ch is involved in JA biosynthesis, a sieve element occluding protein pre venting the loss of photoassimilates after wounding and catalases which are known to serve as common antioxidant enzymes and to induce suberization and other protective mechanisms after wounding . Four proteins with EST counts 100 were peptidyl prolyl cis trans isomerases which are also known as cyclophilins and accelerate the folding of pro teins, proteasome subunits responsible for pro tein degradation and turnover, auxin repressed proteins known to affect auxin signaling as negative reg ulators and methionine synthase, which catalyses the last step in the pro duction of the amino acid L methionine used by plants for many essential direct or indirect cellular processes.

Two further proteins almost unique to the EF li brary in these elms were the enzyme methionine sulfoxide reductase, which functions in plant defense via the regulation of the cell redox status and is known to be involved as an antioxidant in repairing proteins damaged by oxidative stress, and the transport pro tein SFT2, which in yeast is involved in traffic to the Golgi complex and vesicle associated membrane fusion. The R statistic was applied in order to detect differences in relative transcript abundances between the elm treatments. Transcripts with R 3 were considered to be dif ferentially expressed between the Cilengitide libraries. For all these protein types, the R statistic revealed a significant differ ence in transcript abundances between the treatments.

Discussion The large scale EST sequencing results shown here repre sent the first step in studying the defensive responses of field elms to egg laying by the specialist elm leaf beetle Xanthogaleruca luteola, at a molecular level. 361,196 expressed sequence tags were assembled into 52,823 unique transcripts. Although the gene discovery rate among the transcripts was low due to the low number of Ulmus genes in public databases, we were nevertheless able to identify a large number of candidate genes with possible roles in the response of elm to egg lay ing by the elm leaf beetle. Normalization based on se quence sample size and analysis using R statistics provided the basis for comparative gene expression analysis using EST frequencies across five different biological treatments, egg laying and feeding by X. luteola, feeding, transfer of egg clutches, methyl jasmonate spraying and an untreated control.

The function of these candidate genes must now be confirmed in further studies. Despite a similar sample size and the fact that clonal plant material, identical sequencing technologies, and sequence assembly were used, the EST frequencies of the five treatments showed astonishingly small intersec tions as can be seen in the Venn diagrams and visualization of metabolic pathways. Therefore, although the influence of X. luteola feeding on transcripts cannot be ruled out, the ten fold larger library EF F is still capable of being used for detecting the less abundant