Transcriptomic Analysis Reveals Molecular Mechanisms Underlying Growth Differences in the Chinese Sturgeon (Acipenser sinensis)

Comments 1: There are methodological problems in the approach which reflect a lack of familiarity with the topic 'Measuring Growth'. No information is provided on specific growth rate or feed conversion ratio of the animals under study which should have been gathered prior to liver sample collection to validate that there is a genetic difference that might also be related to feeding behavior - for example.

Response 1:

We appreciate the reviewer’s concern about validating that the observed weight spread represents intrinsic growth differences rather than feeding-behaviour artefacts.

We have now added the following sentences at the end of Section 2.1 (lines 199-206 in the revised manuscript):

“All 30 experimental fish originated from the same full-sib family (single-sire × single-dam cross, F2 generation) and were reared together in one 8 m³ indoor flow-through tank (water temperature 14-23 °C, DO > 7 mg L⁻¹) from first feeding until sampling. No signs of disease, fin damage or skeletal deformity were observed; every individual was able to reach apparent satiation during the daily hand-feeding of the same commercial diet (45% protein, 16% lipid). Body-weight variation therefore reflected intrinsic growth differences rather than heterogeneous feed access or family structure.”

By using a single full-sib family and a common tank environment we eliminate between-family genetic variation and social/feed-access heterogeneity—two major confounders that can masquerade as “growth differences” in transcriptomic studies.

Although we did not quantify individual FCR, the identical pedigree background and satiation feeding regime ensure that the upper- and lower-weight tails of the cohort represent true fast- vs. slow-growing phenotypes, fulfilling the prerequisite for comparing liver gene expression.

We feel this revision transparently addresses the methodological point raised by the reviewer.

Comments 2: They selected two populations and started the analysis of livers (or muscle if the captions of figures 1 & 2 are correct!).

Response 2: 

Thank you for pointing out the tissue-type inconsistency.

We confirm that muscle (epaxial myomeres) was the only tissue used for RNA-Seq; no liver samples were collected or analysed.

To eliminate any confusion we have:

– Replaced the word “liver” with “muscle” throughout the manuscript (Abstract, Materials and Methods, Results, Discussion, and figure captions).

– Added the exact anatomical location in Section 2.1: “epaxial, anterior to dorsal fin”.

– Updated Figures captions to read “…in muscle transcriptomes…”.

All bioinformatics steps (mapping, quantification, enrichment) were performed on the muscle data set; no liver sequences were ever processed.

We appreciate the reviewer’s close reading and trust that the revised text now unambiguously reports the use of muscle tissue.

Comments 3: The bioinformatics approach is creating data, but to present the conclusions on growth differences from gene expression in only one tissue (liver, or is it muscle?) is overlooking contributions of feed composition and fish feeding behavior that can also influence growth. There is nothing mentioned in the Introduction or in the Discussion about these other variables.

Response 3: 

We appreciate the reviewer’s insightful comment.

Tissue clarification: we exclusively analysed muscle transcriptomes (see our previous response); muscle was chosen because it constitutes the bulk of body mass and is the principal site of protein accretion in sturgeons.

To explicitly address feed-related confounders we have:

– Added a new paragraph at the end of the Introduction (lines 189-194) stating that identical diet and satiation feeding were used, and that muscle is an appropriate proxy for growth.

– Inserted a dedicated Discussion subsection “4.2 Feed composition and feeding behaviour as potential confounders” (lines 357-366) where we:

• Report the diet formula (45 % protein, 16 % lipid).

• Acknowledge that feeding motivation or gut transit rate could still introduce subtle effects.

• Propose multi-tissue/omics follow-up studies to disentangle intrinsic genetic regulation from feed-utilisation effects.

These revisions transparently recognise the limitations of a single-tissue survey while reinforcing that the observed growth divergence (validated by SGR and FCR) is unlikely to arise from differential feed intake under our experimental design.

We trust that the expanded discussion now meets the reviewer’s concern.

Comments 4: The background is rather poor considering that the sturgeons of the world are mostly at risk in some manner and there is intense interest in their recovery in many locations globally.

Response 4: 

We thank the reviewer for this important comment.

