ADEPT Progress Report

Davis, White and Martin, University of Florida
April 2004

Disease Resistance Phenotyping:

Data analysis was completed on the RSC pitch canker screen. Calculations included heritabilities as well as parental, clonal and full-sib family rankings. (UF pitch canker analysis completed spring 2003, progress reported April 2003). Based on the high (0.86) genetic correlation between the RSC and UF pitch canker screens, the datasets were combined and used to identify families and clones representing phenotypic extremes. Families within each phenotypic extreme (five most susceptible and five most resistant) were related, while no relationships were found across the classes, suggesting host responses to F. circinatum are genetically distinct.

The 1-gall (narrow) fusiform rust screen was phenotyped in April 2003. Identical phenotypic measurements were collected as for the 10-gall (broad) screen (number of galls, number of receptive shoots, the number of receptive shoots with galls, and the gall length, gall width and width of the healthy stem for 2 randomly chosen galls per ramet). Phenotypic measurements for each screen were used to generate datasets for disease incidence (score), gall length, gall width, and gall volume. Calculations included heritabilities as well as parental, clonal and full-sib family rankings. The genetic correlation between the 1-gall and 10-gall screens was high (0.8 for score and 1.0 for gall length) and the datasets combined. Genetic correlations between score and gall length, using the combined dataset, were low (0.04), indicating that gall length cannot predict a family’s tendency to form a gall and vice versa.

For the RSC pitch canker and 10-gall fusiform rust screens (largest number of genotypes), the clonal mean-based broad sense heritabilities (0.75 for pitch canker lesion length, 0.72 for fusiform rust score and 0.32 for fusiform rust gall length) suggest good quality phenotypic evaluations of the clones. This metric is appropriate for assessing the extent of genotypic control on phenotypic variation in studies utilizing rooted cuttings.

Correlations between lesion length (pitch canker) and gall length (fusiform rust) were extremely low (0.0), indicating that host responses to pitch canker and fusiform rust diseases are independent.

Verification of pitch canker phenotypes:

Gene expression array experiments were designed to verify pitch canker phenotypes and determine differential gene expression between resistant and susceptible clones. The experimental design consisted of highly susceptible and highly resistant clones, multiple ramets per clone, and harvests at 0 and 8 days after challenge. Inoculations and tissue collections have been completed and will be used to synthesize probes for microarray analyses scheduled for late summer 2004.

Identification of Additional Disease Resistance Candidate Genes:

Consensus at the 2003 annual meeting was to identify additional candidate genes with a role in resistance to pitch canker disease. Based on data from the disease screens conducted under ADEPT, a pitch canker resistant and susceptible hedge was challenged with F. circinatum. Once disease symptoms were visible on the susceptible hedge (purpling of the stem), fully diseased tissue was discarded and ~1 to 1.5cm of the remaining shoot was collected. A similar harvest was performed for the resistant hedge except in this case the tissue we discarded showed signs of desiccation, not disease. Tissue was sent to UGA for library synthesis and EST sequencing.

ADEPT Progress Report

Santiago Gonzalez Martinez, University of California, Davis
April 2004

Association study for drought

Detection of allelic variation: nucleotide diversity and linkage disequilibrium in drought tolerance candidate genes. We have sequenced a total of 323.712 bp from 18 candidate genes (32 alleles) related to drought tolerance. The frequency of SNPs was approximately 1 SNP per 50 bp of sequence. Within-gene linkage disequilibrium was generally low, decaying quickly after 100-150 bp. No linkage disequilibrium among genes has been found. Two genes (ccoaomt1 and erd3) showed significant values of Tajima’s D and might be under natural selection.

Development of SNP markers. Twenty-eight SNPs (20 of them frequent SNPs, i.e. presence in the discovery panel at frequency higher than 12.5%) are ready to be genotyped in the FBRC association population. This battery of markers cover 12 major candidate genes for drought tolerance (2-3 SNPs per gene). Other 17 SNP primers covering the 6 remaining candidate genes are under testing.

Population genomic study. The objective of this study is to find allelic variation associated with environmental variables in nature. To accomplish our goal nineteen SNPs from 8 drought tolerance candidate genes (most of them expressional candidates) have been genotyped in 306 trees covering the natural range of loblolly pine. Other 40 SNPs from wood trait candidate genes will be used in this study.

ADEPT Progress Report

Sarah Covert, University of Georgia
April 2004

The clone arrays containing potential defense genes from pine were hybridized to cDNA probes derived from healthy xylem, healthy phloem, galled xylem and galled phloem. To account for biological variation, each tissue type was isolated from three separate samples. Then each biological replicate was divided in two before cDNA synthesis, labeling and hybridization. The results from these duplicate probes (or technical replicates) were compared in order to account for variation introduced during probe preparation. A preliminary analysis of the hybridization results has identified several pine genes whose expression is influenced by fusiform rust infection, but not by pitch canker infection. The raw data has been sent to John Davis and Dudley Huber at the University of Florida for statistical analysis. Northern blots to confirm the array results are planned for the near future in our lab.

ADEPT Progress Report

Lee Pratt, University of Georgia
April 2004

A total of 12,927 3′ ESTs have been obtained from the nine RTDR, RTWW, and RTDS libraries. They have all been deposited in GenBank and are available at With separate NSF funding, 5′ ESTs have also been obtained for the same cDNA clones. Java server pages accessible from that location display sequences in text format and provide links to GenBank accessions. A Java GUI additionally provides a graphical display showing phred quality scores for each base call, and identifies vector/adapter, polyT, and high quality trimmed length as submitted to GenBank.

This GUI also provides a viewer to download sequences as fasta files, either individually or by library. A second Java GUI, MAGIC Gene Discovery Viewer, also provides the results of clustering. Clustering was done using a new algorithm developed by Dmitri Kolychev in our group. Many sequences from these IFAFS library are included in the first-generation pine microarray that we are preparing, representing 13,616 genes. Once the pine data have been selected from the “Select a Cluster Run” window, a user can either view the IFAFS data in the context of a much larger pine dataset (default option), or can view it in isolation by selecting Cluster Group 169 from the list box in the “Gene Discovery” window.

A viewer can explore electronic expression profiles, cluster alignments, annotation, and much more, and can also query the Oracle database in a number of ways. In addition, a BLASTable pine EST database is made available from this same web page. In addition, we have obtained 2871 high quality 3′ ESTs for two newer libraries, both from shoot tips. One is resistant to pitch canker (STRR1) and the other sensitive (STRS1). ESTs from these two libraries are not included in the initial Milestone pine clustering now available at our public web site (see above). The sequences, however, can be explored at that site. Our target is a total of 6000 3′ ESTs, divided approximately equally between these two libraries.