Life stage-dependent RNA interference in the tropical rock lobster Panulirus ornatus: mechanistic insights from RNA-seq and functional studies
Thomas M. Banks, Susan Glendinning, Avani Bhojwani, Quinn P. Fitzgibbon, Gregory G. Smith, Tomer Ventura
TL;DR
This study explores why RNA interference works better in juvenile tropical rock lobsters than in larvae, offering insights for improving RNAi-based tools in aquaculture.
Contribution
The study reveals life stage-dependent RNAi efficacy in Panulirus ornatus and identifies key RNAi pathway components and mechanisms.
Findings
Juveniles show efficient systemic RNAi response, while larvae have limited silencing capacity.
SID1 facilitates dsRNA uptake in juveniles but is less active in larvae.
RNA-seq analysis shows upregulation of RNAi machinery and antiviral response genes in juveniles after dsRNA exposure.
Abstract
RNA interference (RNAi) has emerged as a key molecular tool in various commercially important decapod crustaceans, offering potential biotechnological applications in aquaculture. However, in the tropical rock lobster Panulirus ornatus, gene silencing through RNAi has proven difficult to achieve despite the availability of extensive omics data. This study investigates the RNAi response across life stages in P. ornatus, focusing on larvae and juveniles to determine when the species is most receptive to RNAi. Late-stage phyllosoma larvae and early juveniles were microinjected with dsRNA for the insulin-like growth factor binding protein encoding gene to determine silencing efficiency. Our results show that while juveniles exhibit an efficient systemic RNAi response with robust silencing across tissues, larvae display limited silencing capacity. A key finding is the differential expression…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsInvertebrate Immune Response Mechanisms · Crustacean biology and ecology · Neurobiology and Insect Physiology Research
Introduction
The tropical rock lobster, Panulirus ornatus, has emerged as a promising aquaculture candidate. Over two decades of research have ultimately led to closed-lifecycle production and greatly enhanced culture practices (Uy et al. 2023). This unprecedented access to P. ornatus individuals at all life stages has allowed for the development of genetic resources and molecular techniques to promote aquaculture biotechnology in this species (Hyde et al. 2019; Lewis et al. 2022; Ventura et al. 2020). RNA interference (RNAi) is a commonly used molecular technique for sequence specific post-transcriptional knockdown, which can complement the increasing genetic resources and knowledgebase of P. ornatus to functionally characterize biotechnologically relevant genes and pathways (Sagi et al. 2013). The increased omics resources have specifically aided in the identification and characterization of developmental and biochemical pathways, which can provide novel targets for RNAi-based manipulation both to study gene function and develop desirable phenotypes for commercial P. ornatus aquaculture (Glendinning et al. 2023; Kelly et al. 2023; Lewis et al. 2024; Nguyen et al. 2023).
RNAi is a highly conserved biological mechanism of gene silencing in eukaryotic organisms, which both regulates development and combats threats to the genome (for example from viruses and transposons) (Claycomb 2014; O’Brien et al. 2018; Rojas-Rios and Simonelig 2018; Schuster et al. 2019). Generally, RNAi refers to the microRNA (miRNA), piwi interacting RNA (piRNA), and small interfering RNA (siRNA) pathways, each playing specific and distinct roles (Ketting 2011). The siRNA pathway is most relevant in the context of aquaculture biotechnology, as it allows for exogenous double stranded RNA (dsRNA) induced silencing, which leads to the knockdown of specific host genes, as well as inducing virus specific silencing (Nguyen et al. 2018). Briefly, the mechanism of RNAi in invertebrates begins with long dsRNA (> 50 bp usually, derived from exogenous sources (Lozano et al. 2012), which acts as a substrate for the endoribonuclease Dicer 2, which cleaves the dsRNA into double stranded siRNAs (Bernstein et al. 2001; van Cleef et al. 2014). Following this, siRNAs are bound by Argonaute 2, which unwinds the RNA duplex, degrades the passenger strand, and complexes with other cytoplasmic dsRNA binding proteins to form the RNA induced silencing complex (RISC) (Tabara et al. 1999; Zhang et al. 2018). The RISC subsequently surveys the cell for complementary mRNAs, and if there is perfect or near perfect homology to the siRNA, Argonaute 2 degrades the mRNA via its PIWI domain (Nakanishi 2016; Wu et al. 2020). This mechanism effectively silences the genes at the post-transcriptional level, and is harnessed naturally in response to viral infection or artificially to knockdown genes of interest (Schuster et al. 2019).
Upstream of the intracellular silencing cascade are the mechanisms that facilitate systemic RNAi, which allow for efficient uptake and dispersal of dsRNA and the silencing signal at a whole organism level (Melnyk et al. 2011). A systemic RNAi response enables an effective RNAi response in distal or difficult-to-access tissues, and is mediated by endogenous mechanisms which naturally recognize and uptake dsRNA as part of the antiviral immune response (Joga et al. 2016). Genes of interest therefore can be investigated in any tissue with relatively straightforward delivery methods, such as direct injection into the hemolymph (Banks et al. 2020). Effective uptake of dsRNA greatly contributes to the systemic RNAi response, where systemic RNAi-defective 1 (SID1) proteins are known to play an important role in this process in crustaceans (Maruekawong et al. 2018; Winston et al. 2002). SID1 is a conserved transmembrane channel found across metazoans which is selective for dsRNA, facilitating passive uptake from the extracellular space (Shpak et al. 2017). The intracellular retention of dsRNA is then mediated by RNA binding proteins and the RISC, which prevent export back into the hemolymph (Shih and Hunter 2011; Shih et al. 2009). In crustaceans, SID1 has been implicated in the uptake of dsRNA in the gill, muscle, hemocytes, androgenic gland, and likely many other tissues not yet tested (Maruekawong et al. 2018; Shpak et al. 2017). In the hepatopancreas, however, dsRNA uptake is governed by clathrin-mediated endocytosis via an unknown receptor, which works in tandem with SID1 to enable systemic uptake of dsRNA in Litopenaeus vannamei (Maruekawong et al. 2019, 2022). Environmental dsRNA uptake may therefore occur via the oral route in L. vannamei, facilitated by receptor-mediated endocytosis, followed by export to the hemolymph/circulatory system where SID1 is then largely responsible for dsRNA uptake (Banks et al. 2022; Maruekawong et al. 2022), analogous to the SID1 and SID2 mechanisms in Caenorhabditis elegans (Feinberg and Hunter 2003; McEwan et al. 2012; Winston et al. 2007).
