Genes of Immediate Interest Based on Their Specific Functions in Our

Current Working Models

We begin below with a general discussion of our strategies for the choice of genes for silencing and then describe more details regarding our particular target genes (see table).

Please note that this page is under relatively frequent revision as we refine our knowledge of each gene family.

Choice of Specific Genes for Silencing and Development of Silencing Sequences

 In developing sequences to use in the silencing of a given gene or family of genes, one must consider one’s objectives for a given experiment.  In certain cases, it might be of interest to silence only the expression of a specific gene (or member of a family of genes), while for other experimental objectives, it might be more important to silence multiple (or even all) members of a gene family.  One of the strengths of post-transcriptional gene silencing over mutation is that the silencing of multiple genes within a family can often be achieved with a single silencing construct, whereas it would take the “stacking” of many independent mutations to achieve the same result through classical mutation.

 Silencing of a Specific Gene

 If one is attempting to silence a very specific gene (perhaps it is the only gene with this function, or it is the only one expressed in a particular tissue and/or under a particular condition of interest to you), then one can develop a silencing construct with high specificity to that target gene.  However, an important consideration in taking such an approach is that other members of the same gene family or even less related genes with functional redundancy to the target gene may functionally substitute for the gene product’s activity and silencing of the phenomenon of interest will not occur even though effective silencing of the target gene has been successful.  The same limitations are true, of course, for mutational analysis, and thus this is not necessarily a problem, but simply something that must be kept in mind in the analysis of results of an experiment.  If effective silencing of a specific gene is seen (e.g., via quantitative RT-PCR) and the phenomenon of interest is strongly or totally compromised, like mutational analysis this is good supporting evidence for the importance of the particular gene in the phenomenon under observation.  On the other hand, if the phenomenon is not compromised or is only partially compromised, as with mutational analysis one cannot make the simple conclusion that the gene is not involved without further experimental support.  Despite these limitations, the silencing of a specific gene can be very powerful if that gene’s function is important under the very specific conditions of the experiment being done (e.g., response to a specific elicitor).

 Silencing of Multiple Genes of a Given Function

 As noted above, one of the strengths of gene silencing is that it is potentially easier to silence multiple (if not all) members of a functional gene family.  In certain cases, then, it may be possible to completely silence a given biological or biochemical function and examine its effects on the phenomenon of interest.  A good example, described in more detail below is the silencing of both genomic copies of isoflavone synthase (IFS), which we have shown has led to near complete suppression of isoflavone accumulation (1,2).  A complication of any gene mutation or silencing approach, but which may be more common to multiple gene silencing, is that completely knocking out a biochemical function could lead to serious if not lethal effects on cellular function.  In the case of IFS we saw no such deleterious effects on soybean plants, suggesting that isoflavones are not critical to normal cell viability or growth.  In working with mutations, researchers have sometimes tried to avoid such effects through the use of conditional effects (e.g., temperature sensitive mutations).  Likewise, to avoid such "pleiotrophic" problems with gene over-expression, transient expression assays are sometimes employed.  In our particular experiments with Agrobacterium rhizogenes mediated RNAi gene silencing of roots, we are not making permanent changes to the plant germplasm.  This has two potential advantages: 1) plant-wide changes to gene expression are not made and 2) the introduction of silencing constructs and observations of their effects is normally observed over short periods.  Both of these factors can minimize the chance for deleterious effects.  While our systemic silencing protocols allow us to extend silencing to organs other than roots, these silencing effects are transient (2), once again potentially minimizing pleiotrophic effects.

 Thus, post-transcriptional gene silencing is a powerful tool and has some important advantages that make it nicely complementary to mutational analysis.  In particular, it can be very rapid and high throughput and can be used to simultaneously silence the expression of functionally equivalent genes.

