Two de novo transcriptome assemblies and functional annotations from juvenile cuttlefish (Sepia officinalis) under various metal and pCO2 exposure conditions

This study presents two high-quality, annotated *de novo* transcriptome assemblies from juvenile *Sepia officinalis* at different developmental stages and under various metal and pCO2 exposure conditions, providing valuable genomic resources to support environmental and neurobiological research on this model organism.

The Gist

Imagine the cuttlefish (Sepia officinalis) as a high-tech, underwater superhero. It's famous for its ability to change colors, solve puzzles, and even dream. Scientists love studying them to understand how brains work and how animals behave. But here's the problem: to truly understand how these superheroes react to a changing world (like polluted water or acidic oceans), we need their "instruction manual"—their genetic code. Until now, that manual was either missing or very incomplete.

This paper is like a team of librarians finally finishing a massive, two-volume encyclopedia for the cuttlefish. Here is the story of what they did, explained simply:

1. The Mission: Two Different "Snapshots"

The scientists didn't just take one picture of the cuttlefish; they took two very specific snapshots to see how the animal changes as it grows up.

• Snapshot A (The Newborns): They looked at brand-new baby cuttlefish (just hatched). Think of this as looking at a baby's entire body to see what "starter skills" it has.
• Snapshot B (The Toddlers): They looked at one-month-old cuttlefish, but only focused on their heads (brains). This is like zooming in on a toddler's brain to see what complex thoughts and behaviors are starting to form.

2. The Stress Test: "The Ocean Pollution Gym"

To make sure these instruction manuals are useful for real-world problems, the scientists didn't just study happy, healthy cuttlefish in a perfect tank. They put them through a "stress test" in the lab.

They exposed the cuttlefish to different combinations of:

• Acidic water (simulating ocean acidification).
• Heavy metals (like mercury and silver, simulating pollution).

It's like putting a car through a crash test while also running it on bad fuel. By doing this, the scientists captured the cuttlefish's genetic "emergency response" manual. They didn't just record how the animal looks when it's calm; they recorded how its genes scream when it's under pressure.

3. The Result: Two Massive Dictionaries

Using powerful computers, the team translated millions of tiny genetic snippets (RNA) into readable text. They built two massive dictionaries:

• Dictionary 1 (Newborns): Contains about 230,000 entries. It's full of instructions for building the body and setting up the basic nervous system.
• Dictionary 2 (One-month-olds): Contains about 370,000 entries. It's much bigger and more complex.

Why is the second one bigger?
Think of it like a video game.

• The Newborn is at "Level 1." It has the basic controls: walk, swim, eat. Its genetic list is shorter because it's mostly running on "factory settings" (innate behavior).
• The One-month-old is at "Level 10." It has unlocked new features: complex hunting strategies, better memory, and sophisticated social skills. Because the brain is so active and plastic (changeable), it generates many more genetic variations. The scientists found that the older cuttlefish's brain is like a super-computer that constantly rewrites its own code to adapt to new situations.

4. The "Missing" Chapters

When they compared the two dictionaries, they found something fascinating:

• The Overlap: About 40-50% of the instructions are the same in both (the basics of being a cuttlefish).
• The Unique Parts: The rest is unique to the age.
  • The Newborn dictionary is heavy on "construction blueprints" (building nerves, handling stress).
  • The Older dictionary is heavy on "operating manuals" (processing fats, sending complex signals, fine-tuning neurotransmitters).

This proves that you can't just use one genetic map for the whole life of the animal. A baby cuttlefish is biologically very different from a teenager cuttlefish.

5. Why This Matters

Before this paper, scientists trying to study how pollution affects cuttlefish were like mechanics trying to fix a Ferrari with a manual for a bicycle. They had to guess what the genes did.

Now, they have a high-definition, annotated map.

• If a cuttlefish is exposed to mercury, scientists can now look up exactly which genes are turning on or off.
• If the ocean gets more acidic, they can see how the brain's "software" is being rewritten.

The Bottom Line

This paper is a gift to the scientific community. It provides the first complete, high-quality "instruction manuals" for the cuttlefish at two critical stages of life. It tells us that as these animals grow, their genetic complexity explodes, especially in their brains. With these new tools, we can finally understand how these intelligent, colorful creatures will survive (or struggle) in our changing, polluted oceans.

Technical Summary

1. Problem Statement

The common cuttlefish (Sepia officinalis) is a critical model organism for behavioral and neurobiological studies and faces significant threats from anthropogenic global change, specifically combined exposure to heavy metals (e.g., silver, mercury) and ocean acidification (high pCO2). Despite its ecological and economic importance, genomic resources for S. officinalis have been scarce. While a chromosome-level reference genome is imminent, high-quality transcriptomes are essential complements to capture tissue-specific expression, alternative splicing, and RNA editing (a phenomenon highly prevalent in cephalopods). Existing resources were limited to a single tissue-specific assembly, lacking the breadth required to study environmental stress responses across different developmental stages.

2. Methodology

Biological Material and Experimental Design
The study utilized two distinct experimental datasets involving S. officinalis reared under varying environmental stressors:

• Project rnasep1 (Newly Hatched): Collected in 2016. Whole bodies of hatchlings were exposed to six conditions: Control, Acidification (pCO2), Silver (AgNO3), Mercury (HgCl2), and combinations of acidification with metals.
• Project rnasep2 (One-Month-Old): Collected in 2021. Head tissues (central nervous system) of one-month-old juveniles were exposed to four conditions: Control, Acidification, Methylmercury (MeHg) via diet, and combined Acidification + MeHg.