Numerous mechanisms have been linked to reduced effectiveness of the systemic RNAi response, and specifically in insects this has been reviewed thoroughly as many of these species represent commercially relevant agricultural pests (Cooper et al. 2019; Joga et al. 2016). Overall, since RNAi has co-evolved with viruses, many mechanisms which likely contribute to antiviral defense may likewise hinder exogenous dsRNA induced gene silencing, by targeting dsRNA molecules specifically (Cooper et al. 2019; Mongelli and Saleh 2016). This is due to the fact that dsRNA, which is produced by all viruses independent of genome structures (Chen and Hur 2022), acts as a pattern associated molecular pattern (PAMP) for the innate antiviral immune system (Kingsolver et al. 2013). As a result, many mechanisms, such as the well-established nuclease repertoire of lepidopteran insects, have emerged to prevent systemic transport and uptake of dsRNA to combat viruses, which can in turn reduce RNAi efficiency as the same substrate is utilized in both processes (Garbutt et al. 2013; Guan et al. 2018; Singh et al. 2017). The application of dsRNA, independent of silencing, can also lead to overexpression of the RNAi machinery, which essentially primes the system and allows for increased silencing efficiency (Guo et al. 2015; Maruekawong et al. 2019, 2018; Yang et al. 2020; Ye et al. 2018). As the expression of the core RNAi machinery is suspected to be a major determinant of silencing efficiency (Banks et al. 2022; Cooper et al. 2019; Guo et al. 2015), repeated exposure to dsRNA (and increased priming of the system) may overcome the effects of RNAi-inhibiting mechanisms. Mechanisms inhibitory to the RNAi pathway have not been well studied in decapod crustaceans, since silencing is usually potent in both model and commercially relevant decapod species (Sagi et al. 2013).
RNAi has been employed in decapod crustaceans primarily to study gene function via gene knockdown studies; however, a number of applications in aquaculture biotechnology have additionally been proposed, particularly for use as an antiviral tool against severely detrimental pathogens such as white spot syndrome virus (Black et al. 2024; Hong and Kim 2024; Nilsen et al. 2017; Sagi et al. 2013; Thammasorn et al. 2015). RNAi is widely accessible among a number of decapod species where high silencing efficiency and a capacity for systemic RNAi is observed (Banks et al. 2020; Black et al. 2024; Ge et al. 2020; Li et al. 2018; Ponprateep et al. 2012; Smith et al. 2023; Ventura et al. 2012), however, to date, there are very limited accounts of efficient gene silencing in any palinurid lobster (Banks et al. 2020, 2022). In P. ornatus and the related eastern spiny lobster, Sagmariasus verreauxi, repeated attempts at gene silencing via RNAi across multiple life stages revealed very little silencing capacity (T. Ventura; personal communication). Further research into the limiting RNAi mechanisms in P. ornatus has revealed some deficiencies in the core pathway machinery, which may contribute to reduced silencing capacity; however, these mechanisms remain to be further characterized (Banks et al. 2020, 2022).
Here, we aim to further elucidate the RNAi mechanism in P. ornatus and examine the capacity for gene silencing in both larval and juvenile lobsters. In this study, the core RNAi machinery and systemic RNAi response have been investigated in larval and juvenile stages of P. ornatus to determine which life stage may be most amenable to silencing. Critical differences in the expression of the RNAi machinery were observed between life stages, which correlated with silencing efficiency. Furthermore, transcriptomic analysis was performed to identify mechanisms relating to RNAi and dsRNA exposure. This has enhanced the knowledge of the decapod crustacean RNAi pathway and led to the identification of innate immune mechanisms which may contribute to antiviral defense in spiny lobsters. Understanding and characterizing these mechanisms will allow for the development of better RNAi methodology and delivery methods in P. ornatus, which in turn will expand the molecular and genetic tools for this commercially relevant aquaculture species.
Materials and methods
Experimental animals
Larvae of P. ornatus were reared from egg at the University of Tasmania’s Institute for Marine and Antarctic Studies and cultured using commercial in confidence protocols based on previously described methods (Fitzgibbon and Battaglene 2012; Smith et al. 2009). Juveniles of P. ornatus were obtained following metamorphosis of the larvae and were maintained according to previously developed protocols at the Institute for Marine and Antarctic Studies (Kelly et al. 2023).
Evaluating stage-specific silencing capacity
To assess stage-specific silencing capacity of P. ornatus, late stage phyllosoma larvae (100–150 mg) and recently metamorphosed juveniles (80–120 mg) were microinjected with either 5 µL of 5 µg/µL dsRNA (n = 6 per stage) for the Insulin-like Growth Factor Binding Protein-encoding gene (dsIGFBP; see Supplementary Fig. S1) or 5 µL MilliQ water as a control (n = 6 per stage). Animals were kept overnight in standard culture conditions as described above and then snap frozen 24 h post injection for downstream RNA extraction and silencing validation via qPCR. RNA was extracted using RNAzol^®^RT reagent (Sigma-Aldrich, #R4533, Australia) supplemented with 1% ß-mercaptoethanol, as previously described, with quality and yield assessed via nanodrop spectrophotometer (Hyde et al. 2019). 1 µg of RNA was then reverse transcribed using the Tetro cDNA synthesis kit (Bioline, # BIO-65043, Australia) using an oligo dt primer to mitigate genomic DNA contamination, prior to quantification of relative IGFBP expression via qPCR with the 18S gene used to normalize expression. Primers were designed using an online service to reproduce the Assay Design Center from Roche (https://primers.neoformit.com/) with the relevant probe used for each assay (see Supplementary Table S1, for primers and probes), and reactions were performed in Rotor Gene 6000 thermocycler. Relative gene expression of IGFBP compared between the dsRNA treated and control individuals was calculated using the 2^−ΔΔCT^ method, followed by a Student’s T test to determine significance. Data are presented as mean ± standard deviation.
