FRISTENSKY LAB RESEARCH

Figure 1. Evolution of gene expression patterns.

I. GENE EXPRESSION IN PLANT DEFENSE RESPONSES

Defense responses: The global picture

We wish to determine whether there are distinct "themes" of gene expression underlying different types of resistance, such as resistance moderated by single loci, basic resistance governed by many loci, and resistance to different types of pathogens, such as biotrophs versus necrotrophs. As illustrated in Fig. 1, any given gene may be differentially regulated with respect to pathogen. The expression pattern in a would be typical of the response of a non-host to an incompatible fungus, while the same gene might be repressed by a compatible fungus (c).

During my Ph.D. work, genes were found which hybridized more strongly with in-vivo  labeled RNA populations from Fusarium solani -treated plants than with control RNA [ 4 , 5 ]. In a sense, this was one of the early uses of gene arrays, leading to the first clones for pathogenesis-related protein PR10 (pI49, pI176) [ 4 ], defensin (pI230, pI39) [5 ] and DRR206 (pI206) [4 ]. At the time, the absence of automated sequencing limited this study to a few genes.

Today, Expressed Sequence Tags (ESTs) provide a cross-section of genes expressed in a given developmental or environmental context. Our lab has sequenced cDNAs from genes expressed in canola ( Brassica napus ) in a resistant interaction with the blackleg fungus (Leptosphaeria maculans ) [ 20 ]. Among the ESTs we have found homologues of the N (tobacco) and Lr 10 (wheat) resistance genes, defense genes (polygalacturonase inhibitor, CXc750, and PAL), non-defense related genes (eg. cab, RUBISCO) and others of unknown function.

We are currntly employing gene arrays to measure gene expression in mRNA populations from Brassica expressing resistance or susceptibility to Brassica -specific pathogens, or basic resistance to pea pathogens. The diversity of the gene array should allow us to detect processes that would not ordinarily be uncovered by focusing only on defense genes. For example, Hahlbrock and coworkers have shown that cell cycle-related genes such as histones, cdc2 and cyclin are down regulated in parsley by Phytophthora or elicitor [Logemann (1995) Plant J . 8:865-876]. Possibly the most exciting results will come from genes of unknown identity whose expression patterns match genes of known function (eg. cell death). More intriguingly, a coordinately-regulated set of unknown genes could point to a previously unknown pathway.

Defense responses: Gene expression evolves too

A fundamental principle in evolutionary theory is that natural selection can only act if polymorphism exists in a population. Variations in protein function, timing of expression, or response to stimuli are potential sources of phenotypic variation within a population. Figure 1, panels b and d show that expression of a defense gene might vary within a species, or between closely-related plant species. We refer to such variations in expression as "regulatory polymorphism". In particular, regulatory polymorphism among groups of defense genes could play a role in plant-pathogen coevolution. However, the ways in which gene expression evolves, and the time scales over which that evolution takes place, have not been studied.

The redundancy of multigene families makes it possible for a plant to experiment with gene expression by minimizing the risk of loss of function.  Interestingly, most genes activated in plant defense responses are present in multigene families. Phylogenetic analysis of over a dozen defense multigene families (eg. PR10 [ 4 , 5 , 8 , 10 , 19 ], PR1, chitinase, osmotin etc.) across a broad range of plant species shows that within a species, genes in multigene family are more closely related to each other than to homologous genes from other species [35 ]. The rapid evolution of defense genes implies that differential expression patterns must be frequently reassigned as family members are duplicated or deleted over time.

1. Divergence of infection phenotype
Pea tissue expresses basic resistance to the bean pathogen Fusarium solani f. sp. phaseoli within 8 hours post inoculation (h.p.i.), while it is susceptible to invasion by the pea pathogen F. solani f. sp.. pisi [5 ]. At the same time, both fungi exhibit limited growth on the closest relative of pea, P. humile , while even more rapid and prolific hyphal growth occurs on  the more distantly related P. elatius and P. fulvum [ 29 ].

