Background

Human schistosomiasis, which is caused by parasitic Schistosoma flatworms, remains one of the most important tropical diseases in terms of public health impact despite the continued deployment of control measures [1,2,3]. Globally, over 200 million people are infected with schistosomes across 78 countries, and approximately 700 million people are at risk of infection [4,5,6]. Three main Schistosoma species are responsible for endemic disease, Schistosoma mansoni, Schistosoma japonicum and Schistosoma haematobium.

Schistosomes exhibit gonochorism and have a complex life cycle that involves passage through a molluscan intermediate host and a mammalian definitive host [2, 7]. When voided in the urine or faeces, and upon contact with freshwater, schistosome eggs hatch, releasing miracidia that swim using cilia to locate a compatible snail host, which they then penetrate. Next, each miracidium transforms into a mother sporocyst and undergoes asexual reproduction, producing daughter sporocysts that have the capacity to generate large numbers of cercariae for release [8,9,10]. These non-feeding human-infective cercariae swim using their bifurcated tail to locate the definitive host; they then penetrate the skin with the help of proteolytic enzymes secreted from their acetabular glands [11,12,13]. During penetration, the cercaria loses its tail and the head transforms into a schistosomulum (aka somule). The somules transit the skin, enter the vasculature and migrate via the lungs to the hepatic portal system where they develop into sexually mature dioecious adult worms [2, 14]. The ability of the cercaria to successfully transform into a somule is critical to its establishment as a human parasite [2]. The passage through, and transfer between, two different hosts exposes the schistosome to substantial changes in local environment to which the parasite must adapt to survive, grow and develop [2, 15].

Heat shock proteins (HSPs) are evolutionarily conserved proteins that are expressed in cells constitutively and can also be induced by stress [16, 17]. Generally, HSPs can be broadly classified into two families, the small ATP-independent HSPs of molecular mass 8–28 kDa, and the larger ATP-dependent HSPs of molecular mass 40–105 kDa [17]. Initially discovered as proteins upregulated in heat-stressed Drosophila melanogaster [18], HSPs are now understood to perform vital functions that regulate cellular homeostasis both in stressed and unstressed scenarios. HSPs are involved in multiple cellular processes and function mainly as molecular chaperones, which facilitate native protein stabilization, refolding, translocation, and degradation [19, 20]. Considering that the cercaria-to-somule transformation provides a unique physiological stress involving increases in temperature and salinity as the schistosome moves from freshwater to a warm-blooded environment and loses its tail, it is plausible that HSPs play a vital role in ensuring the survival of the parasite during this transition. Moreover, HSPs are likely to be essential to the continued survival of the schistosome in the hostile environment of the host, where they must fend off immune attack. Upregulation of HSP expression during the earliest stages of intra-mammalian somule development has been observed at the schistosome tegument through proteomics [21]. Furthermore, heat shock factor 1 (a major transcriptional activator responsible for transcribing heat shock genes) was localised to the acetabular glands of S. mansoni cercariae, suggesting a potential role for HSPs in cercarial invasion and transformation [22]. However, of the few investigations on schistosome HSPs, most have studied an individual HSP family member.

In the current study, a comprehensive and comparative molecular bioinformatic analysis of all S. mansoni HSPs was carried out, and HSP family members were mapped within human infective S. mansoni life stages, providing an atlas of HSP expression within the worm. The data provide novel insights into the complexities of HSPs in schistosomes, the factors that govern their regulation, and their potential role in schistosome function.

