Label-free quantitative proteomic analysis of molting-related proteins of Trichinella spiralis intestinal infective larvae

Molting is a key step for body-size expansion and environmental adaptation of parasitic nematodes, and it is extremely important for Trichinella spiralis growth and development, but the molting mechanism is not fully understood. In this work, label-free LC–MS/MS was used to determine the proteome differences between T. spiralis muscle larvae (ML) at the encapsulated stage and intestinal infective larvae (IIL) at the molting stage. The results showed that a total of 2885 T. spiralis proteins were identified, 323 of which were differentially expressed. These proteins were involved in cuticle structural elements, regulation of cuticle synthesis, remodeling and degradation, and hormonal regulation of molting. These differential proteins were also involved in diverse intracellular pathways, such as fatty acid biosynthesis, arachidonic acid metabolism, and mucin type O-glycan biosynthesis. qPCR results showed that five T. spiralis genes (cuticle collagen 14, putative DOMON domain-containing protein, glutamine synthetase, cathepsin F and NADP-dependent isocitrate dehydrogenase) had significantly higher transcriptional levels in 10 h IIL than ML (P < 0.05), which were similar to their protein expression levels, suggesting that they might be T. spiralis molting-related genes. Identification and characterization of T. spiralis molting-related proteins will be helpful for developing vaccines and new drugs against the early enteral stage of T. spiralis.


Introduction
Trichinella is an obligate parasitic nematode of animals worldwide. Human T. spiralis infection primarily results from eating raw or semiraw animal meat infected with Trichinella spiralis muscle larvae (ML). Trichinellosis outbreaks were reported in 55 countries around the world, resulting in 65 818 cases and 42 deaths, from 1986 to 2009 [1]. In Chinese Mainland, 14 of 15 trichinellosis outbreaks were due to infected pork during 2004-2009, and pork is the predominant infection source [2,3]. Trichinella is not only a public health harm but also a serious hazardous to animal food safety. Vaccine development is needed to eliminate the transmission of Trichinella among domestic animals [4][5][6].
After being eaten, ML encapsulated in animal meat emerges in the host's stomach and develops into its first stage, called intestinal infective larvae (IIL1), at 0.9 hours post-infection (hpi). The IIL1 penetrate the intestinal epithelium and undergo the first molting to develop into IIL2 at 10 hpi. They subsequently molt three times into IIL2-IIL4 to develop into the adult worm (AW) within 10-30 hpi [7][8][9]. A cuticle covers the surface of the Trichinella worm body, and molting is the basis of Trichinella growth and development, which occurs in the development of all members of the nematode. The presence of various enzymes in cuticles of T. spiralis IIL suggests that the cuticles may contain the molting-related enzymes [10]. Each molting consists of the new cuticle formatting, old cuticle loosening and rupturing, and larval escape from previous-stage larval cuticle [11]. Molting is controlled by ex-sheathing fluid secreted by the larvae; however, which kinds of proteins in ex-sheathing fluid participate in the shedding of old cuticle and the formation of neocuticle are not yet understood. Hence, agents that inhibit or block molting have been a focus for developing new drugs or vaccines against nematode infection. In addition, studies on molting will be valuable to understand the growth and developmental processes of nematodes [12].
Molting is an important strategy for T. spiralis to adapt to the environmental intestinal milieu. Previous studies revealed that T. spiralis infective larvae molted and developed to the adult stage after they were inoculated onto intestinal epithelial cell monolayers cultured in vitro. Approximately 50% of IIL larvae molted when they were cultivated in an epithelial cell monolayer for 5-9 days [13]. If the old cuticle was not completely shed, the larvae were still enclosed in the sheath and could not develop into the adult stages [14]. However, the biological processes that regulate T. spiralis molting are not fully clear.
The encapsulated-stage ML lodge in the nurse cells of the host's skeletal muscle, and their surrounding environment is relatively stable. The ML may live from 1 to 2 years to 10-15 years without any major harm [15]. Hence, ML has been used as a normal control for T. spiralis protein expression. Following ingestion, T. spiralis ML are liberated from their capsules in the host's stomach and activated into IIL1 by exposure to enteral content or bile at 0.9 hpi. After they invade the host's intestinal epithelium, the IIL completes four molts from IIL1 to IIL4 at approximately 10, 10-14, 15-22 and 23-30 hpi, respectively [16]. The 10 h IIL might express multiple molting-related proteins that are involved in new cuticle generation, old cuticle loosening and ex-sheathing for the first molting of this parasitic nematode. Furthermore, since the enteral-stage IIL and AW parasitize the complicated intestinal environment, the IIL at 10-30 hpi and AW possibly express various proteins that participate in other processes of Trichinella development, such as digestion, immune escape, detoxification, reproduction, and so on [6,[17][18][19]. To screen and identify the T. spiralis molting-related proteins as far as possible, 10 h IIL were selected for quantitative proteomic analysis. In this study, the proteomic profile variation between T. spiralis ML and 10 h IIL was determined using label-free quantitative proteomic technology with bioinformatics, which is expected to provide new insight to guide the study of T. spiralis larval molting, growth and development [20].

Worm and mice
Trichinella spiralis strain (ISS534) from a domestic pig in central China was kept by mouse serial passaging in our department [21]. Specific-pathogen-free (SPF) 8-weekold female mice were obtained from the Experimental Animal Center of our university. The animal experiment protocol was approved by the Life Science Ethics Committee of Zhengzhou University (No. SCXK 2017-0001).

