INVESTIGATION OF GENES CONTRIBUTING TO WHOLE PLANT SENESCENCE IN SOYBEAN (GLYCINE MAX)

Plant senescence is a genetically determined developmental program characterized by systematic degradative processes that involves activation of new gene activity and down-regulation of other genes that ultimately leads to cell, tissue, organ and whole plant death. Elucidating senescence regulatory pathways and participating genes will allow for the development of strategies to improve crop yields and also curtail postharvest losses. Three genes are known to be primary regulators of senescence in soybean; namely, g, D1, and D2. In double and triple mutant combinations these genes confer an evergreen leaf and seed phenotype. The double mutation ggd1d1d2d2 shows an inhibition of degradation of chlorophyll and chlorophyll binding protein, but photosynthesis declines and the leaves still abscise. In the triple mutant GGd1d1d2d2 the leaves maintain the normal photosynthetic capacity, but still abscise. So, while the senescence program is not entirely blocked in the mutant background, it is altered. Studies have shown that the expression patterns of soybean senescence associated genes (SAGs) are regulated differentially by g, D1 and D2. Due to the pivotal regulatory nature of these three genes for senescence, it is important to identify their specific nature. In the first study, an analysis using available soybean genome resources (SoyBase, Phytozome, COGE, etc.) was undertaken. This has resulted in the identification of a gene, Glyma01g41610.2, which encodes a putative transcription factor residing within the marker boundaries of the D1 locus on chromosome 1 which also shares a high level of synteny with a region on chromosome 11 and includes a paralogous gene, Glyma11g03770.2, within the D2 marker boundaries. Similarly, another gene, Glyma01g00510.1, also encodes a putative transcription factor and is located within the marker boundaries of the g locus on chromosome 1. These genes were selected as candidates representing g, D1, and D2 for RT-PCR analysis. None of the initial candidates exhibited a differential expression profile when comparing wildtype and mutant allelic versions in isogenic genetic backgrounds. The D2 and g genes were then selected for Sanger sequencing to determine if sequence differences were responsible for the observed phenotypic variations. No differences in sequence were observed when comparing wild type and mutant allelic forms. However, sequence variations were observed when comparing g in Harosoy versus the reference genome cultivar, Williams 82. A second candidate g gene (Glyma01g00520.4) was selected from the defined marker boundary interval on chromosome 1 but did not demonstrate a differential pattern of expression using RT-PCR. In the second study, additional SAGs up-regulated in various plant species were used to identify candidate soybean ortholog genes that could possibly contribute to whole plant senescence. Some SAGs have been shown to contain a unique senescence response element (SRE) within their promoters that confers a senescencespecific pattern of expression. This is best exemplified by the SAG12 (cysteine protease) gene of Arabidopsis thaliana. The 33 base pair SRE for the SAG12 gene has been shown to harbor a well-conserved 7 base pair sequence that is also found in SAGs from other plant species. To determine if a related SRE could be responsible for regulating soybean SAGs, a genome-wide study of previously identified as SAGs in other plant species was performed using publically available databases to find related genes in soybean. This search has led to the identification of several soybean genes that harbor this SRE. These genes were bioinformatically analyzed using various structural criteria to identify the best potential soybean ortholog for each gene type. Structural criteria included measures of alignment similarity with the Arabidopsis SRE, proximity of the SRE to the transcription start site, gene architecture, polypeptide sequence identity, and phylogenetic and syntenic relationships. Genes meeting the defined structural criteria underwent evaluation for a functional role in soybean senescence through RT-PCR analysis using a suite of isogenic lines exhibiting normal as well as delayed senescence phenotypes. The selected mutants represent different combinations (single, double and triple mutants) of genes g, D1 and D2 that give rise to evergreen leaves and green seed phenotypes. None of the genes selected for functional analysis demonstrated evidence of differential expression among the selected isolines. However, many aditional genes harboring an SRE have yet to be investigated.

shows an inhibition of degradation of chlorophyll and chlorophyll binding protein, but photosynthesis declines and the leaves still abscise. In the triple mutant GGd1d1d2d2 the leaves maintain the normal photosynthetic capacity, but still abscise. So, while the senescence program is not entirely blocked in the mutant background, it is altered.
Studies have shown that the expression patterns of soybean senescence associated genes (SAGs) are regulated differentially by g, D1 and D2. Due to the pivotal regulatory nature of these three genes for senescence, it is important to identify their specific nature. In the first study, an analysis using available soybean genome resources (SoyBase, Phytozome, COGE, etc.) was undertaken. This has resulted in the identification of a gene, Glyma01g41610.2, which encodes a putative transcription factor residing within the marker boundaries of the D1 locus on chromosome 1 which also shares a high level of synteny with a region on chromosome 11 and includes a paralogous gene, Glyma11g03770.2, within the D2 marker boundaries. Similarly, another gene, Glyma01g00510.1, also encodes a putative transcription factor and is located within the marker boundaries of the g locus on chromosome 1. These genes were selected as candidates representing g, D1, and D2 for RT-PCR analysis. None of the initial candidates exhibited a differential expression profile when comparing wildtype and mutant allelic versions in isogenic genetic backgrounds. The D2 and g genes were then selected for Sanger sequencing to determine if sequence differences were responsible for the observed phenotypic variations. No differences in sequence were observed when comparing wild type and mutant allelic forms. However, sequence variations were observed when comparing g in Harosoy versus the reference genome cultivar, Williams 82. A second candidate g gene (Glyma01g00520.4) was selected from the defined marker boundary interval on chromosome 1 but did not demonstrate a differential pattern of expression using RT-PCR.
In the second study, additional SAGs up-regulated in various plant species were used to identify candidate soybean ortholog genes that could possibly contribute to whole plant senescence. Some SAGs have been shown to contain a unique senescence response element (SRE) within their promoters that confers a senescencespecific pattern of expression. This is best exemplified by the SAG12 (cysteine protease) gene of Arabidopsis thaliana. The 33 base pair SRE for the SAG12 gene has been shown to harbor a well-conserved 7 base pair sequence that is also found in SAGs from other plant species. To determine if a related SRE could be responsible for regulating soybean SAGs, a genome-wide study of previously identified as SAGs in other plant species was performed using publically available databases to find related genes in soybean. This search has led to the identification of several soybean genes that harbor this SRE. These genes were bioinformatically analyzed using various structural criteria to identify the best potential soybean ortholog for each gene type. Structural criteria included measures of alignment similarity with the Arabidopsis SRE, proximity of the SRE to the transcription start site, gene architecture, polypeptide sequence identity, and phylogenetic and syntenic relationships. Genes meeting the defined structural criteria underwent evaluation for a functional role in soybean senescence through RT-PCR analysis using a suite of isogenic lines exhibiting normal as well as delayed senescence phenotypes. The selected mutants represent different combinations (single, double and triple mutants) of genes g, D1 and D2 that give rise to evergreen leaves and green seed phenotypes.
None of the genes selected for functional analysis demonstrated evidence of differential expression among the selected isolines. However, many aditional genes harboring an SRE have yet to be investigated.