Results and discussion

Comparative analysis of S. mansoni HSPs

A total of 69 human HSPs were mined from databases of the National Center for Biotechnology Information (NCBI) [23] across the five HSP families (HSP 10, HSP 40, HSP 60, HSP 70 and HSP 90). The J domain-containing proteins found on the NCBI and InterPro databases, which are also known as DNAJ(HSP40)C proteins, are listed here as DNAJC23-30 (as in [23]) (Additional file 1: Dataset S1). The human HSP amino acid sequences (from UniProt) were BLASTed against S. mansoni proteins on WormBase ParaSite [24]. After removal of duplicate and partial sequences, and after ensuring that all S. mansoni HSPs contained the necessary functional domains (using InterPro), 55 S. mansoni HSPs were identified across the five HSP families (Table 1; Additional file 1: Dataset S1). As in humans, the S. mansoni HSP 10 and HSP 60 families each contained one member, Smp_097380.1 and Smp_008545.1, respectively. However, 12 HSP 70 family members were identified, one less than in humans, seven of which are almost identical copies replicated in two different parts of the genome. Furthermore, in contrast to humans that have 49 HSP 40 and five HSP 90 family members, S. mansoni was found to possesses 38 HSP 40s and three HSP 90s (Table 1; Additional file 1: Dataset S1).

Table 1 Heat shock proteins (HSPs) identified in Schistosoma mansoni and the corresponding Schistosoma japonicum and Schistosoma haematobium HSPs

Based upon currently available genomic data, the composition of the Schistosoma HSP families appears broadly similar across the Schistosoma species investigated here, with most variation in protein number seen in HSP 40 and HSP 70 families (Fig. 1). Schistosoma bovis contains the most HSP 40s and HSP 70s (40 and 17 members, respectively), whereas Schistosoma rodhani possesses the fewest HSP 40s (30 members), and Schistosoma curassoni the fewest HSP 70s (five members). Interestingly, the free-living flatworms Macrostomum lignano (a marine basal flatworm) and Schmidtea mediterranea (a freshwater planarian) have considerably more HSPs than schistosomes (Fig. 1), with 163 and 75 HSPs, respectively; the striking expansion of HSPs in M. lignano is mainly due to a very large 123-member HSP 40 family. The significance of this expansion of HSPs in the free-living flatworms, when compared to schistosomes, is unknown, but could be linked to their free-living rather than parasitic habit. In planarians, DNAJA1, a HSP 40 family member, is enriched in neoblasts [25] and is required for stem cell maintenance, regeneration and homeostasis [26]; the function of the equivalent protein in schistosomes is worthy of investigation, particularly given the importance of stem cells to schistosome survival in the host [27].

Fig. 1
figure 1

Comparative analysis of the total number of heat shock proteins (HSPs) identified in humans, free-living flatworms and Schistosoma species. HSPs were identified through bioinformatic similarity searches for each HSP family (HSP 10, HSP 40, HSP 60, HSP 70 and HSP 90)

To determine the similarity of individual HSPs between the three most common human-infective Schistosoma spp., S. mansoni HSP amino acid sequences were compared against those of S. japonicum and S. haematobium. While HSP 60 was found in all three species, HSP 10 was surprisingly absent from both S. japonicum and S. haematobium (Table 1). Most of the S. mansoni HSP 40 family members showed greater similarity to S. haematobium (~ 87%) compared to S. japonicum (~ 77%) (Table 1). The S. mansoni HSP 70 and HSP 90 family members were also more similar to those of S. haematobium compared to S. japonicum, with the exception of Smp_072320 (chain A heat shock cognate 71 kDa protein-like). The high similarity between the HSP members across these three Schistosoma spp. indicates that these HSPs likely play a vital role in the parasite and might be essential for its survival. That S. mansoni HSPs show more similarity with S. haematobium HSPs than those of S. japonicum agrees with data published by Young et al. [Full size image

Gene ontology-based functional enrichment analysis revealed that the HSP family networks were mainly enriched for proteins involved in “protein processing in the endoplasmic reticulum” Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway, except for HSP 10 and HSP 60 networks, which were dominated by proteins involved in “ribosome” and “metabolic pathways”, respectively (Fig. 5). As the HSP 60/HSP 10 chaperonin complex functions in the mitochondrial matrix [47], it is not surprising that most of their interacting proteins are mitochondrial matrix proteins. Proteins involved in “protein export” were enriched in the HSP 40 and HSP 70 networks; these HSPs are known to intimately engage in protein folding and trafficking mechanisms [48]. In Plasmodium falciparum, HSP 40 members are among the exportome proteins involved in the correct presentation of P. falciparum erythrocyte membrane protein 1 on the host cell surface [49,50,51]. The HSP families were also dominated by other interesting KEGG pathways important for schistosome biology [e.g. “RNA degradation”, “autophagy—animal”, “citric cycle (TCA) cycle”, “carbon metabolism”, “mitophagy—animal”, “endocytosis”, “phagosomes”, “glycolysis or gluconeogenesis”, “spliceosome”, “peroxisome”, “oxidative phosphorylation”, and “biosynthesis of amino acids”]. The HSP 90 network was the only network enriched for proteins involved in “FoxO signalling” (5/96) and “mTOR signalling” (4/96) pathways. Collectively, these data provide a comprehensive framework for further understanding the biological and functional actions of these HSPs in S. mansoni.

