Author Correction: RNAi mediated myosuppressin deficiency affects muscle development and survival in the salmon louse (Lepeophtheirus salmonis)

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

The salmon louse, Lepeophtheirus salmonis, (Copepoda: Crustacea), is a marine ecto-parasite of Atlantic salmon (Salmo salar) that causes large economic loss to the farmed salmonids 1 and constitute a major threat to the wild fish populations 2 . Salmon louse feeds on the host's mucus, skin and blood, decreasing fish growth and causing lesions that constitute the gateway for other pathogens 3 . Until now, fight against salmon louse has been based mainly on the medicinal treatment and various physical methods 4 . However, the medicine-based delousing strategy has progressively lost the efficacy, since the lice have developed resistance against the various chemotherapeutics 5 , while, physical methods cause stress and damage to the fish, leading to high post-treatment mortality. Therefore, there is an urgent need to develop new and efficient means to control salmon louse infections.
In recent years, increased interest has been aimed at endocrine-based pest control strategy that involves development of drugs targeting neuropeptides in invertebrates [6][7][8] . Neuropeptides, their receptors and enzymes involved in their metabolism, have long been considered as potential targets for the development of new chemotherapeutics, due to their various physiological and behavioural functions, conserved structure and widespread distribution in Arthropods, and the possibility that such targets could deliver specific and environment friendly treatment [9][10][11][12][13] . However, until now, no commercial insecticide utilising neuropeptide's agonists or antagonists has been discovered or used, mainly due to the missing information and lack of full understanding of neuropeptide and their receptors function and structure. To strengthen the development of such pest control strategy, knowledge about neuropeptide composition, their biological activities and receptor-neuropeptide interactions is of critical importance [9][10][11][12][13] .
Neuropeptides are an important class of messengers and hormones acting as critical signal transducers and modulators in a wide range of developmental, physiological and behavioural processes. The variety of functions www.nature.com/scientificreports www.nature.com/scientificreports/ neuropeptides play in the animals suggest that alternations in peptide expression, synthesis, or signalling could result in abnormalities or impaired function resulting in decreased fitness or death of parasite such as the salmon louse.
Myosuppressin belongs to the FMRFamide-like peptides family with the consensus decapeptide sequence at the C-terminal end of X 1 DVX 4 HX 6 FLRFamide (where X 1 = pQ, P, T or A, X 4 = D, G or V, and X 6 = V or S) [32][33][34] . The last six amino acids of the decapeptide constitute the active core that is identical in all Arthropods studied so far and sufficient for antimyotropic activity 30 . Myosuppressin decapeptides mediate their function by binding to myosuppressin receptors that belong to the G-protein coupled receptors 35,36 .
The majority of publications describe structure-activity relationship and function of myosuppressin in controlling various muscles, whereas, myosuppressin's role in development, and effects of its deficiency or dysfunction are poorly investigated. Thus, we investigated myosuppressin function in the animal development in a number of developmental stages of salmon louse Lepeophtheirus salmonis (LsalMS), from free-living planktonic stages to parasitic stages. We knocked-down LsalMS by RNA interference at various developmental time points and looked for developmental and behavioural abnormalities. Using MS-specific antisera and quantitative qRT-PCR we mapped gene expression and peptide localization to nervous system, reproductive system and alimentary tract.

Results
Sequence analysis and molecular phylogeny of the L. salmonis myosuppressin. The gene encoding myosuppressin in Lepeophtheirus salmonis (LsalMS) was identified by the standard Blastp search of the collection of protein predictions from expressed sequence tags (EST) (sealouse.imr.no) using Drosophila melanogaster myosuppressin peptide (AAF56283.1) as the query sequences. The result revealed one candidate with low e-value (e = 7e-08). To obtain the full cDNA sequences of the identified gene, we carried out 5′ and 3′ Rapid Amplification of cDNA Ends (RACE) using PCR primers specific for exon sequences obtained from sealouse.imr. no database (Supp. Table 1). Genomic organisation of LsalMS is shown on Fig. 1a. The final LsalMS sequence has been submitted to GenBank (GenBank accession number: MH748227). The full cDNA sequence of LsalMS is 763 nucleotides long. It contains a 5′-UTR of 154 bp and a 3′-UTR of 195 bp, including a polyA. The deduced open reading frame is 414 nucleotides long and encoded a prepropeptide of 137 amino acid residues. A signal peptide was identified with the cleavage site located between the 39 and 40 amino acid (S/R), giving a precursor peptide of 98 aa. The deduced amino acid sequence of the decapeptide, identified based on multiple sequence alignment, is flanked by abnormal proteolytic processing sites, tribasic (KRK) and dibasic (KR) sites located before and after the decapeptide 52 (Fig. 1b). Thus, the encoded L. salmonis decapeptide begins with glutamic acid (E) and ends with a phenylalanine residue (F), and it follows the consensus pattern for myosuppressin peptides (Figs 1b  and 2a). In the majority of insects 14,20,[22][23][24][25][26]29 , the mature decapeptide starts with glutamic acid (E) or glutamine (Q), and both amino acids can be changed to pyroglutamic acid during post-translational modifications 53,54 . The glycine residue at the end of the decapeptide is necessary for amidation by peptidylglycine α-amidating monooxygenase 55 . The final sequence of L. salmonis mature peptide is predicted to be pEDVDHVFLRFamide. No transmembrane regions, specific functional domain, or N-glycosylation sites were detected in the predicted peptide sequence.