Plant Senescence
One of the more striking hallmarks that signals the arrival of autumn is the magnificently aesthetic display of color changes in tree leaves. This annual, well orchestrated event is the result of a tightly regulated though not yet well understood process known as senescence. Senescence is the sequence of biochemical and physiological events comprising the final stage of development for a plant tissue, organ or whole plant, from the mature, fully developed state until death. The changes that take place in senescence represent a genetically programmed sequence with close coordination at the cell and tissue levels. Cells remain viable and show tight metabolic regulation until the end of senescence . During senescence, leaf cells undergo orderly changes in cell structure, metabolism, and gene expression. The earliest and most significant change in cell structure is the breakdown of the chloroplast, the organelle that contains up to 70% of the leaf protein. Metabolically, carbon assimilation is replaced by catabolism of chlorophyll and macromolecules such as proteins, membrane lipids, and RNA. Increased catabolic activity is responsible for converting the cellular materials accumulated during the growth phase of leaves and the redistribution of micro-and macro-nutrients, including nitrogen, sulphur, phosphorus and potassium, to growing and reproductive organs . Finally, upon reaching maturity, leaves abscise (Rubinstein and Leopold, 1964).
Thus, although senescence is a deleterious process for the leaf organ, it critically contributes to the fitness of the whole plant by ensuring optimal production of offspring and better survival of plants in their given temporal and spatial niches.
allowing the rest of the plant to continue in its development (Buchanan-Wollaston, 1997).

Soybean as a Model to Study Senescence
In 2011, soybean represented 56 percent of world oilseed production, 33 percent of which was produced in the United States. The United States exported 1.275 billion bushels (34.7 million metric tons) of soybeans, which accounted for 37 percent of the global soybean trade. U.S. soybean and soy product exports exceeded $21.5 billion in 2011 (http://www.soystats.com). A large part of this production is used in the extraction of oil, yielding a cake of high protein quality. Soy products are regarded as economical and nutritious feedstuffs with high crude protein content and a reasonably balanced amino acid profile along with many industrial and practical uses (Gatlin III et al., 2007).
G. max exhibits a monocarpic life pattern (Nooden, 1988), meaning it flowers and fruits once in a life cycle (Simmonds, 1980). During this annual life cycle, soybean senescence occurs primarily in the leaves which appear to be the target of the senescence-inducing influence from the seeds. This system of leaf death may be true for other species as well (Nooden, 1988) making soybean a candidate model for basic/fundamental studies. Evidence that soybean senescence may be delayed naturally has also been observed . Previous work with normally aging plants (in the absence of biotic and abiotic stressors) has shown the role hormones play in promoting soybean senescence. It was discovered that Arabidopsis lines harboring defects for ethylene, a common plant hormone, showed a delayed senescence phenotype . Prior research in soybean has also shown that removal of the epicotyls at 16 or 17 days post-germination reversed the decline in nucleic acid, protein, and chlorophyll content in the cotyledons (modified embryonic leaves that appear in early development). Epicotyl removal at 18 days did not reverse the decline in these components, indicating the cotyledon had passed "the point of no return" developmentally (Krul, 1974). Three mutant genes (g, D1, D2) that delay senescence have also been identified .
In an agricultural setting, leaf senescence may limit yield, contributing to the postharvest loss of vegetable crops. Therefore, studying leaf senescence will not only contribute to our knowledge about this fundamental developmental process, but may also lead to ways of manipulating the senescence process for agricultural applications (Gan and Amasino, 1997) such as improving stress tolerance .

Transcription Factors in Plant Development
The regulation of gene transcription is central both to tissue specific-gene expression and to the regulation of gene activity in response to specific stimuli. While instances of posttranscriptional regulation do exist (miRNA, RNAi, etc.), in most cases regulation occurs at the level of transcription by deciding which genes will be transcribed into the primary RNA transcript. Once this has occurred, the remaining stages of gene expression, such as RNA splicing, ultimately result in the production of the corresponding protein . One of the largest and most diverse classes of DNA-binding proteins responsible for regulating gene expression are transcription factors. Transcription factors, largely confined to the nucleus, regulate cell development, differentiation, and cell growth by binding to a specific DNA site (or set of sites) and regulating gene expression (Pabo and Sauer, 1992). Eukaryotic transcription factors usually consist of several domains. The DNA-binding domain binds to regulatory sequences that can either be adjacent to the promoter or at some distance from it. Most commonly, transcription factors include additional domains that help activate transcription. When a transcription factor is bound to DNA, its activation domain promotes transcription by interacting with RNA polymerase II, by interacting with other associated proteins, or by modifying the local structure of chromatin (Berg et al., 2012).
The DNA site(s) or response elements targeted by transcription factors are conserved DNA bases residing adjacent to the genes they regulate that may repress or activate gene expression. These include promoter elements and enhancers that form a complete set of regulators for each gene that is unique ensuring the right amount of the right protein is expressed at the right time as development proceeds. Transcription factors are aided by crucial proteins such as coactivators, corepressors, chromatin remodelers, histone acetylases, deacetylases, kinases, and methylases, which are present in all eukaryotic cells and contribute to the initiation of every RNA polymerase II primary transcript that eventually becomes messenger RNA (Brivanlou and Darnell, 2002).
The major families of plant transcription factors are: MYB, AP2/EREBP, NAC, bHLH/MYC, bZIP, HB, Z-C2H2, MADS, WRKY, ARF-Aux/IAA, and Dof (Riechmann and Ratcliffe, 2000). These transcription factors employ various structural motifs such as the helix-turn-helix, basic-leucine zipper and Cys2His2 zinc-finger to achieve binding of their particular recognition sequences. A vast majority of these motifs bind in the major groove of DNA and interact with DNA bases through different combinations of electrostatic and Van der Waals forces (Berg et al., 2012).

Transcription Factors in Plant Senescence
Leaf senescence is an active process involving the differential expression of hundreds of genes and therefore it is presumed that numerous transcription factors are involved as central elements of the regulatory network . Genes for 96 transcription factors have been identified in Arabidopsis as being upregulated at least threefold in senescing leaves. These belong to 20 different transcription factor families, the largest groups being NAC, WRKY, C2H2-type zinc finger, AP2/EREBP, Aux/IAA, and MYB proteins , with only few examples having been further demonstrated as having a specific functional role in senescence.
NAC proteins are one of the largest families of plant-specific transcription factors with more than 100 members in Arabidopsis. NAC family genes play a role in embryo and shoot meristem development, lateral root formation, auxin signaling, and defense response. A total of 20 genes encoding NAC transcription factors, representing almost one fifth of the NAC family members, showed enhanced expression during natural senescence and dark-induced senescence .
Recently, a T-DNA knockout mutation of one of these genes, AtNAP, was shown to delay leaf senescence significantly. Induced overexpression caused early senescence, suggesting that AtNAP functions as a positive element in leaf senescence .
Among the plant-specific WRKY transcription factor gene family, AtWRKY53 and WRKY6 have been further characterized in relation to leaf senescence. WRKY53 is upregulated at a very early stage of leaf senescence but decreases again at later stages, implying that WRKY53 might play a regulatory role in the early events of leaf senescence (Hinderhofer and Zentgraf, 2001). A knockout line of the WRKY53 gene showed delayed leaf senescence, whereas induced overexpression caused premature senescence, showing that it functions as a positive element in leaf senescence (Miao et al., 2004). WRKY6 is strongly up-regulated during leaf senescence as well as during pathogen infection. However, although the wrky6 knockout mutation alters expression of SAGs it does not have any apparent effect on leaf senescence. SIRK, a gene encoding a receptor-like protein kinase whose developmental expression is strongly induced specifically during leaf senescence, is dependent on WRKY6 function.
Senescing leaves of wrky6 knockout mutants showed a reduction in SIRK transcript levels while green leaves of WRKY6 overexpression lines showed elevated SIRK transcript levels. Furthermore, the SIRK gene promoter was specifically activated by WRKY6 in vivo (Robatzek and Somssich, 2002).

