrsen and Walbot, 1990; reviewed in Lorkovic et al., 2000; reviewed in Reddy, 2007). Thus, the GC to AT-rich region inside the Mu1 element may be prefer entialy spliced in the mutant allele producing the chimeric ZmUrp Mu1 trasncript. Mu1 influencing alternative splice behavior has been previously desc ribe by Ortiz and Strommer (1990) Amino acid sequence comparison between UHM domains of maize, rice, Arabidopsis and human URP proteins showed high degree of am ino acid identity (93% to 49%). However, aligment of the full length protein sequences showed a much lower amino acid identity level between all URP proteins (61% to 43%). These lower identity levels are due to the high variability in the amino acid sequence of the aci dic N-terminal and RS-rich C-terminal domains. For example, RS domain amino acid identity be tween the maize and rice, Arabidpsis and human URP proteins are 47%, 27% and 30% respectively.

PAGE 62

62 Tronchre and co-workers (1997) showed that the URP RS domain is required for protein protein interaction between URP and SR proteins. Vari ations in the URP RS domain suggest interactions with specific SR-proteins. Moreove r, alternative splicing of some SR genes is controlled in a developmental and tissue-specifi c manner (Gao et al., 2004; reviewed in Reddy 2007). Thus, URP proteins may interact with a certain group of SR pr oteins and pre-mRNAs under specific tissue and/or developmental stage. Pre-mRNA splicing is known to be important in multiple developemtal processes (Veit et al ., 1998; Quesada et al., 2003; Wang and Brendel, 2006; Chung et al., 2007). These observations corre late with the predicte d tissue and temporal onset of rgh3. Materials and Methods Genomic DNA and Total RNA Extraction Genom ic DNA was extracted from seedling a nd adult leaf sample s by grinding 2 g of tissue in 5 mL of urea DNA extraction buffer (168 g urea, 25 mL of 5 M NaCl, 20 mL of 1 M Tris-HCl pH 8, 16 mL of 0.5 M EDTA pH 8, 20 mL of sarkosin e and 190 mL of H2O). The extract was mixed with phenol:chloroform:isoamy l alcohol (25:24:1) fo r 15 minutes and the phases separated by centrifugati on (10 minutes at 2000g). The aqueous phase was transferred and mixed with 500 mL of 3 M sodium acetate pH 5.2 and 4 mL isopropanol. The precipitated DNA was washed with 70% ethanol and resusp ended in TE (10 mM Tris-HCl pH 8, 1 mM EDTA pH 8). Total RNA was isolated from 100 mg of tissue with TRIzol reagent (# 15596-026, Invitrogen, Carlsbad, CA, USA) according to the manufacturers instructions. After precipitation, RNA pellets were resuspende d in 50 mL of ribonuclease-free water.

PAGE 63

63 Southern Blot Analysis Approximately 30 g of genomic DNA was restri ction digested in sepa rate reactions with 20 U of the restriction enzyme in a 100 L fina l volume. The digested DNA samples were size fractionated on 0.8% TBE agarose gels a nd blotted to Hybond-N nylon membranes (# RPN303N, Amersham, Piscataway, NJ, USA) usi ng standard capillary transfer conditions. Hybridization probes were prep ared by gel purification of Mu elements and Mu1 -FST fragment. The Mu1 probe was the 650 bp EcoRI/HindIII fragment from pA/B5 (Chandler et al., 1986). The probe fragments were purified by agarose electro phoresis and extracted with Qiagen purification kits (#28704 Qiagen,Valencia, CA, USA) fo llowing the manufacturer's instructions. The fragments were labeled with Amersham Rea dy-to-Go Labelling Beads (-dCTP) (# 27-9240-01, Amersham, Piscataway, NJ, USA) following the manufacturer's instructions. The blots were hybridized with Church hybr idization buffer (1% BSA, 1 mM EDTA pH 8.0, 0.5 M Na2HPO4 pH 7.2 and 7% SDS) overnight at 65C and wa shed once with 2X SSC, 0.1% SDS, and three times with 0.2X SSC, 0.1% SDS at 65C for 20 minutes each wash. MuTAIL-PCR DNA seque nce flanking the Mu element was amplified by MuTAIL-PCR (Settles et al., 2004). The primary MuTAIL-PCR reaction (20 L) contained 1X PCR buffe r (20 mM Tris-HCl pH 8.4, 50 mM KCl and 2 mM MgCl2), 2 mM MgCl2, 200 M of each dNTPs, 1 U of Taq DNA polymerase, 100 ng of genomic DNA, 100 nM of the Mu -specific primer Mu1iPCRrev (5GGTAAACGGGGACAGAAAAC-3), and 1 M of each arbitrary degenerate primers; CGGC1 (5-GWSIBCTANGAGGGC-3), DMR-AG1 (5-GNGWSASTNGAGC-3) and Geeky1 (5GKYKGCKGCNGC-3). The primary reaction was d iluted 1:50 in water and 1 L was added to a 20 L secondary reaction. Water-only negative c ontrols from the primary reaction were diluted and processed identically to reactions cont aining DNA templates. The secondary MuTAIL-PCR