The alignment of myosuppressin propeptide with propeptides from other species revealed the same organization, with a similar length of 70-100 amino acids. The alignment also showed that the greatest identity is concentrated in the C-terminal region, which corresponds to the mature decapeptide (Fig. 2a). Comparison of full prepropeptide LsalMS cDNA sequence with the genomic sequence of L. salmonis showed the presence of four exons in the LsalMS gene (Fig. 1a).
The LsalMS peptide precursor was compared with those of other species, revealing 38% identity to Drosophila melanogaster, 55% to the Hemiptera Nilaparvata lugens, and 34%, 35% and 37% to other Crustaceans, Scylla paramamosain, Homarus americanus and Procambarus clarkia respectively. The LsalMS decapeptide has identical sequence to Zopat-MS and differs from LMS (leucomyosuppressin, the first identified myosuppressin from the cockroach Leucophaea maderae) by only one amino acid residue at the N-terminal end.
The propeptide sequences were used to construct a phylogenetic tree (for accession numbers see Table 1), shown in Fig. 2b. FMRFamide peptides were used as an out-group. In the rooted bootstraped tree, LsalMS groups together with insects, the sister group to the Crustaceans. www.nature.com/scientificreports www.nature.com/scientificreports/ Ontogeny of LsalMS. Quantitative real-time PCR was used to analyse the ontogenic expression of LsalMS in all L. salmonis stages. Salmon louse has a complex development cycle consisted of eight stages: three free-living larval stages (nauplius I, II and copepodid) and five parasitic stages, attached to the fish (chalimus I, II, preadult I, II and adult) 56 . The level of LsalMS in nauplius I was set to 1 and considered as a baseline for quantification of expression in other stages. The lowest expression level of LsalMS transcripts was detected in adult females, whereas the copepodids showed the highest expression levels (Fig. 3a). In chalimus I, the first tested parasitic stage, the level of transcript decreased dramatically and it reached only 7% of the expression observed in the planktonic copepodids. In chalimus II, the transcript level decreased even more (50% of the value observed in chalimus I) and it persisted on this level in all following parasitic stages (Fig. 3a).

qRT-PCR of LsalMS on tissue extracts from adult lice. Quantitative RT-PCR showed the highest
LsalMS expression in the brain. Lower expression was detected in the mouth, intestine and parts of the reproductive system, mainly in the ovaries and testes and in female genital segment. Very low expression was detected in immature eggs extracted from female genital segment (set as 1 and used as a baseline for the whole analysis), which indicates that LsalMS transcripts detected in the genital segment were localised mainly in the oviduct (Fig. 3b).

LsalMS peptide detection. Immunohistochemistry analysis was performed to identify the localization of
LsalMS peptide in copepodids, adult female and male lice. In all tested stages, the specific signal was detected in the brain, in the internal area consisting of axons (Fig. 4c). In the adults, LsalMS peptide was also detected in the intestine and reproductive organs, in cells that might have neural or endocrine origin (Fig. 4b,c). However, the exact cell type needs to be determined further.
RNAi of LsalMS in copepodids. RNA interference was used to assess effects of LsalMS down-regulation in the larval stages. dsRNA was introduced by soaking during moult from nauplius I to nauplius II. After moulting, animals were left to develop to the copepodid stage and then investigated for any phenotypic defects. Although, www.nature.com/scientificreports www.nature.com/scientificreports/ the qRT-PCR analysis revealed significant down-regulation (more than 90%) of LsalMS in larvae treated with dsLsalMS in comparison to the control animals ( Fig. 5a), no visible morphological or behavioural abnormalities were observed. However, immunohistochemistry with LsalMS antisera showed substantial reduction in staining intensity in the dsLsalMS treated copepodids' brain, indicating decreased level of the peptide (Fig. 5b).
While animals treated with dsLsalMS showed no visible phenotype at the free-living copepodid stage, we have investigated if down-regulation of LsalMS could have any effect on attachment ability and further development of the lice. Copepodids from each treatment group were used to infect fish (n = 3 for each treatment group). Development of lice was followed daily by visual observation. While, the control animals developed normally, dsLsalMS treated animals were able to reach chalimus II stage, but none of them were able to reach preadult I stage. The experiment was terminated after 30 days when all dsLsalMS treated lice had disappeared from the fish.