Instances of transcription factors with domains responsible for interactions
with phytohormones have also been implicated in senescence. AP2 (APETALA2) and EREBPs (ethylene-responsive element binding proteins) are members of a family of transcription factors unique to plants whose distinguishing characteristic is that they contain the so-called AP2 DNA-binding domain. AP2/ REBP genes form a large multigene family, which play a variety of roles throughout the plant life cycle, from being key regulators of several developmental processes, like floral organ identity determination or control of leaf epidermal cell identity, to forming part of the mechanisms used by plants to respond to various types of biotic and environmental stress (Riechmann and Meyerowitz, 1998).
The RAV family transcription factor in Arabidopsis, RAV1, has an N-terminal region containing an AP2 DNA-binding domain. Rav1 mRNA increases at a later stage of leaf maturation and reached a maximal level early in senescence, but decreases again during late senescence. This profile indicates that RAV1 could play an important regulatory role in the early events of leaf senescence. Furthermore, constitutive and inducible overexpression of RAV1 causes premature leaf senescence.
These data strongly suggest that RAV1 is sufficient to cause leaf senescence and functions as a positive regulator in this process . and show an inhibition of chlorophyll degradation and chlorophyll-binding proteins yet they still undergo a decline in photosynthetic activity and leaf abscission (Guiamet et al.1991. When combined with the dominant mutant G (GGd1d1d2d2) a decline in photosynthetic activity does not take place, but the leaves still abscise. It has been suggested that d1d2 may control a central regulatory process in the senescence program and that homozygosity at both nuclear loci is required because the two are homeologous (duplicate) loci in the ancient tetraploid soybean genome . The genes responsible for these primary regulators remain unknown; however, because of their important role in the upstream regulation of the senescence program it is plausible they could encode transcription factors. Their characterization represents a logical and organized approach to better understand the mechanisms that control senescence. Near-isogenic lines are available as single, double and triple mutant combinations for these three genes and can serve as the basis for expression analysis of possible transcription factor genes representing g, D1, and D2 during senescence. As patterns of altered expression are characterized, a better understanding of the genetic regulation of the senescence pathways will unfold.

Map Based Methods of Gene Identification
A genetic map is a list of genetic elements ordered with regard to their chromosomal position according to their inheritance patterns. Previously these elements were inferred to be genes underlying phenotypic characters such as seed shape in pea or eye color in Drosophila, though not having to be restricted to morphological traits. More recently, DNA markers such as restriction length polymorphisms (RFLPs), simple sequence repeat (SSRs), and single nucleotide polymorphisms (SNPs) have become prominent in genetic mapping studies. Currently, genetic maps are thought of as ordered sets of markers, together with inter-marker distances representing milestones along a chromosome or region of a chromosome.
A genetic map serves many practical biological purposes and is a key tool in both classical and modern plant research. For the large majority of plants whose genomes are yet to be sequenced it provides an important resource to understand the order and spacing of markers (and relative order when compared to those of other   To facilitate the identification of the genetic basis of many traits and accelerate the creation of improved plant varieties an accurate genome sequence is needed. The first plant physical map generated was that of the Arabidopsis genome in 2000 due to its suitability as a model species for plant research and its small genome (157Mb) . In 2012, the 1.1-gigabase genome of Glycine max var.
Another free and publically available soybean database, SoyBase, has made available a tool that conveniently allows for a comparison of genetic and sequence maps. While there is a strong association with distance in both genetic and physical maps, genetic map distances are calculated by means of recombinant frequencies and do not represent actual physical distances on chromosomes. However, cytogenetic and molecular analysis has shown that genetic distances are, in fact, roughly proportional to chromosome distances , and this assumption is reflected in the current version of SoyBase maps. The markers available on these maps have been utilized by researchers aiming to discover new genes that may be responsible for soybean mosaic virus resistance (Rsv4) (Maroof et al., 2010), aphid resistance (Rag2) , and soybean cyst nematode resistance (rhg1-b)  using fine mapping techniques. It is by employing a map based approach that this current study aims to identify transcription factors that may ultimately represent the primary soybean regulator genes, g, D1 and D2.

A Conserved Promoter Element Identified in a Senescence Associated Gene (SAG) of Arabidopsis thaliana
Manipulating Senescence through Biotechnology The ultimate downstream targets of these primary regulator genes are important in effecting the degradative processes of senescence. Efforts to inhibit or delay the effects of senescence have been made using fusion proteins produced from gene constructs encoding a cysteine protease that incorporate the SAG12 gene senescence response element (SRE)  originally identified in the model organism A. thaliana . The SAG12 SRE is 33 base pairs in length and included within an essential promoter element located -472 to -784 upstream of the transcription start site. It is required for basal level promoter activity and for full SAG12 expression when in conjunction with an additional upstream enhancer (-1181 to -1345) and basal promoter (-66 to the start codon) ( Figure I). The SRE (and basal promoter) is highly conserved in the orthologous SAG12 gene of rapeseed; however, it does not have significant similarity to any known consensus binding sequences of transcription factors . The lack of similarity indicates that the developmental regulation of SAG12 may involve a new or divergent class of transcription factors that specifically recognize this SRE. That only partial promoter activity was conferred by the -603 to -571 region also implies that the other parts of the conserved essential promoter region are required for full SAG12 promoter activity .
In tobacco (Nicotiana tabacum) the SRE was effectively used with a maize homeobox gene, knotted (kn1), and isopentenyl transferase (ipt), a cytokininproducing gene known to inhibit senescence , to delay senescence. Tobacco plants harboring the SAG:kn1 and SAG:ipt constructs developed with normal morphology but had delayed senescence in both intact and detached leaves. Cytokinin levels were significantly raised in leaves of both experimental constructs compared to wild type, suggesting that the delay in leaf senescence may be mediated through changes in cytokinin metabolism (Ori, 1999). An ipt gene under control of the senescence-specific SAG12 promoter (pSAG12-IPT) significantly delayed developmental and postharvest leaf senescence in mature heads of transgenic lettuce (Lactuca sativa) homozygous for the transgene. Apart from retardation of leaf senescence, mature, 60-day-old plants exhibited normal morphology with no significant differences in head diameter or fresh weight of leaves and roots (McCabe, 2001). Similar results were achieved in other distantly related plant species such as rapeseed (Brassica napus) , tomato (Solanum lycopersicum) (Swartzberg et al., 2006), rice (Oryza sativa) (Liu et al., 2010), broccoli (Brassica oleracea var. italica) (Long-Fang, 2001) and bok choy (Brassica chinensis) (Yuan et al., 2002).
The Arabidopsis thaliana SAG12 SRE is Found in SAG's from Other Plant Species Studies by Davies and King (1993) and King et al. (1995) on asparagus found levels of asparagine and asparagine synthetase (AS) transcripts increase following the harvest of asparagus spears and during natural foliar senescence. Increased AS transcript levels during the course of senescence have also been observed in Arabidopsis, sunflower, M. truncatula, rice and corn (Gaufichona et al., 2010).
Asparagine is thought to be the major transport product in conditions of excess nitrogen or limited carbon supply, which may occur in detached tissues, during senescence, and in photosynthetic tissues during extended dark periods. Thus, the activity of the enzyme responsible for producing asparagine and AS must be controlled in response to a complex combination of metabolic, environmental and developmental signals . Interestingly, upstream promoter analysis identified a conserved senescence-specific sequence motif regulating the expression of the AS gene in asparagus  similar in sequence to the SRE of the Arabidopsis SAG12 gene.
This senescence-specific sequence  up-regulates the expression of the SAG12 gene during senescence. The upstream region of the AS gene was aligned to the SAG12 upstream region of both Arabidopsis and rapeseed and a highly conserved seven base pair region was identified within the well conserved 33 base pair region. This upstream promoter element was also shown to be functionally relevant with deletion assays showing senescence induction occurring when present and delayed when deleted ). An attempt to identify a soybean ortholog of this gene that behaves in a similar pattern is investigated in the present study.