PAGE 64

64 reaction was identical to the primary except the Mu -specific primer TIR8.2 (5CGCCTCCATTTCGTCGAATCCSCTT-3) was used instead. Arbitrary degenerate primers were the same as used for primery MuTAIL-PCR reaction. The primary and secondary MuTAIL-PCR reactions were incubated accordi ng to the method of Settles et al. (2004). MuTAIL-PCR products were size selected using a Sephacryl-400 (# V3181, Promega, Madison, WI, USA) spin column following the manufacturer's instructions. This column reduced the concentration of products sm aller than 500 bp. The size-selec ted products were then cloned into the pCR4-TOPO vector (# 45-0030, Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. Ninety six colonies were randomly selected for sequencing at the UF-ICBR genome core facility. Polymerase Chain Reaction Linkage Analysis Genom ic DNA from mutant and normal seedlin gs was extracted as described above. The Mu1tagged allele was amplified with primer pairs TIR8.2 (5CGCCTCCATTTCGTCGAATCCSCTT-3) and MAGI1713rev3 (5ACCTTGTTTCGAGCGGAAGCCT-3). The normal a llele was amplified with primer pairs MAGI1713fwd2 (5-TTCGGCAAACAGAGGG CTCA-3) and MAGI1713rev3 (5ACCTTGTTTCGAGCGGAAGCCT-3). PCR was completed with Taq DNA polymerase and buffers from Invitrogen (#18067-017, Invitrogen, Carlsbad, CA, USA) with 5% DMSO added to the final reaction. Incubation cond itions were 94C for 1 minute, 62C for 1 minute, 72C for 1 minute for 40 cycles with a 10 minutes extension at 72C. Amplification of cDNA Ends Total RNA was isolated from homozygous 18 DAP seeds. The GeneRACER Kit (# L1502-01 Invitrogen, Carlsbad, CA USA) was used to amplify ZmUrp 5 and 3 ends. RACE was performed as described by manufacturer's instructions using the following primers:

PAGE 65

65 DFB028R (5-TGGCCATCTTTAGGCTTGTT-3) for 5 RACE and (5CAAGCCTAAAGATGGCCAAG-3) for 3 RACE. RACE-PCR products were then cloned into the pCR4-TOPO vector for sequencing. Molecular Analysis of Normal and Mutant Z mURP transcripts ZmUrp cDNA products were obtain by means of reverse transcriptase PCR (RT-PCR). RT-PCR was performed using SuperScript First-Strand Synthesis System (#11904-018, Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. The full length cDNA was obtain using the following primer pair s; DFB043F (5-ATCGA CGAGCTCGCCATGT-3) and DFB033R (5-TGCGAAGTTAAACCGATTCC-3). The ZmUrp internal region was obtained with primer pairs; DFB043F (5-ATCGACGAGCTCGCCATGT-3) and EXONrev12 (5-ACATTCCCTCGAAGATGGAAAGACC-3). ZmURP transcript expression was analyzed in normal and rgh3 mutant tissues with primer pairs; DFB043F (5ATCGACGAGCTCGCCATGT-3) and RFrgh3bridgerev1 (5TCTTCATCTGTGAACTCAAGCCC-3). Maize ubiquitin primer sequences were obtain from Chung et al. (2007) and are as follows; ZmUBCF1 (5AAGATGCAGGCATCTAGGGCAAGG-3) and ZmUBCR1 (5AGGCTCTTGGCTTGGCACATGTTC-3).

PAGE 66

66 Figure 4-1. Cosegregation analysis wa s used to identify and clone a Mu1 insertion linked to rgh3. A) A survey of restriction enzymes and Mu elements as probe, identified a 3.9 kb Mu1 PstI digested fragment linked to rgh3 phenotype (white arrow). B) Schematic representation of MuTAIL-PCR. rgh3 mutant DNA was PstI digested and a gel piece predicted to containing the Mu1 cosegregating fragment was excised and used as a template to amplify and clone the Mu1 -flanking sequence (FST) using MuTAILPCR. C) The FST was hybridized to PstI di gested genomic blot of the segregating rgh3-70 population. The probe iden tified a 3.9 kb fragment that cosegregtes with the rgh3 mutant phenotype (white arrow). Black arrows show the normal allele of the tagged locus.

PAGE 67

67 Figure 4-2. Expanded linkage analysis between the Mu1 -tag and the rgh3 phenotype. A) Schematic of codominant P CR marker. Left and right Mu1 -tagged specific primers were designed on either side of the inse rtion (black arrows) to amplify the normal allele, whereas another PCR reacti on using primers specific for the Mu TIR and the flanking sequence (FST), produced a band specific for the mutant allele (white arrow). B and C) Cosegregation analysis with self-progeny. Normal sibling samples included heterozygous and homozygous normal plant. B) Agarose gel shows the mutant PCR product (wh ite arrow) in all rgh3/rgh3 individuals. C) Agarose gel shows the normal allele PCR product in all normal sibblings (black arrow). Molecular marker 1 Kb ladder = M.