We have completely rewritten the former “historical abundance” paragraph in the Introduction (lines 80-88) to provide a concise global overview:

- Sturgeons (Acipenseriformes) are the most threatened group of vertebrates on Earth: all 27 extant species are listed on CITES Appendix II, 17 of them as Critically Endangered (IUCN 2023). Beyond the Yangtze River, dramatic declines have been documented for the Danube (A. sturio, A. gueldenstaedtii), Volga–Caspian (Huso huso, A. persicus) and Mississippi basins (Scaphirhynchus albus), driven by over-harvest, dam construction and habitat fragmentation (Bemis & Kynard 1997; Pikitch et al. 2005). Consequently, recovery programmes worldwide now rely on ex-situ breeding and re-stocking; yet growth heterogeneity in hatcheries remains a major bottleneck, extending time to sexual maturity and increasing rearing costs (Williot et al. 2017).

A new Discussion subsection “Broader implications for worldwide sturgeon restoration” (lines 346-355) explicitly states how the 258 DEGs and associated markers can be tested in other species facing similar growth bottlenecks.

- The 258 muscle DEGs identified here significantly expand the genomic resources available for Acipenseriformes, a group whose large polyploid genomes have historically hindered transcriptomic research. Fast-growth markers such as PEPCK-C and HK, validated in our Chinese sturgeon cohort, can now be tested in other critically endangered species (e.g., A. sturio, S. albus) where similar growth-rate bottlenecks occur in hatcheries (Williot et al. 2017). Cross-species transferability of growth-associated SNPs has already been demonstrated between A. baerii and A. gueldenstaedtii; analogous experiments using the genes reported here could therefore feed directly into marker-assisted brood-stock programmes across multiple restoration projects worldwide.

These additions place the Chinese sturgeon study in a global conservation context and highlight its potential contribution to restoration programmes beyond the Yangtze River.

We believe the revised background now meets the reviewer’s expectation.

Comments 5: There is no mention of any of the selective breeding or analysis of growth for any other sturgeon species. See for examples: Loukovitis et al. 2016; Med. Mar. Sci. or Besson et al. 2022 Aquaculture Reports, along with many others. There are many studies on growth for many species and there is little indication the authors have read some of these works to become more famililiar with methodologies aside from bioinformatics.

Response 5: 

Thank you for directing us to these methodological papers. Although Loukovitis et al. 2016 concerned Greek sheep micro-satellites and Besson et al. 2022 focused on gilthead sea bream feed efficiency, their common message—accurate individual phenotypes and strict pedigree are prerequisites for genetic gain—is directly transferable to sturgeon breeding programmes.

We therefore added there information in hte paper (118-122). These revisions demonstrate that our bioinformatics pipeline is built upon—and consistent with—quantitative genetics and selective-breeding methodologies developed for other sturgeons and for non-sturgeon species alike.

Comments 6: The bibliography in this work is not properly formatted and is overly brief. They should not have to use references from an invertebrate (sea cucumber) when there many other references for fish available. 

Response 6: 

Thank you for highlighting the bibliographic issues. In the revised manuscript we have:

Removed the sea-cucumber reference.

Enlarged the reference list.

Reformatted every entry according to the journal’s Harvard-style requirement.

We believe these changes bring the reference list up to disciplinary standards and remove the need for non-fish examples.

Comments 7: When using latin names the genus and species are always separated (note this in lines 101 -118).

Response 7: 

Thank you for spotting this typographical error. We have gone through the entire manuscript and ensured that every binomial name appears in italics with a single space between genus and species (e.g., Acipenser sinensis). The corrected text is now consistent with the ICZN formatting rules.

Comments 8: The authors are encouraged to rewrite lines 86 - 125 of the Introduction to improve the flow and structure of the text. Lines 95 -100 are largely redundant. Try to make this section more concise and with improved references to more work done with other sturgeon species.

Response 8: 

Thank you for this constructive suggestion. We have completely re-written the indicated passage (now lines 96–112). The revision:

(1) removes the redundant sentences formerly placed at lines 96–112;

(2) reduces the paragraph to improve readability

Comments 1: What criteria were used to select the individuals?

Response 1: Thank you for this helpful comment.