Transcription Factor Genes in Soybean Senescence
Previous work by Chandlee and colleagues employed microarray analysis to identify genes that are differentially expressed in senescing leaf tissue in soybean ). Although problems can arise with this method from background interference caused by similar enzymes that are also active, or lack of instrument sensitivity due to low transcript levels (Buchanan-Wollaston, 1997), a MYB transcription factor was identified by this technique as highly differentially expressed ).
The structural characteristic common to all known MYB proteins is the DNAbinding domain which has been shown to bind DNA in a sequence-specific manner.
Additionally, these proteins usually contain a negatively charged activator domain that has been implicated in transcriptional activation in certain cases (Martin and Paz-Ares, 1997 repeats of about 52 residues (R2, R3) whereas MYB proteins from animals contain three (R1, R2, R3). These MYB repeats fold into a variant of the helix-turn-helix motif and contain 3 regularly spaced tryptophan residues that play a role in the folding of the hydrophobic core (Dubos et. al, 2010). Although these proteins share the homologous MYB domain, differences in the DNA base contacting residues produce distinct DNAbinding specificities in different members of the family. This gene superfamily participates in a host of processes including regulation of gene expression (Yanhui et al., 2006), secondary metabolism (Mehrtens et al., 2005), hormone signal transduction (Abe et al. 2003), response to environmental stresses (Jung et al., 2008), cell shape, and organ development (Higginson et al., 2003).

Higher plant species usually contain a large number of MYB proteins with
Arabidopsis genome encoding 196 different MYB proteins spread across four classes (Dubos et. al, 2010) and the soybean genome harboring 252 MYB genes divided into 3 classes (Du et. al, 2012 Arabidopsis has been shown to inhibit anthocyanin biosynthesis in the lettuce plant (Park et al., 2008) and thus acts as a negative regulator of anthocyanin synthesis.  (Becker and Theissen, 2003). MADS-box genes have also been implicated in senescence. The MADS-box gene AGL15 is preferentially expressed during embryogenesis and seed development (Perry et al., 1996) and when constitutively expressed in Arabidopsis it strongly delays abscission and senescence in reproductive tissues (Fernandez et al., 2000). that it has a more specialized role in protein breakdown during senescence. Cathepsin cysteine proteases are active at acidic pH, and are therefore assumed to be localized to lysosomes or vacuoles . SAG101 has been identified as an acyl hydrolase. During senescence this protein is involved in degrading lipids  has a miRNA regulatory region that was identified in the LSD. This gene was selected for promoter analysis in the present study based on (RNAi)-mediated knock-down of GmSARK which dramatically retarded soybean leaf senescence .
Protein kinases and especially membrane-associated receptor-like kinases (RLKs) have been found to be involved in many developmental and stress signal transduction pathways. Each RLK consists of three domains, an extracellular receptor domain, a single-pass transmembrane domain and an intracellular kinase domain. In the absence of signal molecules RLKs are usually localized in cell membranes in the form of monomers; once signals emerge the extracellular domain will recognize and bind the signal molecules, resulting in the dimerization of RLKs. The dimerization usually causes the intracellular domains to be autophosphorylated or transphosphorylated, eventually activating the RLKs. Thus, the extracellular signals are transduced into the inside of cells . In two studies on senescence-regulated IPT gene expression, the promoter of a gene encoding a receptor protein kinase upregulated during senescence of Phaseolus vulgaris (common bean) leaves was used . Tobacco plants transformed with the PSARK:IPT construct showed a delay in senescence and an exceptional drought tolerance (Rivero et al., 2007). Recently, peanut plants transformed with the same construct were shown to maintain higher photosynthetic rates and higher transpiration under reduced irrigation conditions. In the field, the transgenic peanut plants produced significantly higher yields than the control plants (Qin et al., 2011). However, control of this IPT gene by PSARK also demonstrated an increase in expression of brassinoste related genes and repression of jasmonate genes causing the development of enhanced root biomass in these transgenic plants (Peleg et al. 2011;Rivero et al. 2010). This signifies that the increased levels of cytokinin during senescence may not be responsible for the longevity of the green pigmentation in the leaves or photosynthetic capabilities but rather due to additional changes in the metabolism of the plant .
The collection of senescence related genes that includes an AS, transcription factors (MYB, MADS and MYB60), proteases (SAG12 and SAG2), a lipase (SAG101), and a transmembrane receptor kinase (SARK) ( Table I)

Comparative Genomics Platform Overview (CoGe)
CoGe is publicly available at http://genomevolution.org. This resource contains four major systems: a data engine storing thousands of genomes, a suite of interconnected web-based tools, a wiki documentation system with hundreds of pages on comparative genomics, and a TinyURL resource for storing links to CoGe to regenerate data and analyses. The data in CoGe is constantly growing as new genomes and new versions of existing genomes become available. Currently, there are nearly 20,000 genomes from 15,000 organisms. There are over 20 tools in CoGe; each of these performs one general task, such as searching for genomes, displaying FASTA sequences, querying genomes, comparing genomic regions, etc. These tools are all interlinked with one another so that results generated in one tool may be seamlessly sent to another tool for downstream analyses. Due to the interlinking of these tools, following no specific workflow or analytical pipeline is not required (Lyons and Freeling, 2008).

Main CoGe Entrance Pages
 OrganismView -Search and get an overview of an organism and its genomic information.
 CoGeBlast -Blast against any number of organisms using the CoGe Blast interface (supports Blast, and BlastZ). PubMed ID numbers and DDBJ/EMBL/GenBank nucleotide sequence databases accession numbers also included (Higo et al., 1999).

Gramene
Gramene's (www.gramene.org) purpose is to facilitate researchers' ability to understand the grass genomes and take advantage of genomic sequence known in one species for identifying and understanding corresponding genes, pathways and phenotypes in other grass species. This is achieved by building automated and curated relationships between cereals for both sequence and biology. The automated and curated relationships are queried and displayed using controlled vocabularies and webbased displays. The controlled vocabularies (Ontologies), currently being used include Gene ontology, Plant ontology, Trait ontology, Environment ontology and Gramene Taxonomy ontology. The web-based displays for phenotypes include the Genes and Quantitative Trait Loci (QTL) modules. Sequence based relationships are displayed in the Genomes module using the genome browser adapted from Ensembl, in the Maps module using the comparative map viewer (CMap) from GMOD, and in the Proteins module displays. BLAST is used to search for similar sequences. Literature supporting all the above data is organized in the Literature database (Jaiswal et al., 2006).

Primer-Blast
Primer-BLAST was developed at NCBI to help users make primers that are specific to intended PCR target. It uses Primer3 to design PCR primers and then uses BLAST and global alignment algorithm to screen primers against user-selected database in order to avoid primer pairs (all combinations including forward-reverse primer pair, forward-forward as well as reverse-reverse pairs) that can cause non-specific amplifications (Ye et al., 2012).

Geneious
Geneious Pro is a commercial, integrated, cross-platform bioinformatics software suite for manipulating, finding, sharing, and exploring biological data such as DNA sequences or proteins, phylogenies, 3D structure information, publications, etc. It features sequence alignment and phylogenetic analysis, contig assembly, primer design and cloning, access to NCBI and UniProt, BLAST, protein structure viewing, automated PubMed searching, and many more applications (Biomatters New Zealand, 2013).