PAGE 68

68 Table 4-1. Nucleotide BLAST search of th e rice Os02g35150 predicted gene identified two maize ESTs and one maize truncated cDNA. Query Subject size (bp) Identity (%) Tissue E-value EST CD435189 815 80 7-20 DAP endosperm 0 EST CD448595 644 84 7-20 DAP endosperm 0 Os02g35150 cDNA BT023976 845 78 7-20 DAP endosperm 4e-154

PAGE 69

69 Figure 4-3. Alternative splic ing of the ZmURP transcript. A) Multiple full lengh ZmUrp RTPCR products were obtained indicating the po ssibility of altern ative forms of the ZmUrp transcript (black arrows). Due to th e size of the products, it is difficult to identify the number of alternative transc ripts by 1% agarose gel electrophoresis. B) RT-PCR of the internal ZmUrp transcript, targeting e xon 1 to 6. RT-PCR products visualized on a 1% agarose gel suggest 6 possible variants of the Mu1tagged transcript (black arrows).

PAGE 70

70 Figure 4-4. Structure of the ZmUrp gene. A) Exon/intron structure of the ZmUrp gene (not to scale). Black boxes correspond to the 5 a nd 3 UTRs and white boxes to the coding region. Bars between the boxes represent in trons. Black triangle with white dots represents the Mu1 element (not to scale) inse rted in the ORF of exon 1. B) Nucleotide and translated amino acid sequence of ZmUrp. Black triangle indicates the Mu1 element insertion site. Black arrow h ead indicates altern ative initiation of transcription site. Gray a rrow heads indicate stop codons. A Mu1 B

PAGE 71

71 Table 4-2. Protein BLAST searches results of the Mu1 -tagged predicted protein. Query Subject G.I. size (a.a.) Identity (%) UHM identity (%) OsURP/Os02g35150 125539895 678 61 93 AtURP/AtU2AF-R 15218489 757 60 84 HsURP/U2AF1-RS2 4827046 482 43 49 Mu1Tagged protein U2AF35 267187 240 40 39 Size and identity columns refer to the full length protein.

PAGE 72

72 Figure 4-5. Sequence alignment of UHM domain s of ZmURP, homologous URPs and U2AF35 protein. Black shading shows identical amino acid residues while gray shading indicates conserved amino acid resi dues. ZmURP-UHM, HsURP-UHM, AtURPUHM, OsURP-UHM and U2AF35-UHM correspond to maize, human, Arabidopsis rice and human proteins, respectively.

PAGE 73

73 Figure 4-6. Molecula r lesion of the zmurp locus. A) ZmURP transcript expression was analyzed by semiquantitative RT-PCR in both normal and rgh3 mutant tissues. RT-PCR was performed with a primer flanking the Mu1 insertion (top gel). An expected 859 bp RT-PCR fragment was amplifie d in only normal tissues (black arrow), however a 997 bp fragment containing 141 nucleotides of the Mu1 -TIR sequence was observed in only mutant tissues (white arrow). Bottom gel shows Ubc control gene. Lane 1, 18 DAP embryo; lane 2, 18 DAP endosperm; lane 3, seedling leaf; lane 4, seedling roots; lane 5, 18 DAP rgh3 mutant seed; lane 6, rgh3 mutant seedling leaf; lane 7, rgh3 mutant seedling root. B) Sche me of alternative spliced zmurp mutant transcript. Only exons 1 to 4 are represented in the diagram (open boxes). Mu1 element is represented by the gray box with the TIRs represente d by the stripe pentagons. Alternative splicing event is indicated by dash line. C) Nucleotide and translated amino acid sequence of mutant ZmURP tras ncript (include c oding region of exons 1 to 4). Boxed sequence belong to the Mu1 -TIR. Gray arrows indicate stopping codons.

PAGE 74

74 CHAPTER 5 CONCLUSIONS Maize seed developm ent is a complex pro cess that involves the specification and differentiation of multiple tissu es in a highly coordinated manner. As products of double fertilization the endosperm and embryo follow di fferent developmental pathways. Synchronized development of the endosperm and embryo sugge sts crosstalk between them upon fertilization (reviewed in Ohad 2007) and through the early stag es of seed development. B-A translocations provide a genetic tool to study these interactions. Analysis of B-A mosaics of a number of dek mutants indicates that the majority of the Dek loci have an aut onomous function in the endosperm or embryo (Neuffer and Sheridan, 19 80; Neuffer et al., 1986; Chang and Neuffer, 1994). Conversely, discolored-1 ( dsc1 ) and globby1 ( glo1) are the only two mutants to show non-autonomous developmental phe notypes in B-A mosaics. Mosaics of these loci show that mutant endosperm tissue can cause a genetically normal embryo to develop as a mutant (Scanlon et al., 1998; Costa et al., 2003). Both of these genes show abnormal cellularization early in endosperm development leading to alterations in cell fate specification and subsequent abnormal differentiation of endosperm and embryo tissues. Thus, complete endosperm cellularization is essential for embryo development. In contrast to dsc1 and glo1 the rough endosperm3 ( rgh3) mutant endosperm appears to cellularize norma lly yet stills shows a non-autonomous function. The late onset of the rgh3 mutation (after 10 DAP) suppor ts the conclusion that Rgh3 function is not primarily involved in cellulari zation of the syncytia. However, Rgh3 could be required for an essential function that is lethal when lost in the endosperm. The succe ssful tissue culture of mutant rgh3 endosperms indicates that the gene is not required for general endosperm cell viability. Instead, phenotypic and immunol ocalization data in dicate a role for Rgh3 in maintaining BETL cell identity late in seed develo pment. This result conflicts with a previous