2. Divergence of PR10 genes [30 ]
In order to study how a defense multigene family evolves, we have sequenced PCR products for PR10.1, PR10.2 and PR10.3 orthologues from P. humile , P. elatius , and P. fulvum . Phylogenetic analysis (Fig. 2) indicates that PR10.1, PR10.2 and PR10.3 diverged very recently, probably in the common Pisum ancestor. Orthologous copies of PR10.1, PR10.2 and PR10.3 exist in P. humile (Ph), P. elatius (Pe) and P. fulvum (Pf), except that no PR10.2 orthologue was identified in P. elatius (Fig. 2). In contrast, the divergence of PR10.4 and PR10.5 appears to have occurred earlier in some ancestral legume.

3. Divergence of PR10 gene expression  [30 ]
Regulatory polymorphism among PR10 genes was seen with both F. solani f. sp. phaseoli and F. solani f. sp. pisi. For the recently diverged PR10.1 and PR10.2 genes, the timing of fungal-induced expression differs greatly among species. For example, PR10.1 was strongly induced in P. sativum by F. solani within 8 h.p.i., whereas little PR10.1 expression was seen in pea's closest relative, P. humile , and in the more distantly-related P. elatius . In P. fulvum , expression did not peak until 48 h.p.i. PR10.2 was expressed in P. sativum and P. humile , but not in P. elatius or P. fulvum. In contrast, PR10.4 and PR10.5, which appear to have arisen by duplication in some ancestral legume, have distinct expression patterns that are conserved among Pisum species (Fig. 3). This was the first documentation of divergence of expression in orthologous copies of genes between species.

Despite the observation that PR10.1 and PR10.3 diverged from a single gene recently, each gene contains a different motif bound by nuclear protein(s) within 2 h.p.i. with F. solani f. sp. phaseoli [ 32 ]. This observation combined with the fact that PR10.3, while not expressed in pods, is expressed in roots, shows that components of defense gene regulation can change drastically over short evolutionary timescales.

Our long term goal is to understand the role of the evolution of gene expression in disease resistance. Gene arrays will enable us to broaden these studies to a large number of genes. Do some genes or pathways become fixed in their expression over time? If so, other cellular processes could come to depend upon their expression. In contrast, newly-duplicated copies of a gene would be under few constraints, and their expression would be at liberty to evolve. Are particular genes specialized for defense against specific pathogens? For example, would one gene be induced by fungi, and another by bacteria? Does regulatory polymorphism play a role in the evolution of defense responses?

II. CONSTITUTIVE EXPRESSION OF DEFENSE GENES CONFERS RESISTANCE TO BLACKLEG DISEASE IN TRANSGENIC BRASSICA NAPUS [ 21 ].

Figure 4. Transgenic plants containing pea DRR206 (right) exhibit lush growth and negligible lesions when inoculated with blackleg fungus. Plants transformed with pea defensin (middle) also exhibit decreased lesion formation, compared to controls. Untransformed control plants die after fungal treatment (left).

To identify genes for genetic engineering of disease resistance, we have transformed Brassica with constitutively-expressed pea PR10, chitinase, DRR206 and defensin, and with Brassica napus PR1. Research associate Yaping Wang has demonstrated that plants containing either pea DRR206 or defensin showed almost no symptoms in stem inoculation assays, compared with untransformed cv. Westar, correlating with high levels of mRNAs for these genes. Resistance cosegregates with DRR206 in a 3:1 ratio. Microscopy reveals that DRR206 transgenic plants inhibit germination of blackleg. Extracts from DRR206 and defensin transgenic plants inhibited fungal germination in-vitro. DRR206 plants also demonstrated decreased hyphal  growth that inoculation sites. More recent results indicate that DRR206 transgenic plants have strong resistance to Rhizoctonia , and delayed lesion development with Sclerotinia [ 25 ].