Significance of Project and Outline
Soybean is a major source of food worldwide for humans and livestock. Improving yields is a desirable objective because of this major economic significance. Whole plant/leaf senescence, an orderly degenerative process leading to death, is a developmental program known to be genetically controlled in soybean. The manipulation of this process can potentially improve yields. An understanding of soybean genes that regulate senescence both upstream, early events in the developmental program and downstream, later events in the developmental program, will be useful in this regard ). Currently little is known about the genes that serve as the primary regulators of senescence or about promoter elements for the downstream responding SAGs in the soybean senescence program.
The auxin response factor and MYB gene families represent two of the largest families of transcription factors that play important roles in many aspects of growth and development in most, if not, all eukaryotes. New knowledge will allow researchers to continue to decipher the many roles of these important transcription factors in soybean development. The correlation of transcription factors that lead to whole plant and leaf senescence and the possibility of their role as major regulators of senescence make them attractive targets to elucidate the regulatory framework of this developmental process. The regulatory regions of genes that usher in the changes normally associated with senescence are also of interest because of the opportunities to exploit them for agricultural benefits. The availability of evergreen mutants as soybean near-isogenic lines provides an important resource to assist in the study of these important processes.
Any further understanding of senescence, especially its regulators, will add to the understanding of the genetic basis for developmental regulation and has direct application to the improvement of crop plants by reduction of spoilage and increased yields.
This project targets three objectives:

1) Identification of primary regulatory genes of soybean senescence using a map based approach
Genetic and physical maps offer genetic markers that are used as boundaries to create a database that can be screened for transcription factors representing g, D1 and D2.

2) Genome-wide identification of soybean genes that harbor a conserved UPE and possibly contribute to whole plant senescence
Screen publically available databases using various structural criteria to identify candidate genes for further functional analysis.

3) Expression analysis of primary regulator and ortholog genes harboring an SRE
Evaluate the expression of candidate genes using RT-PCR to monitor their expression in WT and mutant soybean strains.

Figure I. Conserved promoter sequences used as selection criteria of soybean SAG's.
A highly senescence-specific promoter identified in the Arabidopsis thaliana SAG12 gene. A subsequent study identified a conserved 7 base pair sequence within the 33 base pair Senescence Response Element (SRE). These two sequences serve as the basis for selection of candidate genes.

Literature Cited
Abe H., Urao T., Ito T., Seki M., Shinozaki K., and Yamaguchi-Shinozaki K. (2003)  shows an inhibition of degradation of chlorophyll and chlorophyll binding protein, but photosynthesis declines and the leaves still abscise. In the triple mutant GGd1d1d2d2 the leaves maintain the normal photosynthetic capacity, but still abscise. So, while the senescence program is not entirely blocked in the mutant background, it is altered.

Arabidopsis AtMYC2 (bHLH) and AtMYB2 (MYB) function as transcriptional activators in abscisic acid signaling Plant Cell
Studies have shown that the expression patterns of soybean senescence associated genes (SAGs) are regulated differentially by g, D1 and D2. Due to the pivotal regulatory nature of these three genes for senescence, it is important to identify their specific nature. As such, an analysis using available soybean genome resources (SoyBase, Phytozome, COGE, etc.) was undertaken. This has resulted in the identification of a gene, Glyma01g41610.2, which encodes a putative transcription factor residing within the marker boundaries of the D1 locus on chromosome 1 which also shares a high level of synteny with a region on chromosome 11 and includes a paralogous gene, Glyma11g03770.2, within the D2 marker boundaries. Similarly, another gene, Glyma01g00510.1, also encodes a putative transcription factor and is located within the marker boundaries of the g locus on chromosome 1. These genes the initial candidates exhibited a differential expression profile when comparing wildtype and mutant allelic versions in isogenic genetic backgrounds. The D2 and g genes were then selected for Sanger sequencing to determine if sequence differences were responsible for the observed phenotypic variations. No differences in sequence were observed when comparing wild type and mutant allelic forms. However, sequence variations were observed when comparing g in Harosoy versus the reference genome cultivar, Williams 82. A second candidate g gene (Glyma01g00520.4) was selected from the defined marker boundary interval on chromosome 1 but did not demonstrate a differential pattern of expression using RT-PCR.

Introduction
Senescence is the sequence of biochemical and physiological events comprising the final stage of development for a plant tissue, organ or whole plant, from the mature, fully developed state until death. The changes that take place in senescence represent a genetically programmed sequence, with close coordination at the cell and tissue levels . During senescence, leaf cells undergo orderly changes in cell structure, metabolism, and gene expression which ultimately results in redistribution of nutrients to growing and reproductive organs before death and ultimately whole leaf abscission . Thus, although senescence is a deleterious process for the leaf organ, it critically contributes to the fitness of the whole plant by ensuring optimal production of offspring and better chance of survival.
Leaf senescence is thus an evolutionarily selected developmental process and comprises an important phase in the plant life cycle .
Soybean (Glycine max L. Merr.) is an agriculturally and economically important crop (http://www.soystats.com). During its annual life cycle, soybean senescence occurs primarily in the leaves which appear to be the target of the senescence-inducing influence from the seeds. This system of leaf death may be true for other species as well (Nooden, 1988) making soybean a candidate model for basic/fundamental studies.
The ability to regulate gene transcription is central both to tissue specific-gene expression and to the regulation of gene activity in response to specific stimuli . Leaf senescence is an active process involving the differential expression of hundreds of genes and therefore it is presumed that numerous transcription factors are involved as central elements of the regulatory network . Three different genes (g, D1, D2) are known to be involved with regulating the whole plant/leaf senescence program of soybean because mutations contribute to an evergreen (non-senescing) phenotype. Homozygous d1d1d2d2 lines remain green and show an inhibition of chlorophyll degradation and chlorophyllbinding proteins yet they still undergo a decline in photosynthetic activity and leaf abscission (Guiamet et al.1991. When combined with the dominant mutant G (GGd1d1d2d2) a decline in photosynthetic activity does not take place, but the leaves still abscise. It has been suggested that d1d2 may control a central regulatory process in the senescence program and that homozygosity at both nuclear loci is required because the two are homeologous (duplicate) loci in the ancient tetraploid soybean genome . The genes responsible for these primary regulators remain unknown; however, because of their important role in the upstream regulation of the senescence program it is plausible they encode transcription factors.
Using genetic and physical maps it's possible to identify putative candidates for the transcription factor genes representing g, D1, and D2. By employing a map based approach the current study aimed to identify transcription factors that may ultimately represent the primary soybean regulator genes. For analysis of these three genes, near-isogenic lines are available as single, double and triple mutant combinations and provided the basis for expression analysis of these candidate transcription factor genes. To examine a functional role for these candidates, RT-PCR was performed with candidates derived from the map-based strategy. Candidates not demonstrating differential expression were analyzed by Sanger sequencing.