PAGE 75

75 model for BETL differentiation, where it was propos ed that BETL cell fate is irreversible and occurs within a narrow developmental window (Costa et al., 2003). Thus, the rgh3 mutant reveals the presence of a novel developmental process required later to maintain BETL cell identity. A similar model is observed in starc hy endosperm and aleurone cell fate. Starchy endosperm cells have full flexibility to res pond to positional information by converting between starchy endosperm and aleurone cell fate (Becraft and AsuncionCrabb, 2000; Gruis et al., 2006). An impaired BETL would produce significant change s to the flow of solutes through the mutant endosperm, potentially causing the non-autonomous rgh3 endosperm defect. However, the reduce grain filling1 ( rgf1) mutant suggest that nutrient flow is not the primary mechanism for endosperm-embryo development interactions. The rgf1 mutant is characterized by a reduction of specifically expressed BETL genes, yet it ha s a normal BETL morphology and develops a normal viable embryo. The BETL is also known to produce multiple secreted peptides, some of these are likely to be involved in embryo defense roles, while others are predicted to have intercellular signaling roles (Hueros et al., 1995; Muiz et al., 2006). I speculate that some aspects of the loss of BETL gene expression leads to deffect s in endosperm-embryo interact ions. Moreover, the endosperm surrounding region (ESR) is known to produce a numb er of secreted peptides (Opsahl-Ferstad, 1997; Baladin et al., 2005). ESR cel ls may also be affected in rgh3 mutants and ESR-specific genes could be the underlying cause of the development interactions. However, I have not yet tested rgh3 for defects in ESR gene expression. Analysis of rgh3 lethal seedlings also s uggests a specific role of Rgh3 in embryo shoot and root development. A small fraction of rgh3 mutants produce viable seeds but are seedling lethal at the 2 or 3 leaf stage. The bulk of rgh3 mutants fail to produce seeds with embryonic shoot.

PAGE 76

76 The shoot apical meristem (SAM) initiates embryoni c leaves at about 12 to 14 DAP (reviewed in Bommert and Werr, 2001), the onset of the rgh3 phenotype is around the same time suggesting Rgh3 may have a role in meristem initiation or maintenance. The rgh3 locus is most likely mutated in a splicing factor homologous to the human U2 small nuclear ribonucleoprotein auxiliary factor-rel ated protein (URP) The zmurp mutation is tightly-linked to rgh3. ZmURP is normally expressed in endosperm tissues. The zmurp mutation produces a transcript that truncates the ZmURP protein. Finally, ZmURP is a single copy gene suggesting a mutation in ZmURP is more likely to give a phenotype. Based on phylogenetic analysis (Mollet et al., 2006) URP seems to play a non-redundant splicing func tion. There are several precedents for splicing factors having roles in plant development (Que sada et al., 2005; Wang and Brendel, 2006; Chung et al., 2007). A maize locus of particular relevance is the terminal ear1 ( te1 ) mutant. TE1 is required for initiation and development of the leaves and encodes a putative RNA-binding protein and regulates differential spli cing of pre-mRNAs (Veit et al., 1998). Pre-mRNA splicing provides a means of regulat ing gene expression. Splicing factors have been shown to have an important role in plan t development. Disruption of two U2AF35 isoforms in Arabidopsis causes mutant plants to be late flowering as well as changes in flower and leaf morphology (Wang and Brendel, 2006). Moreover, the maize TE1 and zmsmu2 are two loci involve in leaf initiation and endosperm composition respectively, that encode proteins involved in differential RNA splicing (Vei t et al., 1998; Chung et al., 2007). Thus, my data are consistent with a model in which maize Urp ( ZmUrp ) has a regulatory role in pre-mRNA splicing and is required to develop embryonic shoots and roots as well as to maintain BETL cellular id entity. The distinctive rgh3 mutant phenotype and its predicted molecular function provide further insight in to the factors governing endosperm-embryo

PAGE 77

77 interaction and endosperm tissue differentiati on. An understanding of endosperm developmental processes could provide a basis fo r more efficient approaches to seed improvement and serve as a model for organ development in maize.