Lewis et al. (1999; Chem. & Biol. 6: 143) have identified DRR206 homologues in Forsythia as extracellular (+)-pinoresinol synthases, which catalyze an early step in lignin production. We propose three hypotheses for DRR206 in plant defense:

1. DRR206 protects plants through lignification of plant cell walls

2. DRR206 protects plants through lignification of fungal hyphae, preventing growth

3. DRR206 encodes a phenolic compound with antifungal activity.

III. BIOINFORMATICS

An Integrated and Distributed Bioinformatics Platform for Genome Canada
 http://www.genomeprairie.ca/research/bioplatform.htm

SubProject Title: Bias in sequence databases
Beyond acting as repositories of biological information, databases become models of the population from which the data is derived. For example, we would like to be able to use databases such as the GenBank nucleic acid database to make inferences about gene evolution in real species. However, these databases contain biases inherent in the fact that their data is collected from directed experiments, rather than random sampling. The research will examine the following questions:

1. What kinds of biases exist in biological databases?

2. How can bias be quantified?

3. Which analytical methods are affected by bias, and how are they affected?

4. How can analytical methods be adapted to correct for the biases inherent in biological databases?

Figure 5. Maximum likelihood tree of DRR206 coding sequences.

EXAMPLE: We originally cloned plant defense gene DRR206 from peas expressing resistance to Fusarium solani [4,10]. This gene was not identified in any other species until recently, when it was independently cloned in a number of woody shrubs, and in Arabidopsis. In the dataset of DRR206 genes, the bias is therefore toward woody shrubs. As well, the predominant number of DRR206 homologues come from one species, Thuja plicata.

These biases can have two independent effects on the results. First, the multiple alignment used as input to the phylogeny may be weighted toward woody shrubs. Secondly, the species bias may have also affected the construction of the phylogenetic tree. That is, if DRR206 sequences were available from a more diverse set of species, a different branching relationship might be seen even among those species already present in the existing tree.
In databases such as GenBank, SwissProt or PDB, we have seen that parameters like sequence length or number of sequences per taxonomic group appear to be described by an Extreme Value Distribution (EVD), which is characterized by a strong peak at low values a long tail at high values. We are developing metrics that allow us to measure the biases in these types of distributions, and to compare biases between different distributions. For example, the mean protein lengths reported in SwissProt for bacteria appears to be  smaller than for plants or animals. Given a good set of metrics for comparing these types of distributions, we will study the effects of database bias on analytical methods, such as multiple sequence alignment, phylogeny, pattern discovery and recognition of regulatory motifs, and training of gene recognition algorithms. These investigations will include both real datasets, as well as simulated datasets with specific biases deliberately introduced.  The results of these experiments should point to approaches for eliminating the effect of biases on analytical methods, through normalization of datasets or modification of existing analytical methods.

Biological Research Computing Hierarchy (BIRCH) [22 , 26 , 27, 33 ]
http://home.cc.umanitoba.ca/~psgendb

Since 1990 I been developing and operating the BIRCH bioinformatics facility on our campuswide Sun Unix system. BIRCH provides software and databases for use in molecular biology free of charge to researchers in the University community. BIRCH is unified through Steven Smith's Genetic Data Environment (GDE), a graphic user interface that provides a seamless flow of data between programs. BIRCH now serves more than 140 molecular biologists on the Ft. Garry and Medical campuses. 

BIRCH can now be downloaded for installation at other sites. See http://home.cc.umanitoba.ca/~psgendb/birchadmin/birchadmin.html

PUBLICATIONS

1. Fristensky, B., Lis, J.T. and Wu, R. (1982) Portable microcomputer software for nucleotide sequence analysis. Nucleic Acids Research 10:6451-6463.

2. Yang, R.C.A., Fristensky, B., Deutch, A.H., Huang, R.C., Tan, Y.H., Narang, S.A., and Wu, R. (1983) The nucleotide sequence of a new human repetitive DNA consists of eight tandem repeats of 66 base pairs. Gene 25:59-66.