The infection experiment was repeated in order to investigate phenotypic defects in chalimus stages caused by dsLsalMS treatment. In total, nine fish were infected with lice from each treatment group. Lice were collected after 8, 12 and 18 days post infection (n = 3 fish per time point, per treatment group), when control lice were in the middle of chalimus I, chalimus II and preadult I/II stage, respectively. Animals were counted, photographed  Table 1) were used to generate multiple sequence alignment with ClustalOmega. Identical amino acids are highlighted in black and similar amino acids are highlighted in grey. Databases corresponding to the mature peptide are in a red frame (b) Analysis of phylogenetic relationship of Lepeophtheirus salmonis myosuppressin propeptide with sequences from other Arthropods generated by Bayesian method. LsalMS in red. FMRFamide sequences of Arthropods species were used as an outgroup. Branch numbers represent bootstrap values in percent and bar represents 0.2 substitutions per site. Branch lengths are proportional to sequences divergence. The accession numbers listed in Table 1. www.nature.com/scientificreports www.nature.com/scientificreports/ and carefully examined for potential morphological abnormalities. Similar infection rates were observed in both treatment groups, where in average 18 animals were recovered from each fish at day 8 (control 18,3 (±8) vs. dsL-salMS 17,7 (±5,5)) and 27 animals at day 12 post infection (control 28,3 (±4,2) vs. dsLsalMS 26 (±6)) ( Fig. 5c).  www.nature.com/scientificreports www.nature.com/scientificreports/ While control animals developed normally, delayed development was observed in dsLsalMS treated group, and copepodids were still present at day 12 post infection. A significant difference (p value = 0.001, n = 3 for each group) in the number of collected lice was obvious at day 18 post infection, where 96% less dsLsalMS treated lice were recovered from fish than controls, and only 34% of these were at the preadult stage (Fig. 5c).
Down-regulation efficiency was evaluated in lice collected at each time point by qRT-PCR. Comparison of LsalMS transcript level in dsLsalMS treated lice was done with controls in the corresponding developmental stage. Since copepodids were not present in the control group, they were excluded from analysis. To obtain sufficient number of preadult animals, infection was repeated (100 copepodids per fish, 6 fish used), and all lice were collected on day 18 post infection. On day 8, transcript level was tested in chalimus I (n = 10 for each group), on day 12 in chalimus II (n = 10 for each group) and on day 18 in preadult I males and females (n = 5♀/5♂ for each group). While majority of dsLsalMS treated chalimus I and II revealed significant down-regulation of transcript level, few animals (20% chalimus I, 10% chalimus II) showed LsalMS transcript levels comparable to controls. On the other hand, all recovered preadult I lice from dsLsalMS treatment group showed expression level similar to control group (Fig. 5d), suggesting that only animals with insufficient LsalMS down-regulation were able to survive and progress with development. Thus, the knock-down of LsalMS clearly affects normal developmental progress of the lice.

RNAi of LsalMS in adults.
To investigate phenotypic defect in adult animals, we injected preadult I animals of both sexes with dsRNA. In total, 160 pre-adult I female and 160 pre-adult I male lice were injected with dsL-salMS, in addition to animals injected with control dsRNA (four separate experiments, in total eight fish for each fragment). Animals were placed on the fish and observed daily. Lost animals were collected from the filters at the water outlet. The experiment was ended after 16 days, when all dsLsalMS treated females were lost from the fish and only a few males were still visible. In average, 70% of control females and 90% of control males were recovered from the fish, while only 10% of dsLsalMS treated males were still on the fish at the termination. The majority of female lice were lost at the preadult II stage or during moulting to adult stage (Fig. 6a). None of them were able to successfully complete moult to the adult stage.
While the control animals developed normally, they were able to swim and successfully attach to the fish and flat surfaces, dsLsalMS treated animals failed to do so. In all dsLsalMS treated animals (collected from filters and fish) the intestine wall appeared thinner, with numerous granules and gut content was reduced or absent. In adult males, spermatophores were smaller and vasa deferentia appeared swollen (Fig. 6b). The LsalMS transcript level in lice collected from fish/filters, evaluated by qRT-PCR, was 90-95% lower than in control animals (Fig. 6c). Close examination of the intestinal movement of recovered dsLsalMS treated preadult II females revealed significantly increased frequency of contractions (p = 0.0005) (Samples of the movies in Supp. Video 1). While in control lice 0.53 contraction per second was observed (n = 10), 0.64 contractions per second were counted in lice injected with dsLsalMS (n = 12) (Fig. 6d). www.nature.com/scientificreports www.nature.com/scientificreports/ To evaluate mating potential of adult males, pre-adult I male lice were injected with dsLsalMS or control dsRNA, and placed together with non-injected females. The experiment was terminated after 6 weeks when all males reached adult stages and all control females had developed egg strings. In total, four males (13.3%) www.nature.com/scientificreports www.nature.com/scientificreports/ were recovered from the dsLsalMS group in comparison to 13 (43.3%) in the control group. Whereas all control females were mated, had spermatophores attached to the genital segment, and extruded normal, fertilized egg strings, with viable offspring, none of the females placed together with dsLsalMS injected males had spermatophore attached (Fig. 6e), and all collected egg strings (3 pairs) were not fertilized. However, typical mating behaviour, evident as pairing with females was observed in males treated with dsLsalMS.