Examining expression of RNAi pathway genes in gene-silenced larvae and juveniles
Exposure to dsRNA is known to modulate the expression of the RNAi pathway in arthropods, and the degree of upregulation can provide insights into the silencing mechanisms and which components may be interacting with dsRNA (Shpak et al. 2017; Ye et al. 2021). The expression of the core siRNA machinery Dicer 2, Argonaute 2, and SID1 was determined with qPCR as above in silenced (n = 6 per stage) vs. non-silenced individuals (n = 6 per stage) in both life stages, using previously identified sequences and primers (Banks et al. 2022). A Shapiro–Wilk test was performed on relative gene expression data to determine normality, followed by either a Student’s T test (normally distributed data) or a Mann–Whitney U test (non-normally distributed data) to determine statistically significant differences in expression between the silenced and control individuals. Data are presented as mean ± standard deviation.
Fluorescent dsRNA production
Fluorescent dsRNA (fdsRNA) was produced as previously described with slight modification (Banks et al. 2020). The sequence encoding gonad inhibiting hormone of P. monodon was used as a template for dsRNA production, to ensure no cross-reactivity with endogenous P. ornatus mRNA. Briefly, RNA was extracted from the P. monodon central nervous system and reverse transcribed into cDNA, which was used as a template for PCR to produce a 660 bp amplicon. An on top PCR was then performed with T7 promoters to generate two templates for each strand of RNA to be produced. For the single stranded RNA synthesis (ssRNA), 0.4 µL of UTP-Atto-488 (Jena Bioscience) was added per 20 µL ssRNA reaction. Following overnight incubation at 37 °C, ssRNA was cleaned with acid phenol chloroform, chloroform, and ethanol precipitation. The forward and reverse strands were then annealed at 70 °C for 10 min and incubated for 1–2 days at room temperature without light exposure. dsRNA integrity was then assessed via gel electrophoresis on 1.2% agarose gels.
Phyllosoma microinjection and fluorescent microscopy
Phyllosoma were microinjected once with 2 µL of Goat anti-rabbit IgG–Alexa Fluor Plus 594 (red antibody; ThermoFisher Scientific) or 2 µg/µL of fdsRNA, using pulled and sharpened glass needles with the IM-9B microinjector (NARISHIGE group, Japan) in the 5th walking leg sinus. Animals were transferred to a clean petri dish with seawater for photography using the Leica MZ16FA fluorescence stereomicroscope with a 100-W mercury vapor lamp, Fluo-III fluorescence module, and GFP2 (excitation 460–500 nm) and Texas Red (excitation 540–580 nm) filters, on a transmitted light base. Photos were taken using the Leica DMC4500 CCD sensor with a 0.6 × camera mount (to capture the whole animal) using the Leica LAS V4.1 software at 10 min and 3 h post injection. For most treatments, a 25-ms exposure time was used, with the exception of the P. ornatus fdsRNA injections, where the highest exposure (60 ms) had to be used as the fluorescent signal was too low to visualize the processing of the dsRNA.
Tissue specific silencing capacity in juvenile Panulirus ornatus
Juvenile mixed-sex P. ornatus individuals (15–50 g) were injected in the ventral abdominal sinus at the base of the 5th walking leg with either 5 µg/g body weight of ds-IGFBP (n = 6) or an equivalent volume of MilliQ water (n = 6). 24 h post injection, animals were culled on an ice slurry, and the antennal gland (AnG), hepatopancreas (HP), gills (G), white muscle (MS), pleopods (PP), and eyestalk (ES) were dissected and snap frozen in liquid nitrogen for downstream RNA extraction and silencing validation via qPCR (see section “Experimental animals”).
Repeated dsRNA injections into P. ornatus juveniles to prime the RNAi pathway
Juvenile mixed-sex P. ornatus individuals (25–50 g) were injected in the ventral abdominal sinus proximal to the 5th walking leg with either 5 µg/g body weight of ds–IGFBP (n = 6) or an equivalent volume of MilliQ water as a control (n = 6). Prior to this injection, 2–3 pleopods were dissected, snap frozen, and then stored at – 80 °C as a baseline for silencing validation. 24 h post-injection, 2–3 pleopods were again dissected from each animal, snap frozen, and stored at − 80 °C, followed by a second injection of either ds-IGFBP or MilliQ water as above. 24 h after the second injection, a final 2–3 pleopods were dissected, snap frozen, and stored at – 80 °C for downstream RNA extraction and qPCR (see section “Experimental animals” and Supplementary Fig. S2), generating three timepoints where animals received no doses of dsRNA (T0), one dose of dsRNA (T1), or two doses of dsRNA (T2). Pleopods were selected for this analysis as they could be sampled non-lethally to determine how silencing changes across time points in the same individual animals.
Pleopod RNA extraction and RNA sequencing
Pleopods RNA from each silencing timepoint (T0, T1, and T2) collected non-lethally from the same organisms, were extracted using the RNAzol^®^RT reagent (Sigma-Aldrich, #R4533, Australia) supplemented with 1% β-mercaptoethanol as previously described, followed by quality and yield assessment via nanodrop spectrophotometer (Hyde et al. 2019). RNA from each time point (n = 3 per timepoint) was then sent for sequencing at Novogene (Singapore) for paired end sequencing on the Illumina HiSeq2500 platform (Illumina, USA).