 Genes for Which Successful Silencing Has Already Been Accomplished in Our Laboratory:

 Isoflavone synthase (IFS).  IFS is the entry enzyme into the pathway that leads to all isoflavonoids.  There are two genomic copies of IFS, both of which have been cloned.  The two genes share over 90% homology at the nucleic acid level.  By choosing a sequence from a highly conserved region of the two genes (1), we have successfully developed a high efficiency A. rhizogenes mediated silencing protocol that has led to 95-100% reduction of isoflavone levels in root tissues and nearly 80-100% systemic silencing of induced isoflavones in non-transformed tissues (2).  The silencing of IFS has led to marked effects on isoflavone responses to elicitor and infection of both roots and cotyledon tissues by Phytophthora sojae (2).

 Chalcone reductase (CHR).   Chalcone reductase is required for the reduction of chalcone to isoliquiritigenin, the precursor of the isoflavone daidzein.  There are 3 TCs which correspond to chalcone reductase activity (Oliver Yu, unpublished).  Silencing expression of this enzyme should block the accumulation daidzein, while not affecting genistein accumulation.  Indeed, we have developed a silencing construct effective against all three CHR genes and while daidzein accumulation is reduced by >90%, accumulation of the alternate isoflavone, genistein, is enhanced (Oliver Yu, unpublished).

Genes for Which Silencing Efforts are Planned or Underway:

 Isoflavone-specific b-glucosidase (ISBG).  We have isolated a highly specific 7-O-isoflavone-b-glucosidase from soybean (3).  To illustrate its high level of specificity, it is nearly inactive on flavonol glycosides or a wide range of synthetic glucosidase substrates. Throughout its purification, we only detected one protein band with activity (3).  The corresponding gene has been cloned from roots (M.Y. Graham, unpublished).  One other soybean glucosidase shows close homology to this cloned gene, but the rest that we have identified in the soybean EST database have much lower homology.  Thus, consistent with its high specificity, the gene belongs to a unique class of glucosidases.  We will begin our silencing efforts using consensus sequences unique to these two gene family members.

 Cinnamyl alcohol dehydrogenase (CAD).  There are as many as 8 CAD genes in soybean.  Among these, we found that only 3 are expressed in infection EST libraries and only two of these respond to glucan elicitor treatment in preliminary Northern analyses (M.Y. Graham, unpublished).  For silencing, we will begin with a conserved region for these induced species.

 Type II NADH Oxidase (Nox II). We have identified a specific peroxidase isoform which is induced systemically by the glucan elicitor (4), correlates to induced distal defense potentiation (5) and which possesses Type II NADH oxidase activity (Nox II) that is specifically activated by the isoflavone genistein (6).  This specific peroxidase has been hypothesized by us to function in the establishment of the HR and competency for responses to the glucan elicitor.  Although we have not cloned the gene, a single peroxidase TC is strongly induced in glucan elicitor expression libraries and we hypothesize that this specific gene corresponds to the isoform that we have studied biochemically.  This TC, which (consistent with its unique activity) has distinct areas of sequence divergence from other peroxidases, will be used to develop appropriate silencing constructs. 

NPR-1.   Several years ago, we identified a single EST in soybean with sequence homology to the Arabidopsis NPR-1 gene and made a 500bp probe against it.  In preliminary Northerns using this probe we found that expression of the gene was constitutive, though relatively weak and evenly distributed in soybean seedling tissues, except in roots where it was somewhat (~2X) stronger (M.Y.Graham, unpublished).  Bhattacharya and coworkers have reported two genes in preliminary studies (Soybean 2004, University of Missouri, Columbia, MO, July, 2004).   For our silencing studies, we will use a conserved region from the two genes. 

Kunitz Trypsin Inhibitor, Variant S (KTI-S).  A proteinaceous factor, termed Competency Factor 1 (CF-1) has been implicated in the establishment of competency for the phenolic polymer response to the glucan elicitor (7).  Purification and partial amino acid sequencing of this factor (8) led to its identification as a variant of the Kunitz trypsin inhibitor (KTI).  Although there are as many as 10 family members for KTI, which show a high degree of homology, a single gene, Variant S (KTI-S), which corresponds to CF-1, is quite divergent from all other family members.  This is consistent with the unique role proposed for CF-1 (KTI-S) as a regulatory protein involved in elicitation competence.  We will use sequences from this gene with divergence from the other gene family members for silencing efforts. 