Representative animals (preadult II female and adult males) were sectioned and stained with toluidine blue. All animals treated with dsLsalMS were smaller than controls and revealed abnormal morphology compared to the controls. The most striking abnormality in both sexes was significantly reduced content of various tissues. A severe reduction of muscle tissue was evident, fibres of longitudinal muscle stretching from cephalothorax to genital segment were discontinuous with visible ectopic tissue (Fig. 7a,b). The intestinal wall contained numerous vacuoles, and the connectivity between cells seemed affected (Fig. 7c,d). Not all cell types were identified in the www.nature.com/scientificreports www.nature.com/scientificreports/ intestinal wall of dsLsalMS treated animals, and circular muscles surrounding the intestine were thinner and lacked surrounding tissue (Fig. 7c,d). Numerous vacuoles were present in the brain. Moreover, the brain was smaller and empty spaces were visible in the inner part, where LsalMS was localized (Fig. 7e,f). The cellular structures in the subcuticular tissue were strongly reduced, and separated with wide empty spaces. Numerous cells and the tegumental glands are missing or abnormally formed (Fig. 7g,h).
In the transverse section of the male genital segment large, empty spaces were visible (Fig. 8b,c). The spermatophores at different stages of maturity were still present but they were abnormally shaped and smaller compared to the controls. While all typical layers of spermatophore wall were identified, they were much thinner than in the control animals (Fig. 8b,c). Males treated with dsLsalMS had smaller testes, with tubules tightly compressed. In the posterior zone of testes, spermatogonia were visible and the midzone contained primary spermatocytes. However, the anterior zone with spermatids contained a higher number of "A" granules than in control animals (Fig. 8d,e). The "A" granules are normally formed from degenerating spermatocytes 57 , indicating increased degeneration process in dsLsalMS treated males.
LsalMS knock-down preadult II females were generally frailer than corresponding controls, with reduced tissue content in the whole body (Fig. 9a,b). Moreover, the ovaries of dsLsalMS treaded females showed substantial changes. The ovaries were smaller and the ovarian tubules thinner and their lumen narrower compared to control animals (Fig. 9c-f). The chromosomes in the oocytes were clearly visible, but the nuclei seemed to lack nucleoplasm (Fig. 9e,f).

Discussion
Myosuppressin peptide has previously been detected in many Arthropod species, but was reported to be missing in the salmon louse Lepeophtheirus salmonis 16 . In this study, we report for the first time, the identification of a myosuppressin peptide and characterisation of its deficiency in various developmental stages of the salmon louse. Compelling evidence shows that LsalMS is essential for normal development, reproduction, and feeding of this economically important parasite. Moreover, to our knowledge this is the first report describing complex effects of myosuppressin deficiency on normal development throughout the life cycle in a single species. www.nature.com/scientificreports www.nature.com/scientificreports/ The sequence analysis of LsalMS revealed that this peptide belongs to the FMRFamide related peptide family (FaRPs) with the decapeptide sequence following the general pattern characteristic for myosuppressins identified in other Arthropod species. Protein sequence analysis revealed that the putative decapeptide of L. salmonis, pEDVDHVFLRFamide, is flanked by abnormal proteolytic cleavage sites of tribasic and dibasic amino acids, therefore, unusual processing into the active substance could be involved 58,59 . While the final mature sequence of LsalMS needs to be experimentally validated in the future, the predicted one varies with only one amino acid on the N-terminal end of the peptide from leucomyosuppressin, LMS -QDVDHVFLRFamide 20,21 , and it is identical to one of the peptides from the beetle Zophobas atratus 14 .
Only one gene encoding MS has been identified in Arthropod species, and in the majority of these only one peptide product has been found 20,22-24,26-30 . However, despite of the single copy of the MS gene, in a few species additional copies of MS peptide products have been detected, with distinct tissue-specific distribution and function 14,53,60,61 . The exact mechanism of their production is unclear in most of the cases but post-transcriptional modifications and alternative processing sites could play a role 60 . In our study, only one gene encoding myosuppressin was found in salmon louse, with a single putative decapeptide released from the C-terminal end of the precursor. No alternative splicing or post-transcriptional modifications were detected, thus it seems that a single molecular form could account for all MS activity in L. salmonis tissues.
MS transcript level and tissue-specific distribution have been reported in a number of species, mainly in adult animals 34,62-65 , but to a lesser extent in larvae 34,51,62 . However, none of these studies have investigated the MS levels during the whole life cycle for any species. LsalMS expression was detected in all developmental stages. LsalMS transcripts can be detected just after hatching, in the nauplius I stage, with slightly higher expression in nauplii II stage, to reach the highest (13-fold higher than in nauplii I) expression in the copepodids (Fig. 3). After infection www.nature.com/scientificreports www.nature.com/scientificreports/ of the host, expression of LsalMS gene decreases gradually until adulthood, with adults expressing the lowest level of LsalMS.