Transcriptome assembly, quantification, and differential gene expression analysis
Quality control, assembly, and quantification was performed with an in-house Nextflow (v23.04.2) pipeline. Samples included pleopod RNA from the same three individuals before dsRNA injection (T0, n = 3), 24 h after the first injection (T1, n = 3), and 24 h after the second injection (T2, n = 3). The pipeline includes fastp (v0.23.4) for trimming, and trimming for all reads was performed with the following parameters: ‘–trim_front1 15 –trim_front2 15 –trim_tail1 5 –trim_tail2 5’. Reads from one representative sample from each time point were pooled together for de novo transcriptome assembly. Trinity (v2.13.2) and rnaSPAdes (v3.15.5) were used for assembly, followed by Evidential Gene’s tr2aacds script (v2022.04.05) for redundancy reduction. Assembly quality was then assessed with BUSCO (v5.5.0, arthropoda_odb10 dataset) and rnaQUAST (v2.2.3). Salmon (v1.10.1) was used for quantification, and Python (v3.9.18) was used to generate transcript to gene mappings based on the transcript IDs outputted by Evidential Gene. Tximport (v1.30.0) was used to aggregate transcript level abundance to gene level abundance, and DESeq2 (v1.42.0) was used for performing differential expression analysis. Log2 fold changes (LFC) were shrunk with the ‘ashr’ method as described previously (Stephens 2017). Differentially expressed genes (DEGs) were those with p adjusted < 0.01 and absolute LFC > = 1, and only genes with average expressed level > = 20 TPM were included. A principal components plot (PCA) was produced using counts that had undergone regularized log transformation in DESeq2. Annotation of the reference assembly was conducted with the TransPi pipeline (v1.3.0-rc) using default parameters (Rivera-Vicens et al. 2022). The top BLASTx hit from SwissProt (accessed May 1st, 2023) outputted by the pipeline was used for annotating genes. Heatmaps of the top 50 DEGs were created using the top 25 upregulated and top 25 downregulated annotated DEGS based on LFC. Heatmaps were visualized with pheatmap (v 1.0.12) in R (v4.3.2). Proteins involved in the micro-RNA and small interfering RNA pathways were then specifically identified among the DEGs based on previously identified sequences alongside IGFBP (Banks et al. 2022).
Results
Gene silencing capacity is stage specific in Panulirus ornatus and double stranded RNA uptake may be a limiting factor
As RNAi has been deemed ineffective in spiny lobster species (Banks et al. 2020, 2022; Hyde et al. 2019), an in vivo silencing experiment (with dsIGFBP as a marker for silencing efficiency) was performed on larval and juvenile P. ornatus individuals to better determine silencing capacity and mechanisms in these two life stages. IGFBP was chosen as a marker for silencing due to its high levels of expression in virtually every tissue in P. ornatus (Ventura et al. 2020), and non-lethal phenotype when knocked down, which allows for sampling at any timepoint and in any tissue post dsRNA injection.
In P. ornatus larvae, silencing was ineffective, where the mean IGFBP expression was not significantly different between silenced and control individuals (Fig. 1A). In early juvenile individuals, however, gene silencing with an average of 84% IGFBP knockdown was observed (Fig. 1B). To our knowledge, this is the first instance of effective RNAi in any spiny lobster species. To better understand the difference in effectiveness of gene silencing between life stages, the expression of siRNA pathway specific genes (Dicer 2, Argonaute 2, and SID1) was investigated in these individuals in response to IGFBP-dsRNA. After dsRNA injection into juvenile P. ornatus (DS; Fig. 2A–C), all three siRNA specific genes showed significant upregulation when compared to animals injected with ultra-pure water (NC; Fig. 2A–C). This implies that the RNAi system at the juvenile stage is highly inducible, which seems to correlate well with silencing capacity in these individuals (Figs. 1B, 2A–C).Fig. 1. Insulin growth factor binding protein-encoding transcript (IGFBP) silencing in larval A and juvenile B Panulirus ornatus individuals. Animals were injected with either 5 µL of 5 µg/µL double stranded RNA homologous to IGFBP (DS; n = 6 per group) or an equivalent volume of ultra-pure water (NC; n = 6 per group). Whole body RNA was extracted, followed by cDNA synthesis and qPCR to validate silencing capacity in each life stage, with IGFBP expression relative to 18S indicating silencing efficiency. Different letters represent significant differences between groups, and error bars represent standard deviationFig. 2Relative expression of Dicer 2 (A juveniles; D larvae), Argonaute 2 (B juveniles; E larvae), and SID1 (C juveniles; F larvae) in larval and juvenile stages of Panulirus ornatus following injection with dsRNA (DS; n = 6) or an equivalent volume of ultra-pure water (NC; n = 6). Whole body RNA extraction was followed by cDNA synthesis and quantitative PCR to determine relative expression of key RNAi genes, which was normalized to native expression of 18S. Significant differences in expression are denoted by different letters, and error bars represent standard deviation
In the larvae, while Dicer 2 and Argonaute 2 showed significantly higher expression in dsRNA treated individuals compared to the control (Fig. 2D, E), SID1 showed a comparable level of relative expression (Fig. 2F). Interestingly, when a transcript was unregulated in response to dsRNA, the degree of upregulation varied greatly among individuals in both juvenile and larval P. ornatus (Fig. 2). As SID1 is involved in dsRNA uptake and subsequent systemic RNAi, we devised a fluorescent microscopy assay to visualize dsRNA persistence and localization in P. ornatus larvae.