Phosphoproteome analysis of S. mansoni HSPs

Protein phosphorylation is an important post-translational modification, involving kinase-mediated addition of phosphate to serine, threonine, or tyrosine residues in eukaryotes, with removal enabled by protein phosphatases; such “switching” of phosphorylation status governs protein–protein interactions and activation of cellular signalling pathways [52, 53]. Given the importance of HSPs to cellular regulatory processes, experimentally discovered phosphorylation sites were mined from the recently published S. mansoni phosphoproteome [54] and were annotated. At least one bone fide phosphorylation site was found in 17 of the 55 HSPs, with 76 phosphorylation sites detected in total (Fig. 6; Additional file 4: Figure S1; Additional file 5: Dataset S4), including 20 sites within HSP functional domains. Over 80% (67/76) of the identified S. mansoni HSP phosphorylation sites were also successfully predicted by the phosphorylation site prediction tools, the Phosphorylation Site Database (PHOSIDA) and the Human Protein Reference Database (HPRD). The phosphorylation sites amongst the HSPs were 51% phosphoserine, 28% phosphothreonine and 21% phosphotyrosine. HSP 40, HSP 70 and HSP 90 family members were mainly phosphorylated on serine residues, whereas 60% of HSP 60 phosphorylation sites were on threonine (Fig. 6). No phosphorylation sites were found or predicted for HSP 10.

Fig. 6
figure 6

Phosphorylation sites within Schistosoma mansoni HSPs. Data on the phosphorylation sites of S. mansoni HSPs were mined from the S. mansoni phosphoproteome published by Hirst et al. [54]. Four HSPs are shown; the remainder can be viewed in Additional file 4: Figure S1. Domains within each HSP were identified using the conserved protein domain tool within InterPro and the National Center for Biotechnology Information (NCBI) Conserved Domains Database. Putative upstream kinases, phosphatases and binding motifs were identified using the Human Protein Reference Database (HPRD) motif finder and the Phosphorylation Site Database (PHOSIDA). Kinases are written in black and are denoted by a plus or a cross symbol, phosphatases are in red and are denoted by a minus symbol, binding motifs are in blue and are denoted by an asterisk

Protein kinase substrates contain short sequence motifs surrounding the phosphorylated residue which facilitate their recognition by the protein kinase. Therefore, using the PhosphoMotif discovery tools, HPRD and PHOSIDA, motifs surrounding the phosphorylated residue on each of the S. mansoni HSPs could be matched to known kinases. These included PKC, PKA, Akt, ERK1/2, CK1/2 GSK-3, Chk1, CAMK I, II, IV and GPCR, with the CK1/2 (e.g. SPxxS, TxxS) motif being the most represented in the dataset (Fig. 6; Additional file 4: Figure S1; Additional file 5: Dataset S4). TC-PTP and SHP1 phosphatase substrate motifs were also identified on Smp_035200.1 (HSP 40), Smp_106130.1 (HSP 70) and Smp_072330.1 (HSP 90). HPRD also identified consensus sites for protein binding, including 14-3-3, BARDI BRCT, Plk1 PBD, WW, FRIP PTB and MDC1 BRCT domain binding motifs, which function as modular protein domains that mediate protein–protein interaction between the phosphorylated protein and client protein.