The significant variation in the transcript level between copepodid and remaining stages suggests the increased requirement for LsalMS activity in this planktonic and infectious stage. Copepodid is the transitional stage between the free-living and the parasitic phases in salmon louse ontogeny. The development of lice is arrested at this stage until the copepodid finds a suitable host i.e. a salmonid. The successful settlement on the host fish is a critical factor that can interrupt arrest and enable further development. It has been shown that myosuppressin is involved in the regulation of diapause in the cabbage army moth Mamestra brassicae 51 , and the concentration of myosuppressin in the hemolymph after pupal ecdysis was higher in diapause pupae than in nondiapause pupae. The myosuppressin suppresses activation of the prothoracic glands (PGs) stimulated by www.nature.com/scientificreports www.nature.com/scientificreports/ prothoracicotropic hormone leading to diapause 51 . It is therefore possible that LsalMS could possess a similar function in copepodid stage of L. salmonis.
The analysis of the tissue specific peptide distribution suggests that LsalMS plays a role in feeding/digestion and reproduction in L. salmonis. The presence of the myosuppressin transcripts in the tissue surrounding vital organs collected from the adult lice was also validated by quantitative RT-PCR. As expected, the highest transcript level was identified in the extracts from the brain, followed by mouth, intestine, and the reproductive organs: testes and ovaries. The observed LsalMS distribution is consistent with previously reported from other Arthropod species, where the MS peptide was detected in the central nervous system, peripheral neurons and endocrine gla nds 22,50,51,62,65 . Similarly to our results in L. salmonis, the MS-positive nerves were identified in close proximity of the crop muscle, in the crop nerve and in ramifying nerve fibres on the surface of the crop in Drosophila suzukii 64 . Furthermore, in Rhodnius prolixus, MS-positive nerves were detected in the endocrine-like cells in the midgut 50 . However, previous studies did not reveal MS-positive neurons or endocrine-like cells in the reproductive system. In L. salmonis, MS-positive cells were detected in the lumen of tubules of the ovaries. This suggests that in various species MS may play its role in distinct routes, acting as a neurotransmitter directly delivered to the target tissue and/or as a circulatory hormone, released from endocrine-like cells. Nevertheless, the localisation of the LsalMS was consistent with previously described physiological function in regulating contractions of the various visceral muscles and support claims that LsalMS, as myosuppressins in other species, is likely the neuromodulator in the nervous system and plays a role in regulating feeding and reproduction.
Very few data from RNA interference experiments assess effects of the myosuppressin deficiency in Arthropods. In females of B. germanica, RNAi mediated knockdown of LMS caused increase in the heartbeat frequencies 66 , whereas, nothing is known about the influence of the myosuppressin deficiency on the development and growth. Our RNAi studies indicate that myosuppressin directly or indirectly influences proper development, moulting and survival. The RNAi treated L. salmonis larvae, with absence of myosuppressin, were able to settle on the fish in the same rates as control animals, whereas, the development and further growth were affected significantly, with very few animals reaching the pre-adult stage. The survival was significantly reduced and only single animals, those where RNAi mediated knock-down was insufficient, survived to adulthood. Similar results were obtained when pre-adult lice were injected with dsLsalMS. Visual inspection of the intestine of dsLsalMS treated animals in all evaluated stages revealed significantly reduced or no gut content, indicating disturbance of the feeding process. In fact, close examination of the intestinal peristalsis in the recovered dsLsalMS treated preadult II females revealed significant increase in the contraction frequency. Moreover, lice had difficulties with moulting to the next stage, which could be connected to impaired coordination of muscle contractions. The observed phenotype was most severe in the female lice, which were unable to reach the adult stage, therefore, the exact effect of myosuppressin deficiency on reproduction of female lice could not be examined. However, very few male lice reach the adult stage and the complex malformations could be observed, mainly in the genital segment. Adult dsLsalMS treated males placed together with non-injected adult females were not able to complete mating. This effect clearly indicates that lack of LsalMS, apart from other defects, significantly affects mating and reproduction in L. salmonis males. These results are consistent with previously reported function of myosuppressin in reproduction and feeding 45,63,64,67 .
The histological analysis of the adult males and preadult II females indicated developmental abnormalities in several tissues. The most striking feature was severe reduction of the muscle content, with a significant reduction of circular muscles in the intestine, as well as the visceral muscles associated with the vasa deferentia in the genital segment. Moreover, abnormalities in the skeletal muscles were apparent, with discontinuous fibres disturbed by the ectopic tissue. While the function of the myosuppressin on the activity of visceral muscles is well investigated, the influence of deficiency of myosuppressin on muscles development is poorly understood. Moreover, there is very scarce information available regarding myosuppressin function in the skeletal muscles, suggesting that it enhances the force of neurally induced contractions in skeletal muscle 31 . The results from the study in the cockroach Periplaneta americana indicates that the muscles of the legs receive innervation by FMRFamide-related peptide-containing neurons and that may be an additional method of modulating the mechanics of skeletal muscle contraction 68,69 . These findings indicate that myosuppressin has distinct effects on the different types of muscles, inhibiting contractions of the visceral muscles and enhancing the force of contractions in the skeletal muscles.