12-oxophytodienoate reductase (OPDR).  While there are several members of this gene family annotated as TCs in soybean, only one OPDR TC is expressed in defense related EST libraries. We will use both gene specific and non-specific constructs for our silencing efforts. 

Elicitor releasing endo-b-1,3 glucanase (ERBG).  Yoshikawa and co-workers originally obtained a cDNA clone for the soybean glucanase that releases cell wall glucan elicitors from Phytophthora sojae based on its elicitor releasing activity (9).  Its sequence (from the cultivar Harosoy 63) is 98% identical to a soybean genomic clone from a different cultivar (Williams 82).  These clones correspond (98% homology) to a specific Group II β-glucanase as defined by Shoemaker et al. (10), who used sequence information to classify the various classes of glucanases in soybean EST libraries.  We derived a probe derived from Yoshikawa’s gene (ERBG) and performed some preliminary Northern analyses, some of which have been published (11).  Although there is some constitutive expression in soybean cotyledons, wounding and/or glucan elicitor treatment caused a dramatic induction.  While there are many glucanases in soybean, only these Group II ones are apparently involved in elicitor response or defense.  The rest are expressed in under different conditions or in specific tissues (e.g., floral tissues) and share much lower homology to ERBG.  We are thus fairly confident that we will be able to line up the three ERBG clones and derive a gene specific sequence for our silencing efforts.  Moreover, our analysis of expression of various glucanases in the soybean EST libraries suggests that the other glucanase classes will unlikely be induced under the conditions we will be using.  This, and their different compartmentalization in the cell (ERBG is apoplastic), makes it unlikely that they will complement the function of ERBG.  Nonetheless we will use quantitative RT-PCR with specific and non-specific primer pairs to analyze the specificity of silencing.   

ACC Oxidase.   Some reports have noted that ACC oxidase is more responsive to induction in host-pathogen interactions than ACC synthase.  Three ACC oxidase TCs were upregulated in defense related soybean EST libraries.  We will use a silencing construct based on the conserved region of these 3 TCs. 

Glucan Elicitor  Receptor (WGER).  Since the P. sojae wall glucan elicitor (WGE) is one of the best defined elicitors, it is not surprising that several labs have worked to characterize its receptor.  Hahn and co-workers first defined the binding activity with a synthetic heptameric ligand.  Yoshikawa and co-workers reported the cloning of the receptor using a glucanase-released elicitor (12).  Later, a cloned cDNA (for the receptor which bound the synthetic ligand) turned out to be the same gene (13).  The sequence does not have signal sequences or transmembrane regions, but has a glucosyl hydrolase domain (14).  Thus, superficially, it’s characteristics are more like an enzyme than a receptor per se.  Whether the protein is part of a receptor complex or the true receptor remains uncharacterized is unknown.  Silencing of this gene would thus yield potentially interesting information as to its involvement.  We have derived a probe for this gene and done preliminary Northern analyses.  Glucan elicitor treatement led to slight upregulation of its RNA (~2X).   

EREBP.  An ethylene response element (ERE) has been defined from consensus sequences of ethylene responsive genes.  The corresponding class of transcription factors, which bind to this element, is termed EREBP.    In Arabidopsis, 6 EREBPs have been studied.  They were expressed fairly non-specifically.  In tomato, at least 3 EREBPs interact with the Pto R gene (termed Pti 4, 5 and 6).  We found a homolog of Pti6 in soybean.  For our silencing efforts, we will use a conserved AP2 domain in transient silencing experiments.  Many floral related AP2 or EREBPs should not be expressed in cotyledons, so we feel it is unlikely that we will have serious pleiotrophic or deleterious effects. 