3. Hadwiger, L.A., Fristensky, B., and Riggleman, R.C. (1984) Chitosan, a natural regulator in plant-fungal pathogen interactions, increases crop yields. in Chitin, Chitosan, and Related Enzymes ,pp 291-302. John P. Zikakis,Ed., Academic Press.

4. Riggleman, R.C., Fristensky, B., and Hadwiger, L.A. (1985) The disease resistance response in pea is associated with increased levels of specific mRNAs. Plant Molecular Biology 4:81-86.

5. Fristensky, B., Riggleman, R.C., Wagoner, W., and Hadwiger, L.A. (1985) Gene expression in susceptible and disease resistant interactions of peas induced with Fusarium solani pathogens and chitosan. Physiological Plant Pathology 27:15-28.

6.Fristensky, B. (1986) Improving the efficiency of dot-matrix similarity searches through use of an oligomer table. Nucleic Acids Research 14:597-610.

7. Kendra, D., Fristensky, B., Daniels, C.H., & Hadwiger, L.A. (1987) Disease Resistance Response Genes in Plants: Expression and Proposed mechanisms of induction. in Molecular Strategies for Crop Protection pp. 13-24. Alan R. Liss, Inc.

8. Daniels,C., Fristensky, B., Wagoner, W. and .Hadwiger, L.A (1987) Pea genes associated with non-host resistance to fungi are also active in race-specific resistance to bacteria. Plant Molecular Biology 8:309-316.

9.Moody, M. & Fristensky, B. (1987) Database bias and the identification of protein coding sequences. DNA 6:493-495 

10. Fristensky, B., Horovitz, D., and Hadwiger, L.A. (1988) cDNA sequences for pea disease resistance response genes. Plant Molecular Biology 11:713-715.

11. Elliot, R.C., Pedersen, T.J., Fristensky, B., White, M.J., Dickey, L.F. and Thompson, W.F. (1989) Characterization of a single copy gene encoding ferredoxin I from pea. Plant Cell1, 681-690.

12. Falconet, D., White, M.J., Fristensky, B.W., Dobres, M.S. and Thompson,W.F. (1991) Nucleotide sequence of Cab-215, a Type II gene encoding a photosystem II chlorophyll a/b-binding protein in Pisum.Plant Molecular Biology 17:135-139.

13. Alexander, L. Falconet, D., Fristensky, B. W., White, M. J., Watson,J.C., Roe, B.A. and Thompson, W.F. (1991) Nucleotide sequence of Cab-8 , a new type I gene encoding a chlorophyll a/b -binding protein of LHCII in Pisum . Plant Molecular Biol. 17:523-526.

14. White, M.J., Fristensky, B.W. and Thompson, W.F. (1991) Concatemer chain reaction: A Taq DNA polymerase-mediated mechanism for generating long tandemly-repetitive DNA sequences. Anal. Biochem . 199:184-190.

15. Fristensky, B. (1992) BIRCH: Biological Research Computer Hierarchy. User manual. (Unpublished).

16. White, M.J., Fristensky, B.W., Falconet, D., Childs, L.C., Watson, J.C., Alexander, L., Roe, B.A. and Thompson, W.F. (1992) Expression of the chlorophyll a/b protein multigene family in pea (Pisum sativum L.): Evidence for distinct developmental responses. Planta 188:190-198.

17. Fristensky, B. (1993) Feature Expressions: Creating and Manipulating Sequence Datasets. Nucleic Acids Res. 21:5997-6003. 

18. Mailer, R.J., Scarth, R. and Fristensky, B. (1994) Cultivar discrimination of Brassica napus using RAPDs. Theoretical and Applied Genetics 87:697-704.

19. Culley, D.E., Brown, S.M., Parsons, M.A., Hadwiger, L.A. and Fristensky, B. (1995) Cloning and sequencing of disease-resistance response gene DRR49a (Ypr10.PS.1) from Pisum sativum . Plant Physiology 109:722.