Panulirus ornatus larvae specifically sequester and expel double stranded RNA
Previous qPCR analysis (Figs. 1, 2) has suggested that dsRNA uptake may be a limiting factor for P. ornatus larvae RNAi, and this hypothesis was investigated by taking advantage of the unique characteristics (large size, transparency, and non-autofluorescence) of the larval stage in this species to determine how these animals process foreign cargo. Larvae were microinjected with either non-specific fluorescent proteins (red) or non-specific fluorescent dsRNA, which revealed a significant difference in how these molecules were processed. Ten minutes after injection with fluorescent proteins (Red Ab), the red fluorescent signal was observed throughout the animals’ body, with a stronger fluorescence closer to the injection site proximal to the posterior end of the cephalic shield (Fig. 3A). The hemolymph carrying the fluorescent proteins also appears to surround structures of the antennal gland but was not seen entering the labyrinth structure until 3 h post injection (Fig. 3A, B; see Supplementary Figs. S4–S7 for higher resolution images). For the fluorescent dsRNA injected animals (fdsRNA), a fluorescent signal was not observed 10 min post injection when using the same parameters (exposure time, gain, etc.) as in the fluorescent protein group (Fig. 3A). Increasing the exposure to the maximum value (60 ms) likewise revealed no observable fluorescent signal in the whole larval body. When specifically focusing and zooming on the antennal gland region, a green fluorescent signal was present, which indicates that dsRNA is being rapidly and specifically sequestered in this region (Fig. 3A). Unlike in the fluorescent protein injected animals, the fluorescent dsRNA appears to surround and enter the labyrinth structure, preventing dsRNA from escaping to surrounding cells and tissues for silencing (Fig. 3A). Within three hours post injection, the fluorescent signal of the labeled dsRNA intensified in the antennal gland, likely as a continually increasing fluorescent dsRNA was filtered from the hemolymph and concentrated in this region (Fig. 3B).Fig. 3. Double stranded RNA (dsRNA) and protein localization in the phyllosoma stage of Panulirus ornatus. Localization of a red labeled antibody (Red Ab) and fluorescent dsRNA (fdsRNA) was assessed at 10 min A and 3 h B following microinjection. The left image for both treatments show the fluorescent signal in the whole body of larvae, and the right image specifically displays the head region, focused on the antennal gland (AntG), which is labeled with a white arrow
RNA interference is highly effective across juvenile Panulirus ornatus tissues
As RNAi in the larval stages of P. ornatus appears difficult to induce (Figs. 1, 2), but is deemed efficient at the juvenile stages, we have progressed with investigating gene silencing in juvenile stages. Across juvenile gill, eyestalk, antennal gland, hepatopancreas, muscle, and pleopod tissues, highly effective gene silencing was observed, with every sample showing significant knockdown of IGFBP and a high degree of silencing (gill: 98%, eyestalk: 94%, antennal gland: 89%, hepatopancreas: 92%, muscle: 97%, pleopod: 97%; Fig. 4). This demonstrates that P. ornatus juveniles possess a robust RNAi capacity following a single dsRNA injection and provides an opportunity to investigate and functionally characterize key genes of interest.Fig. 4. Relative expression of the Insulin Growth Factor Binding Protein (IGFBP) after 24 h following *IGFBP-*dsRNA (DS) or water (NC) injection across multiple tissues [n = 6 per group; gill (A), eyestalk (B), antennal gland (C), hepatopancreas (D), muscle (E), and pleopod (F)] in juvenile Panulirus ornatus. RNA was extracted from each tissue, and expression of IGFBP was normalized with 18S following qPCR. All tissues display highly potent silencing capacity and statistically significant gene knockdown. Different letters represent a statistically significant difference in expression between groups, and error bars represent standard deviation
Genes involved in foreign double stranded RNA processing upregulate specifically after silencing in Panulirus ornatus
Samples for transcriptomic analysis were derived from the pleopods of the same juvenile P. ornatus individuals over multiple timepoints (Fig. 5A) to identify genes involved in RNAi activation, dsRNA processing, and antiviral immunity, while also further validating the effective IGFBP silencing observed (Figs. 1, 4). IGFBP was identified in the pleopod transcriptome and mapped across each timepoint, which showed highly effective and statistically significant gene silencing (Fig. 5B). Knockdown efficiency positively correlated with the number of IGFBP-dsRNA doses, where each timepoint showed an increasing level of IGFBP silencing (Fig. 5B).Fig. 5. Insulin Growth Factor Binding Protein (IGFBP) silencing and pleopod transcriptomics in Panulirus ornatus. A A pleopod from juvenile P. ornatus individuals was dissected and stored, prior to injection with dsRNA for IGFBP. 24 h later, a second pleopod from the same animal was nonlethally collected, followed by a second IGFBP-dsRNA injection. After another 24 h, a final pleopod was collected, to generate three time points, T0 (no dsRNA), T1 (one dose of dsRNA), and T2 (two doses of dsRNA) which were sent for RNA sequencing. B IGFBP was identified in the pleopod transcriptome and compared across all time points to determine differences in expression (transcripts per million; TPM). C Genes related to the micro RNA (miRNA) pathway, Dicer 1 and Argonaute 1 (Ago1), and the small interfering (siRNA) pathway, Dicer 2, Argonaute 2 (Ago2), and SID1, were mined and identified in the pleopod transcriptome. The expression of these genes (TPM) was then determined for each time point, where siRNA specific machinery upregulates following dsRNA administration. Different letters are indicative of significant differences in gene expression, and error bars represent standard deviation
The genes for the core siRNA machinery were also seen to be upregulated in response to dsRNA, as observed earlier (Fig. 2), however, the degree of upregulation was variable (Fig. 5C). Dicer 2 and SID1 showed upregulation after dsRNA injection, though the increase in expression was relatively small when compared to Argonaute 2, which upregulated manyfold (Fig. 5C). Argonaute 2 may be a more critical component of the RNAi pathway in this species, or perhaps higher amounts of protein are required to facilitate efficient gene silencing. The expression of the micro RNA (miRNA) machinery, Dicer 1 and Argonaute 1, meanwhile, remained constant independent of dsRNA exposure (Fig. 5C), despite the fact that these proteins naturally interact with endogenous dsRNA substrates (O'Brien et al. 2018). The specific upregulation of only the siRNA pathway machinery and not the miRNA pathway demonstrates that genes which specifically interact with foreign dsRNA are likely to be differentially expressed in the pleopod transcriptome and are likely to be involved with antiviral defence and immunity (Wang and He 2019; Wang et al. 2003).