Evolutionary conserved phosphorylation sites are believed to be of core functional relevance among certain species [55]. Therefore, S. mansoni HSP phosphorylation sites were compared with those of humans through manual alignment of each S. mansoni phosphorylated HSP with its human ortholog (Additional file 6: Figure S2). Thirty phosphorylation sites (out of the 76 identified) were conserved with human HSPs (Table 2), with Smp_072330 (HSP 90) containing the most (nine sites). Insight into the function(s) of such phosphorylation sites in the human HSPs could provide a comprehensive framework for understanding their role in S. mansoni HSPs. For example, CK2-mediated phosphorylation of human HSP 90α Ser231/HSP 90β Ser226, equivalent to S. mansoni HSP 90α isoform 2-like (Smp_072330) Ser225, caused dissociation of the aryl hydrocarbon receptor (AhR; a transcription factor associated with cellular response to environmental stimuli such as xenobiotics)—HSP 90 complex and destabilisation of AhR protein in humans [56, 57]. Mutation of this site to a non-phosphorylatable alanine increased the transcriptional activity of AhR and stabilized its interaction with HSP 90 [56]. Also, phosphorylation of human HSPA1A (HSP 70) at Tyr41, similar to S. mansoni heat shock cognate 71 kDa protein isoform 1-like (Smp_106930) Tyr39, regulates HSP 70 protein stability, and inhibition of phosphorylation of HSP 70 at this site with erlotinib (a tyrosine kinase inhibitor) results in increased degradation of HSP 70 [58]. Further understanding of the role of conserved HSP family member phosphorylation sites in humans might provide valuable insight into schistosome HSP function and reveal novel strategies for schistosome control.

Table 2 Phosphorylation sites identified in Schistosoma mansoni HSPs that are conserved with phosphorylation sites present in the human HSP orthologs

Map** the expression of HSPs in S. mansoni

Because HSPs are evolutionarily conserved, commercially available antibodies against HSPs (mainly targeting Homo sapiens sequences) were selected based on conservation in the antibody binding region (where known) with the S. mansoni protein; antibodies recognising regions of high homology were selected (Additional file 7: Figure S3). In total, 13 antibodies were tested by western blotting of S. mansoni protein extracts, and five were found suitable (Fig. 7). A single immunoreactive band was detected with HSP 10 (~ 10 kDa), HSP 60 (~ 60 kDa), HSP 70 (~ 70 kDa) and HSP 90 (~ 90 kDa) antibodies in 24-h somules, adult male and female worms (Fig. 7), at the expected size. However, when testing anti-HSP 40 antibodies raised against human DNAJB1, which has closest homology to S. mansoni Smp_104730, a band at ~ 40 kDa was detected in the adult worms and in the 24-h in vitro cultured somules; a stronger ~ 70-kDa band was also detected, but only in the somules.

Fig. 7
figure 7

Detection of HSPs in Schistosoma mansoni. In vitro 24-h-cultured somules (~ 1000 per lane) or adult worms (one male or female per lane) were homogenised and protein extracts subjected to western blotting with one of the following antibodies: anti-HSPE1 (ARPP54651_P050) (HSP 10) raised against a synthetic peptide located within amino acid 55–102 of human HSPE1; anti-HSP 40 (SPC-100), which recognises the whole amino acid sequence of human DNAJB1; anti-HSP 60 (SMC-111A/B) raised against amino acid sequence 382–447 of human HSP 60; anti-HSP 70 (SMC-162 C/D) raised against the whole amino acid sequence of human heat shock 70 kDa protein 1A and heat shock 70 kDa protein 1B; or anti-HSP 90 (STJ93608) that recognises amino acids 180–260 of human HSP 90β