Selection of Candidate Genes
The putative identity for g, D1, D2, the three genes responsible for regulating the soybean senescence program, was determined using genetic maps available through SoyBase. To resolve all possible candidate genes for g, the chromosome 1 (D1a) soybean 2003-composite map (www.soybase.org) was used to identify DNA markers that straddled a chromosome block surrounding g (Figure 1.1a). The position of g near the telomeric end of chromosome 1 (Figure 1.1a), and the lack of specification of map units in SoyBase, prevented selecting a marker on this telomeric end by this approach. To overcome this problem, the first base pair of chromosome 1 was used as a border marker to ensure all possible candidate genes for g were captured. The marker, SAT_332, representing a simple sequence repeat (SSR) motif of (AT) 25 , was used to set the other marker boundary toward the centromeric end of the chromosome with an end position of 355,784 base pairs. This marker was then verified to be in the correct flanking orientation using a newer version of the soybean genetic map (consensus 4.0 map) that saturates the soybean genome with SNP markers for better resolution. This updated version of the soybean genetic map was then compared with a current sequence map in SoyBase to verify that the marker selected actually was located flanking g (Figure 1.1a).
Next, this information was used with the Comparative Genomics (CoGe) OrganismView tool to screen the Glycine max genome for genes present within the chromosomal block bordered by the marker boundaries. The start and end physical locations of the boundary markers were entered for chromosome 1 and used to extract the most recently annotated coding sequences (CDS) (soybean v1.1) between them and entered into a spreadsheet database. Forty-two (42) gene annotations were identified between these markers. However, because some of these represented alternative transcripts, a total of 32 discrete genes were identified for screening ( Figure   1.2a). All genes were screened for domains with putative transcription factor identity.
To identify D1, the chromosome 1 (D1a) soybean 2003-composite map in SoyBase was used to find DNA markers that could be used as border markers delineating a chromosomal interval containing D1 (Figure 1.1b). A centromeric end marker, BARC-030807-06945, whose sequence starts at 53,063,806 base pairs, and a telomeric end marker, Sat_160, whose sequence end is at 53,236,862 base pairs, were selected as the two closest markers with available genomic sequence information flanking the D1 gene. These markers were then verified to be in the correct flanking orientation using the newer version of the soybean genetic map (consensus 4.0 map).
This updated version of the soybean genetic map was then compared with a current sequence map in SoyBase to verify that the markers selected actually were flanking D1 (Figure 1.1b).
This information was used with the CoGe OrganismView tool screen the Glycine max genome for annotated genes within the border boundaries. The start and end physical locations of the boundary markers were entered for chromosome 1 and used to extract the most recently annotated CDS (soybean v1.1) between them and were subsequently entered into a spreadsheet database. Twenty-eight gene annotations were found between these border markers. However, these included 7 alternative transcripts leaving a total of 21 discrete genes to be screened (Figure 1.2b). All genes were then analyzed for domains with putative transcription factor identity.
As previously, to identify D2, the chromosome 11 (B1) soybean 2003composite map in SoyBase was used to find DNA markers that could be used as borders delineating a chromosomal interval containing D2 (Figure 1.1c). A telomeric end marker, BARC-029533-06211, whose sequence begins at 546,754 base pairs and a centromeric end marker, Sat_272, whose sequence end is at 2,710,583 base pairs, were selected as the two closest markers flanking the D2 gene on either side. These markers were then verified to be in the correct flanking orientation using the newer version of the soybean genetic map (consensus 4.0 map). This updated version of the soybean genetic map was then compared with a current sequence map in SoyBase to verify that the markers selected actually were flanking D2.
Next, this information was used with the CoGe OrganismView tool to screen the G. max genome for annotated genes present within the border boundaries as previously done with the D1 gene interval. The start and end physical locations of the boundary markers were entered for chromosome 11 and used to extract the most recently annotated CDS (soybean v1.1) between them which were subsequently entered into a spreadsheet database. Four hundred twenty-three gene annotations were found between these border markers. However, these included 133 alternative transcripts leaving a total of 290 genes to be screened (Figure 1.2c). All genes were then analyzed for domains with putative transcription factor identity.
Due to the presumed nature of the D1 and D2 duplication, both Synmap and Synfind (CoGe) were used to determine if any transcription factor genes resided within the boundaries delineated for each gene to support the possibility of D1 and D2 representing a gene duplication event. To determine if chromosomes 1 and 11 shared syntenic regions, Synmap was first used to determine soybean whole genome synteny (Figure 1.3). Next, Synfind was used to determine if Glyma01g41610.2 (chromosome 1) and Glyma11g03770.2 (chromosome 11) located in the syntenic regions of these two chromosomes shared high levels of synteny at the gene level by entering these genes into the "specify feature" section and selecting the most recent annotated CDS D2 (Table 1.5). Default settings for the Primer3 program were used except for: 1) the "Exclusion tab" which was then selected to prevent redundant sequences from appearing in the Results field by excluding Refseq transcripts of predicted mRNA and ncRNA and 2) the "Organism" field which was selected for "G. max." The 5 sets of results returned were then used to query the Phytozome soybean database using TBLASTN to ensure the sequence was located within the gene of interest and also to verify the primers were unique in the genome. Primers were also screened against Harosoy genomic DNA to verify amplification of the target genomic sequences in that strain.

RNA Extraction
Total leaf RNA was isolated from several developmental stages throughout the normal life cycle using Harosoy (Table 1.6) and isogenic lines harboring genes that affect leaf senescence (Table 1.7). Leaves were harvested immediately into liquid nitrogen and RNA was extracted the same day, using a standard phenol/chloroform extraction protocol and LiCl precipitation (Maniatis et al., 1986;).
RNA was quantified using a spectrophotometer and assayed for quality and quantity by gel electrophoresis in formaldehyde-containing agarose gels (Figure 1.5). The soybean developmental stages V5 and R7 were used in this study. V5 (vegetative 5) is characterized by five fully expanded, green trifoliate leaves are found on the plant and in R7 (reproductive 7) one major pod has changed to a brown color on the main stem (http://www.ag.ndsu.edu/pubs/plantsci.htm).

DNA Extractions
Whole genomic DNA was extracted using a standard protocol.

RT-PCR Analysis
Preliminary RT-PCR analysis was performed with total RNA from stages V5 and R7 of soybean using seven different isogenic backgrounds (

Sequence Analysis
Verification of the identity of the amplicons was performed by sequence analysis.
Bands were excised from the gels and purified with the Wizard SV Gel and PCR clean-up system according to the manufacturer's instructions (Promega, Madison WI).
Bands were sequenced using the facilities at the University of Rhode Island's Genomics and Sequencing Center (URIGSC). D2 genes. Analysis of these genes for domains with putative transcription factor identity revealed 2 within the D1 boundaries and 41 within the D2 boundaries. One of the two genes identified as a transcription factor, Glyma01g41610.2, encoding a soybean MYB gene located between D1 boundaries, was used to identify a paralogous gene within a highly syntenic region on chromosome 11, Glyma11g03770.2.
Glyma11g03770.2 also putatively encodes a soybean MYB gene located between the defined D2 boundaries (Figures 1.3 and 1.4). Interestingly, the best match in A.
thaliana for Glyma11g03770.2 using the Synfind program within CoGe was AtMYB5 where the promoter has been demonstrated to play a role in plant senescence (Heazelwood et. al, 2011 and D2 in these genetic backgrounds (Figure 1.9 and 1.10 respectively). However, there appeared to be no significant difference in the expression of these transcripts in early and late development in any of the genetic backgrounds.
To determine if an error in the coding sequence of Glyma11g03770.2 could account for the nature of the D2 mutation, Sanger sequencing was utilized. DNA templates from Harosoy and the L69-4266 single mutant (gD1d2) were used and compared to the Williams82 cultivar sequence available through www.phytozome.net . No differences were identified between the Harosoy and L64-2489 genomic sequences and the model cultivar Williams82 (Figure 1.11).
Although the results obtained in this study did not lead to the anticipated outcomes, they do open up other avenues of inquiry that warrant further investigation.
The method for candidate gene identification outlined in this study represent a logical strategy that allows for selection of a gene(s) that can be screened quickly and cost effectively by trial and error compared to the map based cloning technique. Currently, an alternative putative g gene, Glyma01g00520.4, remains to be sequenced. There exists the possibility that an error in the coding region of this gene could adversely affect gene functionality and therefore, the senescence developmental program (Zemach and Grafi, 2006).
The Fast Neutron Database, a resource of the soybase.org repository may aid in future gene selection. Fast neutron radiation was used to induce deletion mutations in the soybean genome followed by cataloging plant variation for seed composition, maturity, morphology, pigmentation, and nodulation traits .  The "Analysis Option" tab was selected and default settings used with the exceptions of "Syntenic Depth" and "CodeML" choices. These advanced analytical tools identify orthologous syntenic regions by the relative evolutionary distance of syntenic gene pairs using synonymous mutation rates and the algorithm "Quota Align" for screening syntenic regions to enforce a specific mapping of syntenic regions between genomes.

