Bighead carp (Hypophthalmichthys nobilis) is a key aquaculture species, with the head and its orbital fat being a commercially valuable product. To elucidate the molecular basis of growth variation, we performed comparative transcriptome analysis of orbital fat from extreme growth phenotypes at juvenile (6 months) and market-size (18 months) stages. In juveniles, slow growth was linked to upregulation of stress-responsive genes (sgk1, fkbp5, lipg), while fast growth correlated with higher expression of stress-buffering (crhbp) and nutrient-signaling (rbp2, mgea5) genes. At 18 months, divergent growth aligned with opposing lipid metabolic states: a pro-anabolic profile (dgat2, fads2) supported fast growth, whereas a catabolic profile (cpt1b, ppargc1a) was associated with slow growth. These results demonstrate stage-specific transcriptional reprogramming in orbital fat underlying growth variation. This study provides a molecular framework for orbital fat-mediated growth regulation and highlights potential candidate genes for molecular breeding in bighead carp.

While chemosymbiosis offers deep-sea bathymodioline mussels a critical metabolic advantage, this dependence makes them vulnerable when removed from their native environment, and their early post-recovery dynamics remain poorly resolved. We maintained adult Gigantidas haimaensis collected from the Haima cold seep at atmospheric pressure and sampled gill tissues at 0, 3 and 9 h after collection. Coupled 16S rRNA amplicon sequencing and RNA-Seq were used to follow changes in the gill microbiota, host transcription and descriptive temporal covariation. Methyloprofundus remained the dominant bacterial genus throughout the experiment, whereas Ca. Vesicomyosocius, the SUP05 cluster, and operational taxonomic units (OTUs) assigned to Vibrio and Pseudoalteromonas had higher relative abundances at 9 h in the descriptive analysis and exhibited positive temporal covariation. RNA-Seq showed a stronger transcriptional shift at 3 h than at 9 h. Stress-related genes, including HSP70, HSP105, Toll-like receptors, fibrinogen-related proteins, dual oxidase, thioredoxin-system genes, ferritin, taurine transporter and sulfide:quinone oxidoreductase, changed markedly, while energy-consuming ribosome biogenesis was suppressed to conserve energy under combined physical and nutritional stressors. The temporal patterns of Ca. Vesicomyosocius and other low-abundance bacterial signals coincided with broad host stress and immune and metabolic transcriptional changes; gene-level microbiome–transcriptome associations were exploratory and did not survive false-discovery-rate (FDR) correction. These results document rapid short-term transcriptional and microbiome changes after recovery, while their causal drivers and consequences require controlled validation.