The in situ distribution of HSPs in cercariae, 3-h and 24-h in vitro cultured somules, and adult male and female S. mansoni were next determined using immunofluorescence and confocal laser scanning microscopy. In all cases, negative controls of the different life stages showed minimal background staining (Additional file 8: Figure S4). Labelling of intact cercariae, 3-h and 24-h somules, male and female adult worms with anti-HSPE1 (HSP 10) antibodies and analysis of image projections and/or individual confocal z-sections revealed prominent expression of HSP 10, an ATP-independent mitochondrial resident protein, in the oesophagus, cephalic ganglia, sub-tegument and parenchyma tissue of cercariae (Fig. 8a). Similarly, HSP 10 localised to the sub-tegument and cephalic ganglia of 3-h and 24-h in vitro cultured somules (Fig. 8b, c). In addition, HSP 10 was observed in the acetabulum and tegument of 3-h somules (Fig. 8b) and the gland duct of 24-h somules (Fig. 8c). In adults, HSP 10 was clearly evident in the testes and surface tubercles of males (Fig. 8d, e), and in the ovary and some vitelline cells of the female (Fig. 8f, g). For HSP 60, another mitochondrial resident protein and a co-chaperone of HSP 10, immunoreactivity was observed in the oesophagus, cephalic ganglia, and spines of cercariae (Fig. 9a, b) and in the acetabulum, cephalic ganglia, gland duct and tegument of 3-h (Fig. 9c) and 24-h (Fig. 9d) somules. In adults, HSP 60 was most noticeable in the testes and tubercles of males (Fig. 9e, f) and tegument/sub-tegument of females (Fig. 9g). HSP 60 was also present in the parenchymal tissue of the adult worms.

Fig. 8
figure 8

ag Map** HSP 10 in Schistosoma mansoni. Cercariae, somules, male and female adult worms were fixed and processed for immunofluorescence using anti-HSPE1 (HSP 10) primary and Alexa Fluor 488 secondary (green) antibodies; rhodamine phalloidin (red) was used to stain filamentous actin. Samples were mounted on slides and images captured on a Zeiss LSM 800 confocal laser scanning microscope. HSP 10 localised in a cercariae; b 3-h somules; c 24-h somules; d, e adult males; and f, g adult females. Representative micrographs are single z-sections through the parasite. Scale bars = 25 µm (for cercariae and somules) and 50 µm (for adult worms)

Fig. 9
figure 9

ag Map** HSP 60 in Schistosoma mansoni. Cercariae, somules, male and female adult worms were fixed and processed for immunofluorescence using anti-HSP 60 primary and Alexa Fluor 488 secondary antibody (green) antibodies; rhodamine phalloidin (red) was used to stain filamentous actin. Samples were mounted on slides and images captured on a Zeiss LSM 800 confocal microscope. HSP 60 localised in a, b cercariae; c 3-h somules; d 24-h somules; e, f adult males; and g adult females. Representative micrographs are single z-sections through the parasite. Scale bars = 25 µm (for cercariae and somules) and 50 µm (for adult worms)

Labelling with anti-HSP 40 antibodies revealed the presence of HSP 40, a co-chaperone of HSP 70, in the cercarial gland duct, the acetabulum, cephalic ganglia (Fig. 10a) and cercarial tail tissue (Fig. 10b). In 3-h and 24-h somules, HSP 40 localised predominantly to the tegument, sub-tegument, acetabulum, and cephalic ganglia (Fig. 10c, d). HSP 40 was mainly found in the tubercles of males (Fig. 10g) and tegument of both male and female worms (Figs. 10f and h); however, it was absent from the testes of the male worms (Fig. 10e). Parenchyma tissue was also stained. Cercariae, 3-h and 24-h somules, male and female worms labelled with anti-HSP 70 antibodies displayed similar localisation patterns (Fig. 11) to that seen for its co-chaperone HSP 40. However, in addition, striking immunoreactivity was observed at the head–tail junction of cercariae (Fig. 11a) and 24-h somules (Fig. 11c), and the oral tip of the 24-h somule (Fig. 11c).

Fig. 10
figure 10

ah Map** HSP 40 in Schistosoma mansoni. Cercariae, somules, male and female adult worms were fixed and processed for immunofluorescence using anti-HSP 40 primary and Alexa Fluor 488 secondary antibodies (green); rhodamine phalloidin (red) was used to stain filamentous actin. Samples were mounted on slides and images captured on a Zeiss LSM 800 confocal microscope. HSP 40 localised in a, b cercariae; c 3-h somules; d 24-h somules; eg adult males; and h adult females. Representative micrographs are single z-sections through the parasite. Scale bars = 25 µm (for cercariae and somules) and 50 µm (for adult worms)

