Beatrice Romagnolo, Min Jiang, Moni Kiraly, Carrie Breton, Rebecca Begley, John Wang, Jim Lund, and Stuart K. Kim
C. elegans genetics
The HS::let-60(G12V) vector was kindly provided by M. Han. Complementary DNA corresponding to the let-60(G12V) mutant was inserted into pPD49.78 (from A. Fire) which carries the heat-shock promoter hsp16-2. Transgenic animals were generated by microinjection of HS::let-60(G12V) DNA (50 ng/ul) and unc-119(+) DNA (100 ng/ul) into unc-119 mutants following the method of Mello et al. . An integrated line (gaIn143) was generated by gamma-irradiation, and then back-crossed twice to wild-type.
Worms were grown and RNA was prepared as in Reinke et al. . Synchronous populations of let-60(G12V) worms were harvested 50 hours after feeding starved worms synchronized in the mid-L1 stage. Developmental stage was verified by observing a small sample of animals with Nomarski optics to score the size of the gonad and the development of the vulva. For heat-shock treatment, the plates were transferred to a water bath at 33oC for 35 minutes, followed by recovery at 15oC for 30 minutes, 1 hour or 2 hours. We isolated eight RNA samples for the 0 hour time point and four samples for time points at 0.5, 1 and 2 hours.
Synchronous populations of wild type (N2), let-60(n1046) or let-23(sy1) worms were harvested from 32 hours to 44 hours after hatching, using protocol outlined in Jiang et al. . Two independent samples were isolated except for let-60(n1046) at 44 hours and let-23(sy1) at 32 or 40 hours, which have one sample each. The reference used in the let-60(G12V) experiments was from a different worm preparation than that used for the let-60(n1046), let-23(sy1) experiments.
DNA microarrays used for the let-60(G12V) experiments are described in Jiang et al. , and for the let-60(n1046) and let-23(sy1) experiments are described in Reinke et al. . RNA preparation, cDNA synthesis, microarray hybridization and microarray scanning were performed as previously described . Cy-3 dUTP was used to label experimental cDNA and Cy5-dUTP was used to label cDNA from reference RNA made from mixed stage population of wild type worms.
In order to identify genes that are expressed differently following heat-shock treatment of let-60(G12V) and wild-type animals, we employed a mixed-model ANOVA analysis procedure as implemented by SAS. The model has a genotype factor (let-60 or wild-type), a time factor (0, 0.5, 1, 2 hours), an interaction factor between genotype and time, and a random factor from the repetitions. More explicitly, our statistical model is:
G, T, and GT are the effects due to genotype (G), time (T)
and the interaction of genotype and time (GT);
For implementation of mixed procedure analysis in SAS, fixed effects for genotype (G), time (T), and genotype*time are placed in the MODEL statement, and random effects for repetitions placed in the RANDOM statement. The interaction effect, GT, indicates that the expression profile of one gene along the time course in the let-60(G12V) strain is different than that due to a pulse of heat-shock to wild type over the time course. We analyzed genes that revealed significant values of GT (p<0.001).
487 out of 708 target genes identified in the let-60(G12V) time course were present on the microarrays used in the let-60(n1046) and let-23 time courses, and we examined the expression for each of these 487 genes in these time courses. To calculate the difference in expression between the two time courses, we calculated the log2(let-60/let-23expression ratio) for each replication of each gene at each time point. We then determined the average log2(let-60/let-23 expression ratio) for each gene at each time point, summed up the average expression change over the seven time points and then selected 41 genes in which the cumulative expression change was more than 5. Out of these 41, 34 display the same type of regulation (induction or repression) in the let-60(G12V) and let-60(n1046) time courses.
Western blot analysis of staged L3 populations were performed as described in Lackner et al., 1998. Polyclonal anti-Erk2 peptide antibody K-23 (SantaCruz Biotechnology, Santa Cruz, CA) and monoclonal anti-Erk-1&2 M8159 (Sigma) were used for immunostaining of MPK-1 and phospho-MPK-1.