PAGE 78

78 LIST OF REFERENCES Balandn M, Royo J, Gmez E, Muniz LM, Molina A, Hueros G (2005) A protective role for the embryo surrounding region of the mai ze endosperm, as evidenced by the characterisation of ZmESR-6, a defensin gene specifically expressed in this region. Plant Mol Biol 58 (2): 269-82 Barkan A, Martienssen RA (1991) Inactivation of maize transposon Mu suppresses a mutant phenotype by activating an outward-reading pr omoter near the end of Mu1. Proc Natl Acad Sci U S A 88(8): 3502-6 Becket JB (1996a) Locating recessive genes to chromo some arm with B-A translocations. The maize handbook. M. Freeling and V Walbot, eds. Springer-Verlag. New York. pp. 313327 Becket JB (1996b) Comprehensive list of B-A tran slocations in maize. The maize handbook. M. Freeling and V Walbot, eds. Springer-Verlag. New York. pp. 336-341 Becraft PW (2001) Cell fate specifi cation in the cereal endospe rm. Semin Cell Dev Biol. 12(5):387-94.Becraft PW, Asuncion-Crabb Y. (2000) Positional cues specify and maintain aleurone cell fate in maize endosperm development. Development 127(18): 4039-48 Becraft PW (2002) Receptor kinase si gnaling in plant development. Annu Rev Cell Dev Biol 18: 163-92 Becraft PW, Asuncion-Crabb Y (2000) Positional cues specify and maintain aleurone cell fate in maize endosperm development. Development 127(18): 4039-48 Becraft PW, Stinard PS, McCarty DR (1996) CRINKLY4: A TNFR -like receptor kinase involved in maize epidermal development. Science 273: 1406-1409 Bensen RJ, Johal GS, Crane VC, Tossberg JT, Schnable PS, Meeley RB, Briggs SP (1995) Cloning and characterization of the maize An1 gene. Plant Cell 7 (1): 75-84 Berger F (2003) Endosperm: the crossroad of seed development. Curr Opin Plant Biol 6(1): 4250 Birchler JA, Alfenito MR (1993) Marker systems for B-A tran slocations in maize. Heredity 84 : 135-138 Bommert P, Werr W (2001) Gene expression patterns in the maize caryopsis: clues to decisions in embryo and endosperm development. Gene 271 (2): 131-42 Boyer CD, Hannah LC (2001) Kernel mutants in corn. In Specialty Corns. Ed. AR Hallauer. CRC Press, Boca Raton, Florida. pp. 1-32 Brutnell TP (2002) Transposon tagging in maize. Funct Integr Genomics 2(1-2): 4-12 Carlson WR (1978) The B chromosome of corn. Annu Rev Genet 12: 5-23 Carson C, Robertson J, Coe E (2004) High-volume mapping of maize mutants with simple sequence repeat marker s. Plant Mol Bio Rep 22(2): 131-143

PAGE 79

79 Chan FK, Chun HJ, Zheng L, Si egel RM, Bui KL, Lenardo MJ (2000) A domain in TNF receptors that mediates ligand-independent receptor assembly and signaling. Science 288(5475): 2351-4 Chan MT, Neuffer MG (1989) A simple method for staining nuclei of mature and germinated maize pollen. Stain Technol 64(4): 181-4 Chandler V, Hardeman KJ (1992) The Mu elements of Zea mays. Adv Genet 30: 77-122 Chandler V, Rivin C, Walbot V (1986) Stable non-mutator stocks of maize have sequences homologous to the Mu1 trans posable element. Genetics 114: 1007-1021 Chaudhury AM, Ming L, Miller C, Craig S, Dennis ES, Peacock WJ (1997) Fertilizationindependent seed development in Arabidopsis thaliana Proc Natl Acad Sci USA 94(8): 4223-4228 Cheng WH, Taliercio EW, Chourey PS (1994) The Miniature1 Seed Locus of Maize Encodes a Cell Wall Invertase Required for Normal Development of Endosperm and Maternal Cells in the Pedicel. Plant Cell 8 (6): 971-983 Chomet PS (1996) Transposon tagging with mutator. The maize handbook. M. Freeling and V Walbot, eds. Springer-Ve rlag. New York. pp. 313-327 Chung T, Kim CS, Nguyen HN, Meeley RB, Larkins BA (2007) The maize zmsmu2 gene encodes a putative RNA-splicing factor that affects protei n synthesis and RNA processing during endosperm development. Plant Physiol 144 (2): 821-35 Costa LM, Gutierrez-Marcos JF, Brutnell TP, Greenland AJ, Dickinson HG (2003) The globby1-1 (glo1-1) mutation disrupts nuclear and cell division in the developing maize seed causing alterations in endosperm cell fate and tissue differentiation. Development 130: 5009-5017 Cresse AD, Hulbert SH, Brown WE, Lucas JR, Bennetzen JL (1995) Mu1-related transposable elements of maize preferentially insert into low copy number DNA. Genetics 140 (1): 315-24 da Costa e Silva O, Lorbiecke R, Garg P, Muller L, Wassmann M, Lauert P, Scanlon MJ, Hsia AP, Schnable PS, Krupinska K, Wienand U (2004) The Etched1 gene of Zea mays (L.) encodes a zinc ribbon protein that belongs to the tran scriptionally active chromosome (TAC) of plastids and is simila r to the transcription factor TFIIS. Plant J 38(6): 923-39 Frey M, Stettner C, Gierl A (1998) A general method for gene isolation in tagg ing approaches: amplification of insertion mutageni sed sites (AIMS). The Plant Journal 13(5): 717 Gao H, Gordon-Kamm WJ, Lyznik LA (2004) ASF/SF2-like maize pre-mRNA splicing factors affect splice site utilization and thei r transcripts are alternatively spliced. Gene 339: 25-37 Gehring M, Choi Y, Fischer RL (2004) Imprinting and seed development. Plant Cell 16 Suppl:S203-13