It was previously shown that muscle mass can dramatically decline in response to pathological signals or lack of normal neuronal signalling [70][71][72] , suggesting that the significant reduction in the content of the muscle tissue in myosuppressin deficient animals could be potentially linked with a lack of the proper neuronal stimulation. In contrary, LsalMS deficiency affects intestinal morphology and function, likely leading to the reduced nutrient uptake, causing decreased tissue development and growth. Inability to acquire nutrients, malnutrition and starvation are well known causes of muscle atrophy in various species [73][74][75] , manifested as decrease in the body size and weight, cell death, intracellular vacuolation and increased intracellular spaces 73,74 . Due to various potential reasons of the observed phenotype in LsalMS deficiency L. salmonis, the exact reasons of muscle mass decline need to be study further.
Data presented here suggests that the LsalMS plays an important role in various physiological processes that are crucial for development, growth and behaviour in the salmon louse. Considering the importance this peptide has in function of vital organs in arthropods, it is not surprising that lack of LsalMS causes major phenotype. The knock-down of LsalMS affects several processes that are critical for survival, like feeding and reproduction. The myosuppressin peptide could be a valuable target for the development of the new and effective anti-lice control method, that could effectively reduce feeding success and reproduction rate of lice and decrease the overall growth of sea lice populations in salmon cages, while targeting adult stages, and significantly decrease the number of lice reaching the adulthood, while targeting myosuppressin signalling in early developmental stages. However, further studies are required in order to unravel the complex roles played by LsalMS in L. salmonis development, www.nature.com/scientificreports www.nature.com/scientificreports/ peptide processing and maturation pathways and interactions with specific receptors to find out how this peptide could be successfully targeted.

Materials and Methods
Ethics statement. The experiments were carried out on salmon louse Lepeophtheirus salmonis, whereas fish Atlantic salmon (Salmo salar) were used only as a host for the experimental animals. All procedures involving animals were performed according to The Norwegian Animal Welfare Legislation. The Animal Ethics Committee by The Norwegian Food Safety Authority approved all experiments.
Salmon lice culture. A laboratory LsGulen strain of salmon lice (Lepeophtheirus salmonis salmonis) was maintained on farmed Atlantic salmon (Salmo salar) 76 . Salmon were hand fed on a commercial diet and reared in seawater with a salinity of 33.5‰ and a temperature of 12 °C. Eggs, nauplii and copepodids were kept in seawater from the same supply. Nauplii were obtained from hatching eggs and kept in single wells in a flow through incubators 76 . Sequence analysis. Candidate L. salmonis myosuppressin peptide gene was identified by blasting (BLASTP) the known propeptide sequence of Drosophila melanogaster (AAF56283.1) against collection of predicted protein sequences from the salmon louse (sealouse.imr.no). Only one high confidence hit was identified (e-value = 7e −08 ).
Full-length cDNA sequence was obtained by rapid amplification of cDNA ends (RACE) using SMARTer ® RACE cDNA Amplification Kit (Clontech) according to the manufacturer's protocol. RACE primers (Supp. Table 1) were designed based on the sequence retrieved from salmon louse genome (sealouse.imr.no), giving 279 bp overlapping fragments at the ends of 5′ and 3′ RACE products (Fig. 1a). Obtained fragments were cloned using the TOPO TA Cloning ® Kit for sequencing (Invitrogen) and sequenced with M13 forward and reverse primers and sequence specific primers used for RACE (Supp. Table 1). Prior to sequencing the PCR products were purified by ExoSAP-it (Affymetrix). Obtained sequences were assembled using MacVector with Assembler vs. 12.7.5 (MacVector, Inc, Software for Scientists, North Carolina, USA). Final sequence of LsalMS transcript, encoding prepropeptide was submitted to the GenBank. LsalMS gene organisation was analysed by aligning obtained cDNA sequence of LsalMS with salmon louse genome sequence (sealouse.imr.no). A precursor peptide sequence was predicted by identification of a signal peptide cleavage site using SignalP 4.1 software (http://www.cbs.dtu. dk/services/SignalP/) 77 . The mature peptide sequence was identified by multiple sequence alignment of the amino acid sequences (accession numbers are listed in Table 1) performed with ClustalOmega 78 .
Phylogeny. The full length of LsalMS predicted precursor peptide sequence was compared to 20 precursors of myosuppressins belonging to other species. Propeptide sequences used in the present study were downloaded from NCBI (http://www.ncbi.nlm.nih.gov/) and their accession numbers are listed in Table 1. Putative domains of LsalMS were predicted and examined by Pfam (http://pfam.xfam.org/). Multiple alignment of the propeptides amino acid sequences was performed with ClustalOmega 78 , and converted to nexus format by Mesquite v. 3.2 79 .
LG + G + I model of substitution was adopted as the optimal evolutionary model, which was predicted by ProtTest 3.4 80 . Phylogenetic analysis was conducted using MrBayes v.3.2 81 . Markov Chain Monte Carlo (MCMC) algorithm was used with four chains running for 1,000,000 generations and sampled every 1000 generations. Output data were analysed with Tracer v1.5 82 , and tree was obtained using FigTree v1.4.3 (http://tree.bio.ed.ac. uk/software/figtree/).

Collection of animals and tissues for ontogenetic analysis.