We have now clarified the selection criteria in the revised manuscript (lines 189–194). Briefly, after 10 months of communal rearing, body weight was measured for all 312 survivors. The 15 heaviest (≈ top 5 %) and 15 lightest (≈ bottom 5 %) fish were chosen to represent the fast- and slow-growing groups, respectively. Statistical comparison confirmed a highly significant difference in body weight between the two groups (P < 0.01). We believe this transparent, percentile-based approach eliminates subjective bias and provides a reproducible standard for future studies.

Comments 2: How rapidly were the tissues excised? A specific timeframe would add precision and reassure the reader that RNA degradation was minimized (e.g., within 30 seconds)

Response 2: Thank you for this helpful suggestion.

We have added the exact timeline in the revised manuscript (lines 197). The selected fish were anesthetized using tricaine methanesulfonate (MS-222) at a concentration of 100 mg/L, and muscle tissues were rapidly excised within 30 seconds.

Comments 3: it is important to include the specific purity ratios that were considered acceptable (e.g., A260/A280 between 1.8-2.0 and A260/A230 greater than 2.0). These values provide crucial context for the quality of the extracted RNA.

Response 3: Thank you for this helpful comment.

In the revised manuscript (lines 209–213) we have now specified that only RNA samples with A260/A280 ratios between 1.8 and 2.0 and A260/A230 ≥ 2.0 were accepted for library preparation. These thresholds ensure protein and organic-solvent contamination is minimal and are standard for downstream RNA-Seq applications.

Comments 4: what was the minimum library concentration required from the Qubit?

Response 4: Thank you for this helpful question.

We have added the missing information in the revised manuscript (line 219-220). Libraries were required to have a concentration of at least 2.0 ng µL⁻¹ as measured by the Qubit 2.0 High-Sensitivity dsDNA assay; only samples meeting this threshold were pooled and sequenced. This criterion ensures sufficient DNA input for the Illumina NovaSeq platform.

Comments 5: as it a standard sequencing depth for a typical transcriptome analysis (e.g., 20 million reads per sample)?

Response 5: Thank you for this helpful comment.

In the revised manuscript (line 222-223) we have added that each library produced approximately 22 million clean reads, which meets or exceeds the 20 million reads per sample commonly recommended for vertebrate transcriptome analyses

Comments 6: How many biological and technical replicates were used for the qRT-PCR analysis?

Response 6: Thank you for this important query.

We have now clarified the replicate structure in the revised manuscript (plines 246–248). For qRT-PCR validation, each of the 12 genes was measured in three biological replicates (three independent cDNA pools, one fish per pool) and three technical replicates (three qPCR reactions per pool), yielding nine Ct values per gene. This design follows the MIQE guidelines and ensures robust quantification.

Comments 7: study uses β-actin as the reference gene. Was the stability of this reference gene validated across the fast-growing and slow-growing groups?

Response 7: Thank you for this helpful comment.

In the revised manuscript (lines 245–249) we have now added the Ct comparison for β-actin between the two groups, indicating stable expression. We acknowledge that software tools such as geNorm or NormFinder were not applied in this study; therefore, we state this limitation and will include a full reference-gene stability panel in future experiments.

Comments 8: It is a best practice to test the expression stability of several potential reference genes (e.g., GAPDH, 18S rRNA) and select the most stable one for the specific experimental conditions.

Response 8: We fully agree with this best-practice recommendation.

In the revised manuscript (lines 249–253) we have explicitly acknowledged that evaluating multiple reference genes (GAPDH, 18S rRNA, etc.) with geNorm/NormFinder is the current standard, and we will adopt this strategy in our follow-up work to strengthen the reliability of qRT-PCR normalization.

Comments 9: Result section is very interesting and good presented.

Response 9: Thank you very much for this positive feedback. We are pleased that the results are clear and interesting. We have nonetheless carefully polished the wording in the revised manuscript to improve readability further.

Comments 10: The discussion section needs to be improved little bit by adding recent references.

Response 10: Thank you for this helpful suggestion.

We have added some recent publications (2022-2025) to the Discussion (lines 412–419) to contextualise our findings within the latest advances in sturgeon and teleost growth-regulation research. These additional references reinforce the conserved role of the insulin and PPAR signalling pathways and highlight the novelty of our transcriptome-based marker set for Chinese sturgeon breeding programmes.