PAGE 80

80 Gomez E, Royo J, Guo Y, Thompson R, Hueros G (2002) Establishment of cereal endosperm expression domains: identification and prope rties of a maize tr ansfer cell-specific transcription factor, ZmMRP-1. Plant Cell 14 (3): 599-610 Grossniklaus U, Vielle-Calzada JP, Hoeppner MA, Gagliano WB (1998) Maternal control of embryogenesis by MEDEA, a polycomb group gene in Arabidopsis Science 280(5362): 446-450 Gruis DF, Guo H, Selinger D, Tian Q, Olsen OA (2006) Surface position, not signaling from surrounding maternal tissues, specifies aleuro ne epidermal cell fate in maize. Plant Physiol 141(3): 898-909 Gutirrez-Marcos JF, Dal Pr M, Giulini A, Co sta LM, Gavazzi G, Cordelier S, Sellam O, Tatout C, Paul W, Perez P, Dickinson HG, Consonni G (2007) empty pericarp4 encodes a mitochondrion-targeted pentatricope ptide repeat protein necessary for seed development and plant growth in maize. Plant Cell 19(1): 196-210 Hueros G, Royo J, Maitz M, Salamini F, Thompson RD (1999) Evidence for factors regulating transfer cell-sp ecific expression in maize endosperm. Plant Mol Biol 41(3): 403-414 Hueros G, Varotto S, Salamini F, Thompson RD (1995) Molecular char acterization of BET1, a gene expressed in the endosperm tr ansfer cells of maize. Plant Cell 7 (6): 747-57 Huh JH, Bauer MJ, Hsieh TF, Fischer R (2007) Endosperm gene imprinting and seed development. Curr Opin Genet Dev 17(6): 480-5 Kielkopf CL, Lucke S, Green MR (2004) U2AF homology motifs: protein recognition in the RRM world. Genes Dev 18(13): 1513-1526 Kiesselbach TA (1949) The structure and reproducti on of corn. Research Bulletin 161, University of Nebraska College of Agriculture Larkins BA, Dilkes BP, Dante RA, Coelho CM, Woo YM, Liu Y (2001) Investigating the hows and whys of DNA endoreduplication. J Exp Bot 52(355): 183-92 Lid SE, Al RH, Krekling T, Meeley RB Ranch J, Opsahl-Ferstad HG, Olsen OA (2004) The maize disorganized aleurone layer 1 and 2 ( dil1, dil2) mutants lack control of the mitotic division plane in the aleurone layer of developing endosperm. Planta 218: 370378 Lid SE, Gruis D, Jung R, Lorentzen JA, Ananiev E, Chamberlin M, Niu X, Meeley R, Nicols S, Olsen OA (2002) The defective kernel 1 ge ne required for normal aleurone development in maize grains encodes a membrane protein of the calpain gene superfamily. Proc. Natl. Acad. Sci USA 99: 5460-5465 Lin BY (1984) Ploidy barrier to endosper m development in maize. Genetics 107 : 103 Lisch D (2002) Mutator transpos ons.Trends Plant Sci 7(11): 498-504 Lopes MA, Lark ins BA (1993) Endosperm origin, development, and function. Plant Cell 5: 1383-99 Lorkovi ZJ, Wieczorek Kirk DA, Lambermon MH, Filipowicz W (2000) Pre-mRNA splicing in higher plants. Trends Plant Sci 5(4): 160-7