The following numbers and groups of animals were collected in quadruple samples. Stages: nauplius I, nauplius II and copepodid (free-living) -approximately 200 larvae per sample, chalimus I -30 animals, chalimus II -20 animals, preadult I, II and adult stagessingle animals. Specific tissue and organs were dissected from 50-100 animals per sample from both sexes (brain, mouth, intestine) or only from one corresponding sex (ovaries, testes, genital segments).
Immunohistochemistry. For detection of LsalMS peptide, the following immunohistochemistry protocol was used. Briefly, slides with paraffin sections were incubated at 65 °C for 30 min and afterwards cleared in Histo-Clear II (National Diagnostics) three times for 10 min, following by rehydration in decreasing ethanol series (2X 100%, 96%, 80%, 50%) for 3 min each and then put into distilled water. After a heat-induced antigen retrieval (HIER) (30 min in a 95 °C sodium citrate buffer (10 mM Sodium Citrate, 0.05% Tween 20, pH = 6.0)) 83 , slides were washed two times for 10 min in PBST (Phosphate buffered saline, Sigma Aldrich, with 1% Triton-X) on the shaker. 2% NGS (Normal Goat Serum, Jackson ImmunoResearch Lab) and 2% BSA (Bovine serum Albumin; Sigma-Aldrich) in PBST was used to block unspecific antiserum binding for 2 h at room temperatures. Primary antiserum was diluted 1:500 in PBST with 2% BSA/2% NGS and left on the sections overnight at 4 °C, in the dark. Polyclonal rabbit anti-MS, the gift form Dr. A. Urbański, were previously used in 62,65 . Incubating tissues without peptide antiserum was performed as a control. After five washes in PBST (10 min each) and one overnight wash at 4 °C, the secondary antibody (1:100, diluted in TBST; Anti-Rabbit IgG (whole molecule)-Alkaline Phosphatase antibody produced in goat, Sigma Aldrich) was applied for 1 h at room temperature. After five 10 minutes washes in PBST on the shaker, the slides were washed three times 10 minutes with processing buffer (100 mM Tris-NaCl, 50 mM MgCl2, pH 9.5) and then incubated in NBT-BCIP solution (Sigman Aldrich). As soon as staining became visible, the reaction was stopped by incubation in distilled water. Sections were mounted using ImmunoHistomount (Sigma Aldrich) and photographed using Zeiss Axicam 105 color (Zeiss) camera mounted with to Zeiss Axio Scope.A1 (Zeiss) microscope.
www.nature.com/scientificreports www.nature.com/scientificreports/ Production of dsRNA. LsalMS gene sequence was too short to test two independent dsRNA fragments. Therefore, only one dsRNA fragment was tested in this study, complementary to the gene sequence that shows lowest similarity to other salmon louse sequences. To test it, selected sequence was blasted (BLASTN) against sea lice genomic sequence and collection of predicted transcripts (sealouse.imr.no). Templates for dsRNA were generated in two steps. First, PCR products were obtained from copepodids cDNA for LsalMS and plasmid containing cod CPY185 gene for control fragment with the specific pairs of primers (Supp . Table 1). Later, the final templates were generated with the same primers but with T7 extensions added (Supp. RNA interference. RNAi on nauplius I larvae was performed as described earlier 84 . The experiment was executed using groups consisting of approximately 2000 nauplius I larvae. For soaking, these were divided in subsamples of 100 animals, each incubated in 200 μl of seawater in the lid of a 1.5 ml microcentrifuge tube. dsRNA was added to the final concentration of 20 ng/μl. The existence of any unspecific effects of soaking nauplii in dsRNA on gene expression was assessed by soaking animals in the control dsRNA, containing a fragment with no significant sequence similarity to the salmon louse genomic sequence (a cod trypsin gene, CPY185) 85 . Animals were incubated overnight. When moulting was finished, all animals within each treatment group were pooled and transferred to flow through incubators, where they were kept until sampling on day 7 post hatching. Before sampling, lice were inspected under a binocular microscope (Olympus SZX12, 0.5× Olympus objective). Photographs of copepodids were taken using a Zeiss Axicam 105 color (Zeiss) camera mounted to Zeiss Axio Scope.A1 (Zeiss) microscope. For fish infection, 12 groups of 100 copepodids from each treatment group were separated. For total RNA extraction copepodids were divided into five parallels with around 100 animals per sample and preserved in RNAlater (Invitrogen). For histological analysis, copepodids were collected in Karnovsky fixative.
RNAi on preadult I females and males was performed as previously described 85 . dsRNA fragments were diluted to a concentration of 600 ng/μl and bromophenol blue was added to visualise dsRNA solution during injection. On the day of injection, preadults I of both sexes were collected from the fish and injected with 0,5 μl of dsRNA solution in cephalothorax, with care not to disturb the intestine. After injection, the animals were incubated in seawater for around 4 h and placed back on fish. For each fragment, in total eight fish were used, each carrying 20 injected preadult I females and 20 injected preadult I males. A cod trypsin dsRNA CPY185 fragment was used as a control. The experiment was terminated after 3 weeks, when all dsLsalMS treated female were lost from the fish. To evaluate mating potential of adult males, a total of 30 pre-adult I male lice were injected with dsLsalMS, and placed together with non-injected females (10 males and 10 females per fish, fish n = 3). The same experimental set up was repeated for males injected with control dsRNA. The experiment was terminated after 6 weeks, when all control female lice had developed at least their first pair of egg strings. At termination, all recovered lice were subjected to a gross morphological inspection before lice were sampled for RNA extraction in RNAlater or histological analysis in Karnovsky fixative. Photographs were taken with Canon EOS 600D camera mounted with an adaptor (LMscope) to an Olympus SZX9 dissecting microscope.