Exposure to double-stranded RNA elicits an antiviral gene expression profile in the pleopods of Panulirus ornatus
Differential gene expression (DGE) analysis was performed in order to identify any potential dsRNA interacting factors tp. Generally, between T1 (one dose dsRNA) and T2 (two doses dsRNA) samples, very few relevant differentially expressed genes were observed, and their expression profile for genes of interest was largely the same (Fig. 6; see Supplementary Fig. S3, Supplementary file S2). In addition, PCA analysis showed that the T0 samples clustered tightly away from the T1 and T2 samples were overlapped and scattered (See Supplementary Fig. S2). As a result of this, DGE analysis was performed between T0 (no dsRNA) and T1 (one dose dsRNA)/T2 (two doses dsRNA) samples, and the top 50 differentially expressed genes of T0 vs T1 were visualized. Many of the genes downregulated in response to dsRNA (T1; T2) were related to DNA replication, specifically the induction of DNA synthesis (Tada and Blow 1998) (blue asterisk; Fig. 6). Three extracellular matrix glycoproteins, Cuticle Protein 7, Papilin, and Fibrillin-2 (green asterisk) (Fessler et al. 2004; Godwin et al. 2019; Willis 1999), were also significantly downregulated after dsRNA exposure, alongside many proteins related to metabolism and ion transport (Fig. 6). As expected, IGFBP was also among the most downregulated transcripts after exposure to *IGFBP-*dsRNA (Fig. 6; gold asterisk).Fig. 6. Top 50 differentially expressed genes between Panulirus ornatus individuals before exposure to double stranded RNA (dsRNA) vs. one dose of Insulin Growth Factor Binding Protein (IGFBP; gold asterisk) dsRNA 24 h post injection across multiple time point samples (T0; no dsRNA, T1; one dose dsRNA, T2; two doses dsRNA). Gradient color shading indicates these expression levels, which have been scaled by row. Transcripts of interest are labeled with a blue (related to DNA synthesis), green (extracellular matrix glycoproteins), or red (dsRNA interacting) asterisk
Many of the upregulated transcripts in the T1 and T2 samples were related to metabolism and growth, which may be compensatory mechanisms more related to silencing IGFBP due to its role in regulating these processes (Rosen et al. 2013), rather than dsRNA exposure (Fig. 6). Two putative dsRNA interacting proteins upregulated in T1 and T2 samples were observed; however, which include Argonaute 2 (known RNAi protein) and RNA exonuclease 1 homolog (red asterisk; Fig. 6). Two proteins possibly involved in immune activation (which may result from dsRNA exposure), Adenylate cyclase and Serine protease easter (Shan et al. 2023), were similarly upregulated after dsRNA exposure (Fig. 6). In the complete list of significantly differentially expressed genes (see Supplementary file S2), several additional transcripts related to the RNAi mechanism and dsRNA exposure, which were upregulated in T1 and T2 samples, were also identified, as detailed in Table 1.Table 1. Significantly upregulated transcripts after double stranded RNA exposure, which may relate to RNA interference or antiviral immunityTranscript annotationPutative functionRNA interference pathway?Endoribonuclease dicerdsRNA cleavage (Ciechanowska et al. 2021)YesProtein argonaute-2mRNA cleavage, silencing (Wu et al. 2020)YesSID1 transmembrane family 1dsRNA uptake (Maruekawong et al. 2018)YesRNA exonuclease 1 homologPossible antiviral roleNoRNA/RNP complex-1 interacting phosphataseDicer binding partner (Duchaine et al. 2006)YesProtein NazoAntiviral effector (Goto et al. 2018)NoNFX1-type zinc finger-containing protein 1Viral sensing and antiviral immune activation (Blasi et al. 2022)NoSmall RNA 2'-O-methyltransferaseRISC assembly (Horwich et al. 2007)YesExonuclease mut-7 homologsiRNA production (Gu et al. 2009)Yes5’-3’ exoribonuclease 1mRNA decay, possible RISC recycling (Orban and Izaurralde 2005)Yes
Discussion
Limiting mechanisms of RNA interference in Panulirus ornatus larvae
This study, to our knowledge, is the first instance of successful RNAi induction in P. ornatus; however, potent silencing was observed solely in juvenile individuals, and not in late-stage larvae (Fig. 1). Silencing capacity has been linked closely with life stage in insect species (Cooper et al. 2019), where in Drosophila melanogaster, for example, larval silencing is largely ineffective via dsRNA injection, while effective RNAi was observed in adult flies (Goto et al. 2003; Miller et al. 2008). In Drosophila suzukii, however, when dsRNA was orally administered, larvae showed greater silencing efficiency than in adults (Taning et al. 2016). The data presented here represent the first time that silencing capacity has been correlated with life stage in any crustacean species, owing greatly to the large size of P. ornatus larvae, which permits accurate injection of dsRNA and allows for higher quantities of RNA to be extracted to validate RNAi efficiency.
In the present study, we have identified several key findings that together provide a rationale for the lack of a gene silencing effect in P. ornatus larvae. The intracellular RNAi components (Dicer 2 and Argonaute 2; Fig. 2) showed significant upregulation after dsRNA injection in both juvenile and larval life stages of P. ornatus (Fig. 2). SID1, however, exhibited similar upregulation uniquely in the juvenile individuals (Fig. 2C), while in the larvae, dsRNA does not appear to enhance SID1 expression (Fig. 2F). As silencing is effective only in juveniles, this suggests that the expression of SID1 and consequently dsRNA uptake may be one of the limiting factors in the RNAi response of P. ornatus larvae. Alternative routes of dsRNA uptake, such as the receptor-mediated endocytosis mechanism described in L. vannamei (Maruekawong et al. 2022, 2019), have been described, which could present in P. ornatus larvae and explain the lack of SID1 modulation (Fig. 2F). If this were the case, however, silencing efficiency would likely be higher in P. ornatus larvae, as following dsRNA injection, Dicer 2 and Argonaute 2 were upregulated and would be able to initiate a potent RNAi response. SID1, although not upregulated by dsRNA exposure, is still expressed in P. ornatus larvae; therefore, limited dsRNA uptake may occur in this life stage. Such a level of initial dsRNA uptake could be sufficient for Dicer 2 and Argonaute 2 to upregulate, while insufficient to modulate SID1 expression (Fig. 2D–F). This is particularly interesting, as previous transcriptomic analysis showed that Dicer 2, Argonaute 2, and SID1 had a comparably low level of expression in the larval and juvenile stages of P. ornatus (Banks et al. 2022). The data presented here, within the context of the previous transcriptomic analysis (Banks et al. 2022), suggests that dsRNA ‘activates’ the RNAi system in P. ornatus, and that the basal expression of the core machinery is not necessarily a determinant of silencing efficiency.