Introduction
Leaf senescence is a highly regulated developmental process that ends with the programmed death of leaf cells . During leaf senescence, cellular components such as proteins, lipids, and nucleic acids are degraded, and the released nutrients are mobilized from the leaves for re-use in other parts of the plant Noodén and Guiamet, 1991;. Understanding the molecular mechanics of senescence has direct application to the improvement of crop plants by reduction of spoilage and increased yields. Efforts to inhibit or delay the effects of senescence have been made using fusion proteins produced from gene constructs that incorporate the Arabidopsis thaliana SAG12 gene (encoding a cysteine protease) senescence response element (SRE) . The SAG12 SRE is 33 base pairs in length and is found within an essential promoter element located -472 to -784 upstream of the transcription start site . The SRE is highly conserved in the orthologous SAG12 gene of rapeseed  and the asparagine synthetase (AS) gene of asparagus  two distantly related taxa. The SRE of the AS gene was aligned to the SAG12 SRE of both Arabidopsis and rapeseed and a highly conserved seven base pair region was identified within the well conserved 33 base pair region (Figure 2.1).
The goal of this study was to examine a collection of soybean genes encoding an AS, transcription factors (MYB, MADS and MYB60), proteases (SAG12 and SAG2), a lipase (SAG101), and a transmembrane receptor kinase (SARK) for the presence of a conserved SRE promoter element. These genes were chosen because of experimental evidence implicating them in the senescence program of various plant species. These genes were used to select soybean orthologs for a comprehensive bioinformatic analysis to identify an associated, potentially functional, SRE element. Candidate genes meeting defined structural criteria were analyzed functionally to elucidate a role in the soybean senescence pathway.
Three different genes (g, D1, D2) are thought to be involved with regulating the whole plant/leaf senescence program of soybean because non-functional mutations in any one of them contributes to an evergreen (non-senescing) phenotype. The fact that mutations in these genes are known to alter the progression of the normal senescence program suggests that they function as major regulators of at least portions of the overall senescence pathway . Homozygous d1d1d2d2 lines remain green and show an inhibition of chlorophyll degradation and chlorophyll-binding proteins yet they still undergo a decline in photosynthetic activity and leaf abscission (Guiamet et al.1991. When combined with the dominant mutation, G (GGd1d1d2d2), a decline in photosynthetic activity fails to take place but the leaves still abscise. It has been suggested that d1d2 may control a central regulatory process in the senescence program . Nearisogenic lines are available as single, double and triple mutant combinations for these three genes . This material can serve as the basis for expression analysis of candidate genes during senescence. As patterns of altered expression are characterized, a better understanding of the genetic regulation of the senescence pathways will unfold. Seven near-isogenic lines of soybean (Harosoy, L64-2489, L69-4266, L69-4265, L69-971, L69-4267 and, L73-54;  representing the SAG family [SAG12 (7) and SAG2 (2) previously implicated in the senescence program. Another gene, Glyma10g11200.1, annotated as a peroxidase harboring the defined SRE but matching the 70 base pair sequence exactly was also examined. Functional analysis, however, did not yield evidence of a differential expression profile (Figure 2.9).
MADS28, another gene previously identified and determined to be involved in the senescence program, yielded conflicting results. While this study determined that 11MADS28, [which harbors an SRE (Glyma11g36890.1) and not a non-SRE containing duplicate version 18MADS28 (Glyma18g00801.1)], is the copy with a likely role in the senescence program, a discrepancy exists between  and the current study. In the previous study a differential expression pattern between wild type Harosoy (gD1D2) and the triple mutant (Gd1d2) showed a perceptible upregulation of the transcript in the wild type isoline at the R7 developmental phase. However, in the current study, upregulation was evident in the triple mutant isoline of the R7 developmental phase (Figure 2.4). A possible explanation for this observation may be the primers used for analysis. The previous study used primers that spanned the fifth and sixth exon junction and the current study used primers that spanned the sixth exon and 3' UTR junction.
While none of the genes examined functionally in this study exhibited differential expression, the method used to derive the candidates has been well developed to serve as an efficient tool for gene mining. Rather than relying on costly and time consuming methods such as microarray analysis, this approach allows for a rapid evaluation of large numbers of orthologs from various species. It incorporates data from various platforms and allows for the integration of information even with continually updating databases. There remain many genes from various gene families in soybean that have been identified as potentially involved in the senescence program using this method but await further functional analysis.

Selection of Candidate Genes for Functional Analysis
The strategy to identify candidate soybean orthologs harboring an SRE involved following guidelines for analysis of 8 specific structural criteria. Genes selected for functional analysis were required to: 1) exhibit an E-value less than 1.0 e-10 in a TBLASTN search; 2) exhibit sequence identity greater than or equal to 48.5% to the SRE 33 base pair consensus; 3) exhibit sequence identity greater than or equal to 71.4% to the SRE 7 base pair consensus; and 4) exhibit a proximity of the SRE sequence upstream to the transcription start site of the gene similar to the model SAG12 gene. Gene architecture, polypeptide sequence identity, phylogenetic and syntenic relationships were also evaluated, but were given less emphasis than the previous four criteria for determining orthology (Figure 2.10). A total of 9 gene types representing 6 gene families were screened for soybean orthologs or homologs ( The ortholog and homolog soybean candidate genes produced from both global and local alignments were then subjected to nucleotide alignments using CLUSTALW and Geneious alignment programs to obtain the best possible nucleotide sequence comparison in a 3000 base pair promoter region upstream of each gene (Figure 2.12). automatically determine sequence's direction (build guide tree via alignment not used) and refinement iterations, 2.
Alignments were performed using the entire gene sequence and 3000 base pair upstream region to confirm the 33 base pair SRE and that the 7 base pair conserved sequence within the SRE was indeed in the region upstream of the transcription start site. Alignments were omitted from further structural analysis if they did not have an SRE alignment of greater than or equal to 48.5% (16 of 33), a 7 base pair alignment of greater than or equal to 71.4% (5 of 7), and were not located in the upstream region.
Gaps in the 33 base pair alignments found in the promoter were deleted and inspected to determine if they met the defined alignment criteria and discarded if they did not.
This analysis was performed for all genes in this study regardless of identification method. Soybean SARK is the only gene whose homologs did not produce candidates meeting these criteria.

Gene Architecture
Comparative gene architecture was examined to determine if conservation of exon/intron structures existed between the gene used for the search and the putative soybean ortholog. The publically available database Gramene (http://gramene.org) was used to search for the gene of interest and the ortholog transcript structures so that a comparison could be made between the two. While no quantitative criteria were established to define of an evolutionary relationship, this analysis contributed to ultimately selecting genes for functional testing. Similarities in length of genes, number of exons/ introns, and nucleotide alignments of over 50% in exons (performed using Geneious alignments with sequences obtained through global and local alignments) were used with other evidence aid in supporting a claim of orthology with the genes examined in this study (Figure 2.13). The sequence for the TMYB was generated from within the soybean genome but because individual one soybean gene could definitively be identified as representing this sequence, a comparison of gene structures could not be achieved.