20. Fristensky B, Balcerzak M, He D-F, Zhang  P, (1999) Expressed sequence tags from the defense response of Brassica napus to Leptosphaeria maculans. Mol. Plant Pathol. Online http://www.bspp.org.uk/mppol/1999/0301FRISTENSKY .

21.Wang, Y., Nowak, G., Culley, D. Hadwiger, L.A. and Fristensky, B. (1999) Constitutive expression of pea defense gene DRR206 confers resistance to blackleg (Leptosphaeria maculans) disease in transgenic canola ( Brassica napus). Mol. Plant-Microbe Interact. 12: 410-418.

22.Fristensky, B. Building a multiuser sequence analysis facility using freeware. In Misener S, Krawetz S. (1999) Bioinformatics Methods and Protocols , pp. 131-145, Humana Press.

23.Fristensky, B. Network computing: Restructuring how scientists use computers and what we get out of them. In Misener S,Krawetz S. (1999) Bioinformatics Methods and Protocols, pp 401-412, Humana Press.

24. Fristensky, B .  Network-centric computing in genomics. AgBiotechNet Vol. 1, ABN032, November (1999) http://www.agbiotechnet.com . Can be downloaded at http://home.cc.umanitoba.ca/~frist/Papers/ABN032/ABN032.html

25.Wang, Y. and Fristensky, B. (2001) Transgenic canola lines expressing pea defense gene DRR206 have resistance to aggressive blackleg isolates and to Rhizoctonia solani. Molecular Breeding 8: 263-271.

26.Fristensky, B (2003) Installing bioinformatics software in a server-based computing environment. In Krawetz, SA and Womble DD, Introduction to Bioinformatics . pp 285-296. Humana Press, USA.

27.Fristensky, B (2003) Bioinformatics software management in a server-based computing environment. In Krawetz, SA and Womble DD, Introduction to Bioinformatics . pp 297-306. Humana Press, USA.

28.Tang J-H, Fristensky B and Scarth R (2003) Effects of  genomic position and copy number of Acyl-ACP thioesterase transgenes on the level of the target fatty acids  in Brassica napus L. Molecular Breeding  12: 71-81.

29. Tewari S, Brown SM, Fristensky B (2003) Plant defense multigene families: I. Divergence of Fusarium solani-induced expression in Pisum and Lathyrus http://arXiv.org/q-bio.PE/0310003

 30. Tewari S, Brown SM, Kenyon P, Balcerzak M, Fristensky B (2003)  Plant defense multigene families. II Evolution of coding sequence and differential expression of PR10 genes in Pisum. http://arXiv.org/q-bio.PE/0310038

31. Srivastava S, Fristensky B, Kav N (2004) Constitutive expression of a PR10 protein enhances the germination of Brassica napus under saline conditions. Plant and Cell Physiol. 45:1320-1324.

32. Srivastava S, Emery RJN, Kruepin LV, Reid DM, Fristensky B, Kav NNV (2006) Pea PR 10.1 is a ribonuclease and its transgenic expression elevates cytokinin levels. Plant Growth Regul. 49: 17-25.

33. Fristensky B (2007)  BIRCH: A user-oriented, locally-customizable, bioinformatics system. BMC Bioinformatics , 8:54

In preparation or submitted

34. Zhang, Y-P. and Fristensky, B. (2006) Analysis of cis-regulatory elements in differential expression of the PR10 multigene family in peas. (in preparation)

35. Zhang, P., Tu, J. and Fristensky, B. (2006) Early events in resistant and susceptible interactions between Brassica napus and Leptosphaeria maculans. (in preparation).

SELECTED ABSTRACTS

36. Fristensky, B. (1995) Evolution of defense multigene families and its consequences for plant/pathogen compatibility. Phytopathology 85:1132. American Phytopathological Society Meeting, Oral presentation 138.

37. Fristensky, B. (1996) Evolution of multigene families encoding plant defense proteins: Implications for host population diversity. International Plant Genome Conference IV, San Diego CA, USA, Jan. 14-18, 1996. Poster P280.