In addition to the stage specific modulation of SID1 expression, the way that dsRNA is processed by the different larval tissues may be important in the RNAi response of P. ornatus. Injection of fluorescent dsRNA demonstrated that dsRNA is not highly persistent in the hemolymph of P. ornatus larvae and localizes rapidly to the antennal gland, which likely competes with SID1 mediated uptake (Fig. 3A). SID1 channels transport dsRNA passively via diffusion (Shih and Hunter 2011), so dsRNA may initially enter cells rapidly upon injection in a concentration dependent manner (Shih et al. 2009). As dsRNA is removed from the hemolymph and sequestered in the antennal gland however (Fig. 3), the concentration of intracellular dsRNA may quickly begin to reach equilibrium with that in the hemolymph, which would dramatically reduce dsRNA bioavailability and uptake potential. In insects, a dsRNA degrading nuclease which inhibits the RNAi response was shown to upregulate prior to and in greater levels than Dicer 2 and Argonaute 2 after dsRNA exposure, which indicates dsRNA-interacting genes may upregulate sequentially after silencing (Guan et al. 2018). A similar mechanism may be present in the larvae of P. ornatus, where small amounts of dsRNA can quickly trigger the upregulation of Dicer 2 and Argonaute 2 (Guan et al. 2018) before being sequestered by the antennal gland (Fig. 3) but a greater period of dsRNA exposure might be needed to upregulate SID1 to comparable levels (Fig. 2D–F),.
The mechanism by which dsRNA is sequestered in the antennal gland is likely a major limiting factor in the silencing capacity of P. ornatus larvae, as systemic RNAi is being prevented (Fig. 3). The antennal gland's function is excretion, so sequestered dsRNA may be subsequently released into the surrounding media, preventing interaction with somatic cells and tissues. This mechanism may occur in part by a receptor highly/specifically enriched in the antennal gland, which is able to bind and rapidly sequester dsRNA. Using the known receptor-mediated endocytosis dsRNA uptake mechanisms (in both insects and crustaceans Aung et al. 2011; Maruekawong et al. 2022, 2019; Saleh et al. 2006; Wynant et al. 2014; Ye et al. 2021)) as a foundation for in silico analysis using existing transcriptomic data for P. ornatus (Hyde et al. 2019; Ventura et al. 2020) may aid in identifying possible dsRNA receptors in the antennal gland of the larvae in future studies. Additionally, the fluorescent signal of the labeled dsRNA was significantly quenched in P. ornatus larvae within as little as ten minutes and could only be detected with substantial overexposure of the antennal gland region, which suggests much of the fluorescent dsRNA was rapidly removed from the animal. dsRNAses may participate in this mechanism, as degradation could dissociate the fluorophore from the larger dsRNA strand, which may enable rapid excretion via the antennal gland, thereby diluting the fluorescent signal. This concept is supported by studies in insects, where dsRNAses are abundant in the hemolymph of some insect taxa, which adversely affects RNAi (Garbutt et al. 2013; Singh et al. 2017). It is difficult, however, to assess dsRNAse concentrations in the larvae of P. ornatus, since their anatomy hinders extraction of large amounts of clean hemolymph for enzymatic studies. Techniques such as encapsulation of dsRNA in chitosan nanoparticles (Zhang et al. 2010) or generating ribonucleoprotein complexes derived from dsRNA binding proteins from crustacean hemolymph (Banks et al. 2020) could be used to increase dsRNA persistence in P. ornatus larvae in future studies.
We have postulated some mechanisms of how silencing might be less effective in phyllosoma, but it remains unclear why P. ornatus larvae display a markedly different RNAi response compared to that of juveniles. The larvae and juvenile populations exist in completely different ecological niches (Dennis et al. 1998), and as such require specialized culture conditions and diets in aquaculture to reflect the different ecology. The larvae of P. ornatus inhabit the epipelagic zone (Dennis et al. 2002), while the juvenile–adult stages are found in shallow benthic habitats (Dennis et al. 1998). Early juvenile individuals appear to initially prefer a solitary lifestyle in situ (Dennis et al. 1998), but become highly gregarious with age (Kanciruk 1980; Zimmer-Faust and Spanier 1987), which greatly increases conspecific interaction. In decapods, viruses can be transmitted orally by cannibalism/ingestion of infected individuals or via direct contact (Qiu et al. 2023; Sanchez-Paz 2010), so the gregarious nature of P. ornatus juveniles may have necessitated a functional antiviral RNAi response. The larvae of P. ornatus meanwhile are largely planktonic, migrating horizontally in response to ocean currents, and only migrate vertically in the water column as a phototactic response (Phillips and Sastry 1980). The potential for conspecific interactions is therefore reduced, hindering any viral transmission through the larval population. This is compounded by the fact that P. ornatus larvae primarily feed on zooplankton (Phillips and Sastry 1980), which may inhibit viral transmission through ingestion of infected decapods/conspecifics. The only known spiny lobster virus, Panulirus argus Virus 1 (PaV1), has been detected solely in non-larval stages (Kough et al. 2015), which may support the hypothesis that virus transmission among Panulirus spp. larvae is uncommon. The RNAi response in the larvae of P. ornatus may therefore have co-evolved with this viral ecology, and opted for immune mechanisms which handle viruses through yet to be elucidated mechanisms (possibly linked to the rapid sequestration of dsRNA in the antennal gland observed in Fig. 3), rather than combating viruses directly with the siRNA pathway. Indeed, there is evidence that immune gene expression profiles do differ between life stages in crustaceans (Chen et al. 2023; Jiravanichpaisal et al. 2007; Quispe et al. 2016). These studies were performed in prawns however which have far more abbreviated larval stages than spiny lobsters. This highlights the need for future research into immune gene profiles of the phyllosoma which may in turn facilitate understanding of the RNAi mechanism.