Polypeptide Sequence Alignment
To further corroborate orthologous relationships, polypeptide sequence identity was examined. As with gene architecture, no quantitative criteria were established to define a relationship but rather the alignments were used to substantiate the previous evidence for an orthologous relationship. Genes were examined to ensure that domains aligned properly with known sequences and a percentage of identical sites was used to assist in selecting genes for functional testing (Figure 2.14).

Phylogenetic Analysis
A phylogenetic approach was undertaken to provide additional insight into any relationships between genes of different species. Soybean genes harboring an SRE within nodes in the same clade as the ortholog of each gene type were considered stronger candidates. To construct trees, the peptide sequence of the model for ortholog/homolog discovery of each gene type was used to screen the Phytozome database using TBLASTN against the Arabidopsis, rapeseed, rice, common bean, sorghum, and corn genomes. The best overall HSP result returned for each species was selected and the amino acid sequences imported to the Geneious program for phylogenetic evaluation. The peptide sequences for these genes and all soybean genes identified during global and local alignments were then aligned followed by evaluation of the phylogenetic tree (Figure 2.15). The default settings of Geneious Tree Builder were as follows: cost matrix, Blosum62, gap open penalty of 12, gap extension penalty of 3; alignment type, global alignment with free end gaps; genetic distance model, Jukes-Cantor; tree build method, Neighbor-joining; and no outgroup. Tree nodes were then examined to see how orthologous soybean genes selected for harboring an SRE clustered with genes presumed to be orthologs or homologs.

Syntenic Analysis
The final criteria used to aid in the selection of orthologs in soybean for functional analysis was synteny. The Synfind portal of CoGe was used to screen A.
thaliana orthologs and soybean homologs against the soybean genome. A. thaliana orthologs and soybean homologs were entered in the name field of the Specify Feature section and the appropriate coding sequence (CDS) selected. In the organism name field under the Specify Organisms section, the genome to be queried, Glycine max, was entered and the appropriate annotated version selected. Default settings were used for the Configure Parameters tab before running the analysis (A. thaliana Col. vs. soybean). Genes were determined to share a "chromosomal neighborhood" if 4 genes existed with similar structures and available annotations were present. More genes in the same syntenic region strengthened the likelihood of an orthologous relationship for this analysis. No results from syntenic analysis decreased the likelihood of an orthologous relationship, but no gene was eliminated based solely on this finding.
Running the analysis in reverse (soybean vs. A. thaliana) did not necessarily guarantee the same results. This can be attributed to the difference in levels of amino acid identity between the two genes in question. As with the gene architecture, polypeptide sequence identity and phylogenetic analysis, and results (or lack thereof) of syntenic analysis (Figure 2.16) were combined with the other analyses to determine the best candidate(s) for functional analysis.

Primer Design
Gene specific primers for the selected candidate soybean orthologs were designed using Primer 3 software (http://www.ncbi.nlm.nih.gov/tools/primer-blast/) available through the NCBI portal. Primers were designed to span either the 5' or 3' UTR and exon junction for unique products ranging from 100 to 300 base pairs.
Default settings were used except for the Exclusion tab which was then checked and Organism field which was changed to "Glycine max". If no 5' or 3' UTR was available, exon regions outside of conserved domains were used to generate unique products. The 5 sets of results returned were then used to query the Phytozome soybean database using TBLASTN to ensure the sequence was within the gene of interest and to verify the primers were unique in the genome (Table 2.4)

RNA Extraction
Total leaf RNA was isolated from several developmental stages throughout the normal life cycle using Harosoy and isogenic lines harboring genes that affect leaf senescence (Table 2.1). Leaves were harvested immediately into liquid nitrogen and RNA was extracted the same day, using a standard phenol/chloroform extraction protocol and LiCl precipitation (Maniatis et al., 1986;).
RNA was quantified using a spectrophotometer and assayed for quality and quantity by electrophoresis in formaldehyde-containing agarose gels (Figure 2.17). The soybean developmental stages V5 and R7 were used in this study. V5 (vegetative 5) is characterized by five fully expanded, green trifoliate leaves are found on the plant and in R7 (reproductive 7) one major pod has changed to a brown color on the main stem (Table 2.5) (http://www.ag.ndsu.edu/pubs/plantsci.htm).

RT-PCR analysis
Preliminary RT-PCR analysis was performed with total RNA from the V5 and R7 developmental stages of soybean leaves from the seven different isogenic backgrounds, with the described experimental primer sets (Tables 2.6 and 2.4). RT-PCR analysis was performed using the AccessQuick RT-PCR System (Promega, Madison WI) with 1ug of total RNA in each reaction. Reaction controls were performed with actin-specific (Sac3) primers (Table 2.4) on each of the RNA samples tested. The Reverse Transcription cycle was run at 45º C for 45 min for one cycle (Eeppendorf thermocycler model 5331). The PCR was carried out as follows: an initial denaturation step at 94º C for 4 min; 40 cycles with 1 min at 94º C, 1 min at 48º C, and 2 min at 72º C; final extension step of 72º C for 7 min; and finally, a hold at 4º C to complete the program. This program was used for all the primer sets analyzed. The products were screened using a 2% molecular biology grade agarose (Fischer Scientific) gel in 100 mM Tris-acetate and 2 mM EDTA.

Sequencing
Verification of the identity of the amplicons was performed through sequence analysis. Bands were excised from the gel and purified with the Wizard SV Gel and PCR clean-up system according to the manufacturer's instructions (Promega). Bands were sequenced using the facilities at the University of Rhode Island's Genomics and Sequencing Center (URIGSC). A highly senescence-specific promoter has been identified in the Arabidopsis thaliana SAG12 gene. A subsequent study identified a conserved 7 base pair sequence within the 33 base pair Senescence Response Element (SRE). These two sequences serve as the basis for selection of candidate genes. Genes selected for functional analysis were required to exhibit A-D. E-H were also evaluated but with less emphasis than the previous four criteria.     Jukes-Cantor Neighbor Joining Phylogenetic Tree made using the Arabidopsis thaliana MYB60 peptide sequence and soybean sequences found using TBLASTN and Smith and Waterman alignments. Highlighted is the node that contains Glyma19g29750.2 and its duplicate which cluster with AtMYB60 suggesting an evolutionary relationship. Highlighted pink boxes represent the syntenic relationship observed between AtMYB60 (At1g08810.1) and its ortholog in soybean (Glyma19g29750.2). Three other genes within a 60K region are also observed with similar functional annotations. This evidence implies a close evolutionary relationship between the two chromosomal "neighborhoods" where the gene resides.  ). The gel shown is a representative agarose gel of RNA extracted from leaf tissue of seven isogenic lines of soybean (Table 1). All RNA's used in this study exhibited similar quality.  A database and literature search for genes associated with senescence in other plant species was conducted. A list was compiled and screened for soybean orthologs for which 3000 base pairs upstream were then evaluated for the presence of an SRE and conserved element. The Arabidopsis thaliana MYB60 gene (At1g08810.1) (starred) will serve as an example of a typical gene screening strategy in this report.  **R7-The developmental stage at which one major pod has changed to a brown color on the main stem (www.ag.ndsu.edu/pubs/plantsci.htm).