WRKY.   There are numerous WRKY transcription factors in any plant that are involved in many diverse processes.  A subset of WRKY has been shown to be involved in defense related gene expression.  We derived a pair of primers based on this class of WRKYs and used a PCR product as a probe in preliminary Northern blots.  These results demonstrated that a WRKY species was transiently induced within 1 hour of wounding.  The primer sets match well with only 2 WRKY TCs.   Because of our primary interest in wound induced elicitation competence, we plan to use a silencing construct based on a conserved region of these two genes.

 1.  Hsieh, M. C. and T. L. Graham. 2001.  Partial purification and characterization of a soybean β-glucosidase with high specific activity for isoflavone conjugates.  Phytochemistry 58:995-1005.

2.   S. Subramanian, M. Y. Graham, O. Yu and TL Graham. 2005.  RNA interference of isoflavone synthase leads to silencing in non-transformed tissues and to enhanced susceptibility to Phytophthora sojae.  Plant Physiology, submitted.

3.  Subramanian S, Hu X, Lu G, Odell JT, Yu O (2004) The promoters of two isoflavone synthase genes respond differentially to nodulation and defense signals in transgenic soybean roots. Plant Mol Biol. 54:623-39.

4.  Graham M Y; Graham T L. 1991.  Rapid accumulation of anionic peroxidases and phenolic polymers in soybean cotyledon tissues following treatment with Phytophthora megasperma f.sp. glycinea wall glucan.  Plant Physiology 97:1445-1455

5.   Park, D.S., Graham, M. Y., Landini, S. and Graham,T.L. 2002.  Induced distal defense potentiation against Phytophthora sojae in soybean. Physiol. Molec. Plant Pathol., 60:293-310.

6.  Graham TL, Graham MY, Rose AR, Poling RS, Omer MA (2000) Identification and distribution of a specific peroxidase isoform functioning as a genistein-activated NADH oxidase central to defense competency in soybean.  Peroxidase Newsletter, 14:103-109.

7.   Graham, Madge Y.; Graham, Terrence L. 1994. Wound-associated competency factors are required for the proximal cell responses of soybean to the Phytophthora sojae wall glucan elicitor. Plant Physiology 105:571-578

8.   Park, D.S., Graham, M. Y. and Graham, T. L. 2001. Identification of soybean elicitation competency factor, CF-1, as the soybean Kunitz trypsin inhibitor.  Physiol. Molec. Plant Pathol. 59:265-273.

9.  Takeuchi Y; Yoshikawa M; Takeba G; Tanaka K; Shibata D; Horino O. 1990. Molecular cloning and ethylene induction of messenger RNA encoding a phytoalexin elicitor-releasing factor beta-1 3 endoglucanase in soybean Plant Physiology 93:673-682

10. Jin, Wei; Horner, Harry T.; Palmer, Reid G.; Shoemaker, Randy C. 1999. Analysis and mapping of gene families encoding beta-1,3-glucanases of soybean. Genetics 153:445-452

11. Graham, MY, Weidner J, Wheeler K, Pelow ML, Graham TL. 2003. Pathogenesis-  Related Protein Gene Activation in Soybean by Wounding and the Phytophthora sojae Cell Wall Glucan Elicitor, Physiological and Molecular Plant Pathology 63:141-149.

12. Umemoto, Naoyuki; Kakitani, Makoto; Iwamatsu, Akihiro; Yoshikawa, Masaaki; Yamaoka, Naoto; Ishida, Isao. 1997. The structure and function of a soybean beta-glucan-elicitor-binding protein. PNAS 94:1029-1034

13. Mithoefer, Axel; Fliegmann, Judith; Neuhaus-Url, Gabriele; Schwarz, Heinz; Ebel, Juergen The hepta-beta-glucoside elicitor-binding proteins from legumes represent a putative receptor family Biological Chemistry 381(8)2000. p.705-713

14. Fliegmann, Judith; Mithoefer, Axel; Wanner, Gerhard; Ebel, Juergen. 2004.  An ancient enzyme domain hidden in the putative beta-glucan elicitor receptor of soybean may play an active part in the perception of pathogen-associated molecular patterns during broad host resistance. Journal of Biological Chemistry 279:1132-1140.