Systemic RNAi and antiviral gene regulation in juvenile P. ornatus
Unlike in the larval stage, the juvenile P. ornatus individuals show extremely effective silencing across tissues, and at the whole organism level (Figs. 1B, 4). The injection site of the dsRNA was at the anterior end of the abdomen, and distal tissues such as the gill, eyestalk, and antennal gland showed highly significant gene knockdown, which indicates a robust systemic RNAi capacity in juvenile P. ornatus (Fig. 4). It was also observed following transcriptomic analysis to further validate silencing capacity, that the miRNA pathway machinery was expressed independent of silencing, and only the siRNA pathway specific machinery was upregulated after dsRNA exposure (Fig. 5C). As the miRNA pathway also uses dsRNA as a substrate, this indicates that only genes that interact with foreign dsRNA via the siRNA pathway specifically were likely to be upregulated, alongside genes that may be involved in the antiviral immune response, as dsRNA is a known viral PAMP (Kingsolver et al. 2013; Nie et al. 2018). To support this, several genes related to the siRNA pathway were significantly upregulated following dsRNA exposure, which have not been well characterized in decapods (Fig. 6; Table 1). Some of these proteins were identified in the shrimp Crangon crangon (Christiaens et al. 2015) as well; however, in both C. crangon and P. ornatus, interestingly the siRNA pathway specific protein R2D2 (Liu et al. 2003) was not identified. R2D2 is involved in loading siRNAs into the RISC (Yamaguchi et al. 2022) and is critical to the RNAi pathway in D. melanogaster, so its absence in two decapod species may indicate that another protein fulfills this function in crustaceans. The classic models for invertebrate RNAi pathway research are C. elegans and D. melanogaster, which are evolutionarily distant from crustaceans, so the expansion of the putative siRNA pathway components here further elucidates these mechanisms in a decapod specific context.
A number of genes that may be involved in P. ornatus’ antiviral immunity (excluding the identified siRNA pathway genes) were also differentially expressed, including antiviral effector proteins and host machinery which viruses may parasitise (Fig. 6; Table 1). Broadly, these findings suggest that dsRNA can modulate the immune system of P. ornatus and could be used as a prophylactic against viruses which may emerge as the aquaculture of this species matures. NFX1-type zinc finger-containing protein 1 (Table 1), alongside several other RNA helicases (see Supplementary file S2), were upregulated in response to dsRNA. RNA helicases are known to be involved in the activation of the antiviral immune response and may act as sensors of viral infection by recognizing dsRNA which can further enhance immune activity (Baldaccini and Pfeffer 2021; Blasi et al. 2022). Nazo was also identified among the dsRNA upregulated transcripts, which is a known antiviral effector protein, though its mechanism is unknown (Goto et al. 2018). Similarly upregulated is RNA exonuclease 1 homolog, though its role in the innate antiviral response is unclear. The homologous protein in humans belongs to the same protein family as ISG20, which is an exonuclease known to be involved in innate antiviral defence against RNA viruses by directly degrading viral RNA (Deymier et al. 2022; Espert et al. 2003). The upregulation of RNA exonuclease 1 in P. ornatus in response to dsRNA may indicate some degree of conserved function as an antiviral effector protein.
Many proteins which regulate DNA replication and cell division were shown to be downregulated after dsRNA exposure (Fig. 6), which may constitute a mechanism to avoid parasitism of replication machinery by DNA viruses (Idrees et al. 2023). Factors involved directly in DNA replication complexes, such as DNA polymerase alpha subunit (Muzi-Falconi et al. 2003) and proliferating cell nuclear antigen (Kelman 1997), were downregulated after silencing, alongside those involved in biosynthesis of dNTPs (thymidylate synthetase (Chu and Allegra 1996)), and numerous DNA licensing factors (Song et al. 2022). As DNA viruses can directly use intrinsic host replication machinery, P. ornatus may have evolved to downregulate a wide variety of factors which contribute to DNA replication in response to dsRNA. The immune signaling pathways which naturally sense and respond to foreign virus dsRNA (Chen and Hur 2022) could also lead to downstream transcriptional repression of DNA replication to prevent viral parasitism. Also downregulated after silencing were cuticle protein 7 (Judith 1999; Stiles 1991), fibrillin 2 (Godwin et al. 2019; Ramirez and Sakai 2010), and papillin (Fessler et al. 2004), which are glycosylated extracellular matrix proteins, and may act as viral attachment sites or entry routes into the cell (Li et al. 2021; Zhai et al. 2024). If so, their downregulation in response to dsRNA may be a mechanism to inhibit further viral entry and uptake in P. ornatus juveniles.
A large number of the differentially expressed transcripts were related to energy metabolism and nutritional pathways, such as the crustacean hyperglycaemic hormone (Chen et al. 2020), which was upregulated after silencing. It is difficult to determine, however, if these genes are related to dsRNA exposure and subsequent activation of RNAi and antiviral pathways, or are differentially expressed in response to IGFBP silencing. IGFBP regulates the function of insulin-like growth factors involved in growth and metabolism, so silencing IGFBP may have led to the activation of compensatory mechanisms to maintain energy production (Chandler et al. 2017; Hwa et al. 1999). Likewise, IGFBP silencing may have resulted in downstream effects which are not necessarily compensatory to energy production. Future studies should replicate this experiment but expose P. ornatus individuals to dsRNA which cannot silence endogenous genes to determine more precisely the specific differential gene expression profile which results following silencing.
Conclusion
In summary, this study highlights the distinct differences in RNA interference (RNAi) efficacy between larval and juvenile P. ornatus. While the larvae exhibit limited RNAi due to potential sequestration of dsRNA and inefficient uptake mechanisms, juveniles demonstrate a robust and systemic RNAi response. The upregulation of specific genes associated with the siRNA pathway in juveniles suggests a sophisticated mechanism for gene silencing and antiviral defence that is not present in the larval stage. These findings not only advance our understanding of RNAi in crustaceans but also open new avenues for enhancing antiviral defences in aquaculture practices for P. ornatus. Further research is required to optimize dsRNA delivery in larvae and to explore the broader implications of RNAi for disease management in this species.
Supplementary Information
Below is the link to the electronic supplementary material.Supplementary file1 (DOCX 2267 kb)Supplementary file2 (XLSX 118 kb)