Expression Profiling to Identify Novel Desiccation Response Transcripts from Tortula ruralis gametophytes

In this study we have tried to utilize the unique aspects of the T. ruralis response to desiccation and rehydration to design a strategy to identify rehydrins that are of low abundance and perhaps completely novel to the desiccated or rehydration transcriptomes. We have constructed two Subtractive Suppression Hybridization (SSH) libraries (Diatchenko et al., 1996) that are designed to enrich for differentially expressed low-abundance transcripts contained within gametophytic cells either in the slow-dried state (mRNP sequestrated rehydrin transcripts) or cells that have been rapidly dried, rehydrated and sampled at 2h of hydration (rehydrin and recovery transcripts) when the translational change in gene expression is at its peak (Oliver 1991). To achieve this aim we constructed SSH libraries using PolyA RNA isolated from the polysomal (mRNP) fractions from the slow-dried and 2h rehydrated rapid dried gametophytes selected against PolyA RNA from hydrated control gametophytes as the source for driver cDNA. Collections of cDNA clones from each library were sequenced and used to generate a small T. ruralis SSH cDNA microarray for expression profiling of both total RNA extracts for transcript accumulation assessments and polysomal RNA extracts for transcript sequestration and recruitment assessments. Overall design: To assess the expression characteristics of the transcripts represented by the SSH contigs we established a cDNA microarray containing the inserts (PCR derived fragment) from each of the 768 individual SSH ESTs, with the exception of thirteen that failed to generate a PCR fragment. Twelve of the missing thirteen SSH EST cDNAs were replaced with PCR fragments from previously isolated T. ruralis cDNAs four of which, representing the ribosomal proteins S14, S16, L23 and L15 (Wood et al., 2000, Zeng and Wood 2000), were previously reported to be constitutively expressed and were added to serve as normalization genes. The remaining eight clones, Tr155, Tr217, Tr403, Tr416, Tr421 (described by Scott and Oliver, 1994), and TrCDPK (U82087) were added as either positive “up-regulated” (Tr155, Tr403, Tr421), negative “down-regulated (Tr217, Tr416), or neutral (TrCDPK) controls based on previous northern analyses. The cDNAs were printed from two 384 well plates in 12 blocks (two columns of 6) of 24 x 8 spots such that each SSH EST and controls were represented in triplicate. Each of the triplicate cDNAs was separated within the blocks to eliminate possible spatial hybridization bias. All hybridizations were duplicated as dye swaps with two separate RNA preparations, from large populations of individual gametophytes (isolated from a minimum of three separate clumps), serving as the source for the sscDNA Cy3 and Cy5 labeled probes. The RNA preparations for the Total polyA RNA were by necessity separate samples from those used to isolate Polysomal poly A RNA.