Comparison of larval growth rates between genotypes produced in a reciprocal cross of families of Crassostrea gigas. (A) Growth rates calculated from regressions of mean size (shell length, μm) against age. Values shown represent means (N=50) ± s.e.m. Growth rates calculated from regressions (μm day−1) are: family 3×3=7.2±0.6; 3×5=9.6±0.5; 5×3=9.6±0.5; and 5×5=5.6±0.6. All regressions are significant (ANOVA: P<0.05). (B) Shell lengths (means ± s.e.m., N=50) at 6 days post-fertilization compared among duplicate cultures of the four larval genotypes. Bars sharing a lowercase letter represent cultures for which no significant differences in average size were observed (Tukey's HSD: P>0.05). The average sizes of larvae from all hybrid crosses (3×5, 5×3) were significantly larger than those from the parental lines 3×3 and 5×5 (Tukey's HSD: P<0.01).
Examples of the four major non-additive gene expression patterns in fast-growing larvae of C. gigas. Each of these patterns was selected for further analysis. Expression levels are given as transcripts per million, based on MPSS counts from GEO accession number GSE3596. For each comparison between reciprocal hybrid families (3×5 and 5×3) and larvae from their inbred parental lines (3×3 and 5×5), a horizontal dashed line indicates the additive expectation (the mid-parental value). Expression values sharing a lowercase letter code were not significantly different. The four examples shown correspond to the following clone numbers: OD (overdominant), clone 72; D+ (dominant-high expression), clone 34; UD (underdominant), clone 122; and D− (dominant-low expression), clone 156.
Annotation of candidate genes from fast-growing larvae of C. gigas. (A) Summary of GenBank database comparisons for the full set of 181 cDNA clones analyzed in this study. Clones were scored as showing similarity with known coding sequences (N=37 clones), similarity with uncharacterized or non-coding sequences (N=41), or no significant similarity (N=103). Assignments based on TBLASTX searches of GenBank and EST databases with an e-value threshold of 10−3. (B) List of 34 different genes identified based on similarity to annotated coding sequences. Assignments based on gene product annotation of the subject showing highest sequence similarity, as determined from TBLASTX analysis. (C) Biological processes associated with the 24 candidate genes showing similarity to genes annotated with Gene Ontology biological process terms. Assignments based on BLASTX search of GenBank using the Blast2GO application, with an e-value threshold of 10−3.
Transcript abundance of ribosomal protein RPL35 during development of C. gigas measured by RNA (northern) blots. Expression compared for four larval families with different genotypes based on hybridization of a gene-specific radiolabeled probe for ribosomal protein RPL35 to blots containing equivalent amounts of RNA from a series of developmental stages from eggs to 6-day-old veliger larvae. (A) Digital image of RNA blot showing relative abundance of RPL35 transcript at different developmental times (age in days shown above each blot) for all four genotypes (3×3, 3×5, 5×3, 5×5). Probe bound to a single transcript of ~800 bp. (B) Relative abundance of RNA for ribosomal protein RPL35 calculated by standardizing the amount of bound radioactive probe (shown in A) to the amount of ethidium bromide-stained 18S rRNA. Relative units of band density calculated from digital image analysis.
Examples of candidate genes showing linear relationships with growth rates in larvae of C. gigas. Growth rate data are re-plotted from Hedgecock et al. (Hedgecock et al., 2007). For each example, expression data were normalized to the lowest expressing family for that gene, based on the primary count data from GEO accession number GSE3596. (A) Positive relationships — clone numbers 58 (ribosomal protein L32), 83 (ribosomal protein S23) and 101; (B) negative relationships — clone numbers 120, 128 and 137.