PAGE 81

81 Luehrsen KR, Walbot V (1990) Intron enhancement of gene expression and the splicing efficiency of introns in maize cells. Mol Gen Genet 225(1): 81-93 Lunde CF, Morrow DJ, Roy LM, Walbot V (2003) Progress in maize gene discovery: a project update. Funct Integr Genomics 3(1-2): 25-32 Maitz M, Santandrea G, Zhang Z, Lal S, Hannah LC, Salamini F, Thompson RD (2000) rgf1, a mutation reducing grai n filling in maize through e ffects on basal endosperm and pedicel development. Plant J 23: 29-42 May BP, Liu H, Vollbrecht E, Senior L, Rabinowicz PD, Roh D, Pan X, Stein L, Freeling M, Alexander D, Martienssen R (2003) Maize-targeted mutagenesis: A knockout resource for maize. Proc Natl Acad Sci U S A 100: 11541-11546 McCarty DR, Settles AM, Suzuki M, Tan BC Latshaw S, Porch T, Robin K, Baier J, Avigne W, Lai J, Messing J, Koch KE, Hannah LC (2005) Steady-state transposon mutagenesis in maize. Plant J 44: 52-61 Mollet I, Barbosa-Morais NL Andrade J, Carmo-Fonseca M (2006) Diversity of human U2AF splicing factors. FEBS J 273(21): 4807-4816 Muiz LM, Royo J, Gmez E, Barrero C, Bergareche D, Hueros G (2006) The maize transfer cell-specific type-A response regulator ZmTCRR-1 appears to be involved in intercellular signalling. Plant J 48 (1): 17-27 Neuffer G, Sheridan WF (1980) Defective kernel mutants of maize. I. Genetic and lethality studies. Genetics 95: 929-44 Nowack MK, Grini PE, Jakoby MJ, Lafos M, Koncz C, Schnittger A (2006) A positive signal from the fertilization of the egg cell se ts off endosperm proliferation in angiosperm embryogenesis. Nat Genet 38(1): 63-67 Nowack MK, Shirzadi R, Dissmeyer N, Do lf A, Endl E, Grini PE, Schnittger A (2007) Bypassing genomic imprinting allows seed development. Nature 447(7142): 312-5 Ohad N (2007) Plant development: parent al conflict overcome. Nature 447(7142): 275-6 Ohad N, Margossian L, Hsu YC, Williams C, Repetti P, Fischer RL (1996) A mutation that allows endosperm development without fertilization. Proc Natl Acad Sci USA 93(11): 5319-5324 Olsen OA (1998) Endosperm developments. Plant Cell 10 (4): 485-8 Olsen OA (2001) Endosperm development: Cellulari zation and Cell Fate Specification. Annual Review of Plant Physiology and Plant Molecular Biology 52 : 233-67 Olsen OA (2004) Nuclear endosperm development in cereals and Arabidopsis thaliana Plant Cell Suppl: S214-27 Opsahl-Ferstad HG, Le Deunff E, Dumas C, Rogowsky PM (1997) ZmEsr, a novel endosperm-specific gene expressed in a rest ricted region around the maize embryo. Plant J 12(1): 235-46 Ortiz DF, Strommer JN (1990) The Mu1 maize transposable element induces tissue-specific aberrant splicing and polyadenylation in two Adh1 mutants. Mol Cell Biol 10(5): 2090-5

PAGE 82

82 Porch TG, Tseung CW, Schmelz EA, Settles AM (2006) The maize Viviparous10/Viviparous13 locus encodes the Cnx1 gene required for molybdenum cofactor biosynthesis. Plant J 45(2): 250-63 Quesada V, Dean C, Simpson GG (2005) Regulated RNA pro cessing in the control of Arabidopsis flowering. Int J Dev Biol 49(5-6): 773-80 Quesada V, Macknight R, Dean C, Simpson GG (2003) Autoregulation of FCA pre-mRNA processing controls Arabidopsis flowering time. EMBO J 22(12): 3142-52 Reddy AS (2007) Alternative splicing of pre-messenger RNAs in plants in the genomic era. Annu Rev Plant Biol 58: 267-94 Robertson D (1978) Characterization of a mutator system in maize. Mutation Research 51: 2128 Scanlon MJ, Myers AM (1998) Phenotypic analysis and mo lecular cloning of discolored-1 (dsc1), a maize gene required for early ke rnel development. Plant Molecular Biology 37: 483-493 Scanlon MJ, Stinard PS, James MG, Myers AM, Robertson DS (1994) Genetic analysis of 63 mutations affecting maize kernel developm ent isolated from Mutator stocks. Genetics 136: 281-94 Shannon JC (1994) Establishment and culture of maize endosperm. The maize handbook. M. Freeling and V Walbot, eds. Sp ringer-Verlag. New York. pp. 719-722 Shannon JC, Batey JW (1973) Inbred and hybrid effects on establishment of in vitro cultures of Zea mays L. endosperm. Crop Sci 13: 491-492. Scott RJ, Spielman M (2006) Deeper into the maize: new insights into genomic imprinting in plants. Bioessays 28(12): 1167-71 Settles AM (2005) Maize community resources for forward and reverse genetics. Maydica 50: 405-414 Settles AM, Latshaw S, McCarty DR (2004) Molecular analysis of high-copy insertion sites in maize. Nucleic Acids Res 32(6): e54 Shen B, Li C, Min Z, Meeley RB, Tarczynski MC, Olsen OA (2003) sal1 determines the number of aleurone cell layers in maize endosperm and encodes a class E vacuolar sorting protein. Proc Natl Acad Sci U S A 100: 6552-6557 Sheridan WF, Neuffer MG (1980) Defective kernel mutants of maize II. Morphological and embryo culture studies. Genetics 95 : 945-960 Singh M, Lewis PE, Hardeman K, Bai L, Rose JK, Mazourek M, Chomet P, Brutnell TP (2003) Activator mutagenesis of the pink scut ellum1/viviparous7 locus of maize. Plant Cell 15(4):874-84 Springer PS, Holding DR, Groover A, Yordan C, Martienssen RA (2000) The essential Mcm 7 protein PROLIFERA is localized to th e nucleus of dividing cells during the G(1) phase and is required maternally for early Arabidopsis development. Development 127(9): 1815-1822