Sequencing and Pre-processing

• RNA Extraction: Performed on whole bodies (rnasep1) and heads (rnasep2).
• Sequencing:
  • rnasep1: Illumina HiSeq 3000 (2x150bp), generating ~358 million raw paired-end reads.
  • rnasep2: Illumina NovaSeq 6000 (2x150bp), generating ~830 million raw paired-end reads.
• Quality Control: Raw reads were filtered using Fastp (removing adapters, low-quality reads PHRED ≤30, keeping reads >120bp) and checked for prokaryotic contamination using Kraken2 (none found).

De Novo Assembly and Thinning

• Assembly: Performed using Trinity (v2.15.1) with in silico normalization.
• Reduction: Assembled isoforms were clustered using CD-HIT (95% identity threshold) to reduce redundancy and retain representative sequences.
• Filtering: Transcripts were mapped to the reference genome (GCA_050097725.1) using GMAP. Chimeras, transcripts with poor alignment (<100bp or <20% length), and sequences <200bp were removed.
• Validation: A pseudo-rarefaction analysis was conducted using Salmon to ensure the subset of samples used for assembly captured the full transcriptomic diversity without excessive computational cost.

Annotation and Analysis

• ORF Prediction: TransDecoder identified Open Reading Frames (≥100 amino acids), retaining the longest ORF per transcript.
• Functional Annotation:
  • BLASTp: Against curated (SwissProt) and full UniProt databases.
  • EggNOG-mapper: For orthologous groups (COG, KEGG, GO, PFAM).
  • GO Enrichment: Performed on specific sequences using GO_MWU.
• Comparative Analysis: Reciprocal Best Hits (RBH) via BLASTp were used to identify shared vs. stage-specific transcripts between the two assemblies.

3. Key Results

Assembly Statistics

• rnasep1 (Newly Hatched, Whole Body):
  • Final assembly: 230,672 transcripts (mean length 677 bp).
  • Putative ORFs: 35,590.
  • Annotation rate: ~70% (26,290 ORFs annotated).
  • BUSCO Completeness: 95% (Metazoa) and 83% (Mollusca).
  • Read mapping: 91.3% of reads mapped back to the assembly.
• rnasep2 (One-Month-Old, Head):
  • Final assembly: 370,613 transcripts (mean length 569 bp).
  • Putative ORFs: 44,233.
  • Annotation rate: ~70% (29,474 ORFs annotated).
  • BUSCO Completeness: 95% (Metazoa) and 83% (Mollusca).
  • Read mapping: 93.4% of reads mapped back to the assembly.

Comparative Transcriptomics

• Overlap: Only 17,526 transcripts (approx. 40–49% of each assembly) were shared between the two stages (RBH).
• Specificity:
  • rnasep1 (Hatchlings): Unique transcripts were enriched in neural network structuring (neural crest formation, dendrite arborization), cellular stress response (unfolded protein response), and metabolism/transport. This reflects the active developmental transition from embryo to hatchling.
  • rnasep2 (Juveniles): Unique transcripts were enriched in lipid processing (sphingolipid metabolism, membrane remodeling), neural signaling (ion channel clustering, catecholamine transport), and post-translational modifications. This aligns with the high lipid content of the cephalopod brain and complex neurotransmission requirements.
• Diversity: The higher number of transcripts in rnasep2 is attributed to the complexity of the mature nervous system, extensive RNA editing, and the inclusion of specific brain tissue versus the whole body of hatchlings.

4. Key Contributions

1. Resource Generation: Provided two high-quality, annotated de novo transcriptome assemblies for S. officinalis, significantly expanding the genomic toolkit for this non-model species.
2. Developmental Insight: Demonstrated distinct transcriptomic landscapes between newly hatched (whole body) and one-month-old (head) stages, highlighting the shift from developmental structuring to mature neural function and lipid metabolism.
3. Environmental Relevance: The assemblies integrate data from individuals exposed to multiple stressors (heavy metals and high pCO2), making them robust resources for studying combined environmental stressors (ecotoxicology) rather than just baseline physiology.
4. Validation: Rigorously validated the assemblies using BUSCO, read mapping, and rarefaction analysis, confirming their suitability for downstream gene expression studies.

5. Significance

This work addresses a critical gap in cephalopod genomics by providing the necessary reference data to investigate how S. officinalis responds to global change. The inclusion of combined stressor conditions (metal + acidification) is particularly valuable for future ecotoxicological studies, as these stressors often act synergistically in the wild. Furthermore, the distinct biological signatures identified between developmental stages underscore the necessity of considering age and tissue specificity in transcriptomic studies of coleoid cephalopods. These resources will facilitate the investigation of transcriptomic endpoints, RNA editing, and isoform diversity, ultimately supporting conservation efforts and a deeper understanding of cephalopod neurobiology under changing ocean conditions.

Data Availability:

• Raw Reads: European Nucleotide Archive (ENA) under PRJEB89715 (rnasep1) and PRJEB88225 (rnasep2).
• Assemblies: ENA accession numbers HCHA01230572 and HCHB01370613.
• Annotations & Code: Zenodo (DOIs provided in the text).