1Institute of Biomedical and Life Sciences, Glasgow University, Glasgow G12 8QQ, UK; 2John Innes Centre, Norwich NR1 4UH, UK

Background and objectives
Infection by compatible viruses usually leads to the expression of symptoms that may involve a programmed response by the host [1, 2]. Little is known about which plant genes are important in host response and symptom expression during infection by compatible viruses. Lipid transfer proteins (LTPS) [3] are small basic proteins, encoded by a multigene family, that can bind fatty acids and acyl-CoA esters and transfer lipids between membranes in vitro. Their function is not clearly understood, but they have been implicated as anti-fungal and anti-bacterial defence proteins [3]. We have investigated the expression of LTPs in Arabidopsis during infection by mild and severe isolates of cauliflower mosaic virus (CaMV), and in plants in which transgenic expression of a CAMV protein (P6) gives rise to a symptom like phenotype [2].

Materials and methods
Plants were infected with CAMV isolates Cabb B-JI or Bari-1 by mechanical inoculation [1]. In transgenic Arabidopsis lines expressing CAMV gene VI protein (P6), expression was driven by a 35S promoter [2]. In transgenic LTP-reporter lines, the promoter from a Brassica napus LTP gene (Bn-LTP) was used to drive expression of GUS. GUS activity was assayed by staining seedlings with X-gluc and by MUG assays. LTP MRNA levels were determined in Northern blots, using a cloned BN-LTP CDNA as probe.

Results and conclusions
Bn-LTP::GUS transgenic plants were inoculated with a severe isolate of CAMV (Cabb B-JI). Plants were assayed for GUS activity at 14 d.p.i., by which time infected plants had begun to develop vein clearing symptoms. Leaves from mock-inoculated and uninfected controls stained only very lightly. Inoculated leaves from infected plants stained strongly, with four or five intense blue spots, presumably representing local lesions; uninoculated leaves also stained strongly. Most of the root tissue was barely stained. However, root hairs in infected, although not control plants, stained intensely. We observed almost identical results in plants infected with the a mild isolate CAMV Bari-1, which is asymptomatic in Arabidopsis. At 7 d.p.i., before systemic spread had occurred, GUS activity in inoculated and control plants was similar.

In leaves harvested 14-21 d.p.i., levels of endogenous LTP mRNAs were 2-4 fold greater in infected plants compared to controls. Levels of LTP mRNAs were up to 3-fold higher in uninfected transgenic plants expressing P6 than in non-transgenic controls. Levels were highest in transgenic lines with the highest levels of P6. We are investigating P6 induction of LTP genes by crossing the Bn-LTP::GUS reporter construct into the P6 transgenic lines.

Up-regulation of LTP genes forms part of the response to systemic infection by CAMV. The extent of this response appears to be independent of symptom severity. The promoter for at least one LTP gene contains element(s) that respond to virus spread. The product of CAMV gene VI, the major viral determinant for symptom character, appears to be important in controlling LTP gene expression. We plan to use LTP genes as markers with which to analyze plant responses to compatible viruses.

1. Cecchini E et al., 1998. J. Exp Botany 49 (in press). 2. Cecchini E et al., 1997. Mol. Plant-Microbe Interactions 10, 1094-1101.
3. Kader J-C, 1996. Ann. Rev. Plant Physiol. Plant Mol. Biol. 47, 627-654.