PAGE 83

83 Suzuki M, Latshaw S, Sato Y, Settles AM, Koch KE, Hannah LC, Kojima M, Sakakibara H, McCarty DR (2008) The Maize Viviparous8 Locu s, Encoding a Putative ALTERED MERISTEM PROGRAM1-Like Peptidase, Regulates Abscisic Acid Accumulation and Coordinates Embryo and Endosperm Development. Plant Physiol 146(3): 1193-206 Suzuki M, Settles AM, Tseung CW, Li QB, Latshaw S, Wu S, Porch TG, Schmelz EA, James MG, McCarty DR (2006) The maize viviparous15 locus encodes the molybdopterin synthase small subunit. Plant J 45(2): 264-74 Thompson RD, Hueros G, Becker H, Maitz M (2001) Development and functions of seed transfer cells. Plant Science 160: 775-783 Tian Q, Olsen L, Sun B, Lid SE, Brown RC, Lemmon BE, Fosnes K, Gruis DF, OpsahlSorteberg HG, Otegui MS, Olsen OA (2006) Subcellular local ization and functional domain studies of DEFECT IVE KERNEL1 in maize and Arabidopsis suggest a model for aleurone cell fate specificati on involving CRINKLY4 and SUPERNUMERARY ALEURONE LAYER1. Plant Cell 19 (10): 3127-45 Tronchere H, Wang J, Fu XD (1997) A protein rela ted to splicing factor U2AF35 that interacts with U2AF65 and SR proteins in splicing of pre-mRNA. Nature 388(6640): 397-400 Yim YS, Davis GL, Duru NA, Musket TA Linton EW, Messing JW, McMullen MD, Soderlund CA, Polacco ML, Gardiner JM, Coe EH Jr (2002) Characterization of three maize bacterial artifici al chromosome libraries toward anchoring of the physical map to the genetic map using high-density bacterial artificial chromosome filter hybridization. Plant Physiol 130(4): 1686-96 Young TE, Gallie DR (2000) Programmed cell death during endosperm development. Plant Mol Biol 44(3): 283-301 Veit B, Briggs SP, Schmidt RJ, Yanofsky MF, Hake S (1998) Regulation of leaf initiation by the terminal ear 1 gene of maize. Nature 393(6681): 166-8 Walbot V (2000) Saturation mutagenesis using mai ze transposons. Curr Opin Plant Biol 3: 103107 Walbot V, Rudenko G (2002) MuDR/Mu Transposable Elemen ts of Maize. Mobile DNA II. N. Craig, R. Craigie, M. Gellert, A. Lam bowitz, eds. American Society Microbiology. Washington D.C. pp. 533-563 Walbot V, Warren C (1988) Regulation of Mu element copy number in maize lines with an active or inactive Mutator transposable element system. Mol Gen Genet 211(1): 27-34 Wang BB, Brendel V (2006b) Genomewide comparative anal ysis of alternative splicing in plants. Proc Natl Acad Sci U S A 103(18): 7175-7180 Wang C, Barry JK, Min Z, Tordsen G, Rao AG, Olsen OA (2003) The calpain domain of the maize DEK1 protein contains the conserved catalytic triad and functions as a cysteine proteinase. J Biol Chem 278: 34467-34474

PAGE 84

BIOGRAPHICAL SKETCH Diego Sebastian Fajardo wa s born in Lim a, Peru. His undergraduate studies were completed at the Universidad Peruana Cayeta no Heredia in Lima-Peru. He obtained his bachelors degree in biological sciences in December 1994. Wh ile an undergraduate student, Diego worked as a research assistant in the Cellular Biology and Viro logy laboratory in the Department of Microbiology at the Universidad Peruana Cayetano Heredia. He worked in the measurement of the effect of native amazon plant ( Brosinum rubescens) extract on Balb-3T3 mouse fibroblast. In January 1995, he started working at the International Potato Center (CIP) in the Genetic Resources Department as a research assistant under the supervision of Dr. Mark Ghislain and Dr. Dapeng Zhang. In August 1997, he was accepted as a Master of Science student in the Horticulture Department at Louisiana State University (LSU). He worked with Dr. Don LaBonte on the genetic diversity in Papua New Guinea sweetpotato ( Ipomoea batatas L. Poir.) germplasm. Upon completion of his masters degree in December 2000, he continued with his graduate studies and was accepted in the Plant Molecular and Cellular Biology Program (PMCB) at the University of Florida in August 2001. In June 2002, Diego initiated his Ph.D. project under the supervision of Dr. Andrew M. Settles in the study of maize ( Zea mays L.) rough endosperm ( rgh) seed mutants.