For evaluation of the intestine contraction rate, preadult II females were placed on petri dish in the drop of seawater and intestinal movement was recorder for 1 minute at a time, using Canon EOS 600D camera mounted with an adaptor (LMscope) to an Olympus SZX9 dissecting microscope. Each louse was recorded twice. Number of contractions was counted on each movie. (Samples of the movies in Supp. Video 1).
Infection of host fish. Atlantic salmon (Salmo salar) were placed in single fish tanks (70 litres) with constant flow of seawater 76 . Seawater temperature was maintained at 12 °C during the whole experiment. 12 fish were infected with copepodids treated with dsLsalMS and 12 with control copepodids treated with dsCPY185. During infection period, water level was decreased to around 15 cm, to enhance contact between fish and copepodids. Fish were washed by 100 copepodids (7-9 days post hatching) and left without water flow for 10 minutes, with water manually mixed after 5 minutes. After that time, the water flow was restarted. Lice were collected from four fish from each group after 8, 12 and 18 days post infection, when control lice were in chalimus I, chalimus II and preadult stages, respectively. At termination, all recovered lice were subjected to a gross morphological inspection, counted and staged. Photographs were taken with Canon EOS 600D camera mounted with an adaptor (LMscope) to an Olympus SZX9 dissecting microscope. Lice were sampled for RNA extraction in RNAlater or histological analysis in Karnovsky fixative.
RNA extraction and cDNA synthesis. Total RNA was extracted as described earlier 86 . Total RNA was diluted in DEPC-treated water (Invitrogen) and the final concentration was evaluated with spectrometry (NanoDrop Technologies Inc.). From each sample 2000 ng of total RNA was DNAse treated in 10 μl reaction (TURBO TM DNase, Ambion). 3 μl of each DNAse treated total RNA was used in cDNA synthesis with oligo(dT) and random primers, using program recommended by the manufacturer: primers annealing: 5 min −25 °C, synthesis: 45 min −42 °C, termination: 5 min −95 °C (AffinityScript qPCR cDNA Synthesis Kit, Stratagene). cDNA was diluted four times before use in qRT-PCR or storage at −20 °C. 2 μl of final dilution was used in each PCR reaction.
For evaluation of LsalMS transcript level in chalimus I and chalimus II Direct zol RNA Microprep Plus (ZYMO Research) kit was used, according to the manufacturer instructions. TotalRNA from single preadult I lice was extracted using Tri Reagent (Sigman Aldrich) as described before 86 . www.nature.com/scientificreports www.nature.com/scientificreports/ Quantitative RT-PCR (qRT-PCR). qRT-PCR was performed and analysed using the elongation factor 1 alpha (EF1α) 87 and filamin A 88 as reference genes. Both genes are considered to be good references, with stable expression rate in all developmental stages and under RNAi treatment conditions 86,88 . All primers and probes are listed in Supp. Table 1. All qRT-PCRs were carried out in technical duplicates. The assays were performed simultaneously for the reference and test genes using the same cDNA and master mix (PowerUp ® SYPR ® Green Master Mix; Applied Biosystems). Thermal cycling and quantification were done on the QuantiStudio3 qPCR Machine (Applied Biosystems) in 10 μl reactions under standard conditions (initiation: 2 min −50 °C, holding: 2 min −95 °C, 40 cycles of 15 s −95 °C, 1 min −60 °C, followed by the Melt Curve Stage). Amplification efficiencies of target and reference genes were evaluated (ef1α -93,8%, filamin A -105,7%, LsalMS -104,6%). Relative expression was calculated using 2 −ΔΔCT formula, with amplification efficiency correction for each gene (Pfaffl Method), utilizing the controls of each experiment as a standard. Independent-Samples T-tests were used to determine if control and test groups were differently expressed. A p-value of 0.05 was chosen as threshold.
Histology. Animals were fixed in Karnovsky fixative, washed twice in PBS, dehydrated in a graded ethanol series, pre-infiltrated with Technovit7100/ethanol (50/50) for 4 hours (Technovit 7100, Heraeus Kulzer Technique) and infiltrated with Technovit7100/hardener overnight before embedding in plastic. Two μm thick sections were produced using a microtome (Leica RM 2165) and stained with toluidine blue (1% toluidine blue in 2% borax) for 30 s. The stained sections were mounted using DPX New Mounting Medium (Merck). Photographs were taken using a Zeiss Axicam 105 color (Zeiss) camera mounted to Zeiss Axio Scope.A1 (Zeiss) microscope.