Fig. 11
figure 11

ag Map** HSP 70 in Schistosoma mansoni. Cercariae, somules, and male and female adult worms were fixed and processed for immunofluorescence using anti-HSP 70 primary and Alexa Fluor 488 secondary antibodies (green); rhodamine phalloidin (red) was used for filamentous actin staining. Samples were mounted on slides and images captured using a Zeiss LSM 800 confocal microscope. HSP 70 localised in a cercariae; b 3-h somules; c 24-h somules; d, e adult males; and f adult females. Representative images are eg maximum projections of confocal z-stacks; the rest are single z-sections through the parasite. Scale bars = 25 µm (for cercariae and somules) and 50 µm (for adult worms)

Labelling with anti-HSP90β antibodies revealed expression of HSP 90 in the cephalic ganglia and gland duct of cercariae (Fig. 12a) and 3-h somule (Fig. 12b). In addition, HSP 90 was observed in the sub-tegument, head/tail junction and tegument of 3- and 24-h somules (Fig. 12b-c). HSP 90 was also evident in the testes and tubercles of male worms (Fig. 12d–f) and the ovary and sub-tegument of the females (Fig. 12g, h). Antibody staining was broadly consistent between HSPs that are known to act as co-chaperones, supporting the idea that the anti-HSP antibodies were reacting with their intended targets in intact worms. All antibodies also reacted with proteins of the expected molecular weights (Fig. 7). However, the possibility that the antibodies might interact with other S. mansoni proteins cannot be ruled out, especially with anti-HSP 40 antibodies, where an immunoreactive band was observed at ~ 70 kDa in 24-h in vitro transformed somules only. Furthermore, unlike HSP 10 and HSP 60 that have only one family member, several members make up the HSP 40, 70 and 90 families; we are unable therefore to identify the specific HSP member(s) of these families that have been mapped within the parasite stages. Notwithstanding these caveats, the immunofluorescence profiling of multiple HSPs across several human-infective schistosome life stages provides a detailed and novel insight into the potential roles of these proteins in coordinating schistosome function.

Fig. 12
figure 12

ah Map** of HSP 90 in Schistosoma mansoni. Cercariae, somules, male and female adult worms were fixed and processed for immunohistochemistry using anti-HSP 90 primary and Alexa Fluor 488 secondary antibodies (green); rhodamine phalloidin (red) was used for filamentous actin staining. Samples were mounted on slides and images captured on a Zeiss LSM 800 confocal microscope. HSP 90 localised in the a cercariae; b 3-h somules; c 24-h somules; df adult males; and g, h adult females. Representative micrographs are single z-sections through the parasite. Scale bars = 25 µm (for cercariae and somules) and 50 µm (for adult worms)

Members of the HSP 90 family interact with steroid hormone receptors and are essential for fertility [59]. HSP 90α has been implicated to play a role in female mouse oocyte meiosis [60]; in males, the absence of HSP 90 caused the disruption of testicular development and azoospermia [61]. Similar to the data published by Xu et al. [62] on S. japonicum, HSP 90 was prominent in the ovary and testes of adult S. mansoni worms, implying a potential role in the regulation of adult fertility and differentiation of the germline stem cell population in schistosomes. Interestingly, HSP 60, unlike its co-chaperone HSP 10, was also present in the testes of the male worm, but was less prominent in the female ovary, suggesting a potential role in spermatogenesis but not in oogenesis.

The tegument proteins of schistosomes are known to play important roles in a variety of cellular processes, including nutrition, excretion, osmoregulation, and signal transduction, and key roles in host-parasite interactions, including immune evasion and modulation [2, 63]. From the data presented here, all HSPs investigated were present in the tegument of 3- and 24-h somules. This agrees with a previously published proteomic study where five HSPs (Smp_106930, Smp_148530, Smp_072330, Smp_008545 and Smp_049550) were detected in the somule tegument [21]. In addition, HSP 10, 40, 60, 70, and 90-like proteins have been identified by proteomics in tegument surface membranes of adult S. mansoni after differential extraction [64]. Collectively, these data demonstrate that HSPs likely play a role in assisting the parasite to adapt to the host immune microenvironment, supporting its transition from an immune-sensitive to an immune refractory state.