Intestinal infective larvae at different times post-infection
The MLs were obtained from artificially digesting T. spiralis-infected mouse muscles with 1% pepsin [22,23]. Each mouse was orally inoculated with 5000 ML, and IIL were isolated from the small intestine at 8, 10, 12, 14, 16, and 18 hpi [24]. The IIL molting process was observed and counted under microscopy. Fifty IIL were randomly selected from each time point, and three replicates were observed. All larvae were washed thoroughly with sterile PBS and then stored in liquid nitrogen.

Protein extraction
The ML and 10 h IIL were homogenized in a tissue grinder. The larval fragment was further homogenized by ultrasonication as reported [25]. The homogenate was boiled for 15 min. Following centrifugation at 14 000 g for 40 min, the supernatant was filtered using 0.22 µm filters. The protein concentration was determined using a BCA kit (Bio-Rad, USA).

Protein digestion
Approximately 200 μg of larval proteins was added to 30 μL of SDT buffer (150 mM Tris-HCl, 4% SDS and 100 mM DTT) and filtered with UA buffer (8 M urea, 150 mM Tris-HCl). To inhibit the cysteine residues, larval proteins were incubated with 100 μL iodoacetamide for 30 min in darkness. After washing, the protein sample was digested with 4 μg trypsin at 37 °C overnight, and the resulting peptides were recovered, desalted and concentrated as described [26]. The contents of peptides were determined using a UV light spectral density at 280 nm.

LC-MS/MS and protein datum analysis
LC-MS/MS was carried out using a Q Exactive mass spectrometer (Thermo Fisher Scientific, USA) as reported [27]. The MS data were analyzed by MaxQuant software (version 1.5.3.17) against the Trichinella data from the UniProt database. The differentially expressed T. spiralis proteins between the ML and 10 h IIL stages were those with a 2.0-fold change compared with one another [12]. Molecular pathway analyses through Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) were conducted to analyze the possible enrichment of differentially expressed proteins with specific biological features. The mapping of protein interactions was performed by statistical analysis at a low confidence score in the STRING database [27].

qPCR
Eight genes from differently expressed T. spiralis proteins were selected to investigate their transcription levels using qPCR [9]. Total RNA of ML and 10 h IIL was extracted with Trizol reagent (Takara, Japan) and transcribed into cDNA as templates for qPCR. The genespecific primers were designed using Primer 5.0 software ( Table 1) and synthesized by Sangon Biotechnology Co. (Shanghai, China). The T. spiralis GAPDH gene (Gen-Bank: AF452239) was also amplified as an internal control gene [28]. The gene transcription level was evaluated using the comparative Ct (2 −ΔΔCt ) method as described [29,30]. Each sample was run in triplicate.

Statistical analysis
The data were analyzed with SPSS for Windows (version 20.0, SPSS Inc., Chicago, IL, USA). Relative gene expression levels in ML and 10 h IIL stages are shown as the mean ± standard deviation (SD), and the differences between two groups were assessed using Student's t test. P < 0.05 was regarded as statistically significant.

Molting and ecdysis
The molting of Trichinella IIL at 8, 10, 12, 14, 16, and 18 hpi is shown in Figure 1. The IIL began to molt at 10 hpi, and the larval molting rate was up to 28% at 12 hpi ( Figure 2). The copulatory appendages of the males were clearly observed at 16 hpi under microscopy.

Protein identification and quantification
The proteomic analysis allowed confident identification of 2885 T. spiralis soluble proteins. Of these proteins, 2740 co-existed in both ML and 10 h IIL stages, 29 solely in the ML and 116 only in 10 h IIL. Out of the 2740 co-existing proteins, 100 were significantly upregulated in the 10 h IIL stage (P < 0.05), and 78 were significantly downregulated in the 10 h IIL stage (P < 0.05) ( Figure 3). In this cohort, the downregulation of ADPribose pyrophosphatase, serine proteinase, angiotensin-converting enzyme, glycine c-acetyltransferase and carbonic anhydrase was more significant at the ML than at the 10 h IIL stage. At the same time, the upregulation of fatty acid synthase, purine nucleoside phosphorylase, prolyl 4-hydroxylase subunit alpha-2, cuticle collagen 14 was over twofold at the 10 h IIL compared with the ML stage. The significantly upregulated proteins that might be involved in molting are listed in Table 2. A clustering analysis of those 178 proteins common to ML and IIL is shown in Figure 4. A clear set of proteins was significantly downregulated at 10 h IIL, while the majority of IIL larva proteins were upregulated in comparison to the ML stage.

Gene Ontology annotation
To determine the biological functions of the differentially expressed T. spiralis proteins, the proteins were categorized into three types: biological process, molecular function and cellular components, based on their GO hierarchy. Based on their P value, 20 of the most prominent terms for ML and IIL larval proteins  133 are presented in Figure 5. In the molecular function ontology, the majority of protein function is relevant to ADP-ribose diphosphatase activity, endonuclease activity, structural components of cuticle, serine hydrolase activity and catalytic activity. Binding of iron ion, carboxylic acid, L-ascorbic acid and organic acid were the significant and enriched terms of binding activities. Under biological processes, most of the regulated proteins were involved in cellular processes and metabolic processes. DNA-templated transcription, initiation and protein maturation by iron-sulfur cluster transfer were the significant biological processes. There is a close relationship between a protein's subcellular location and its function. The cellular components collagen trimer and collagen-containing extracellular matrix were among the most significant 20 terms. These findings suggest that new cuticle formation and a large amount of energy metabolism may be a crucial adaptive method for T. spiralis larval molting. Interestingly, among the structural constituents of cuticle, we found that all upregulated proteins were collagens. Two proteins (E5RZT0 and E5SQP6) were upregulated at IIL compared with ML. At the same time, many cuticle collagen proteins (E5SQP9, A0A0V1BMR6, E5SR67, E5SAG9, A0A0V1BL19 and E5S488) were expressed only at the 10 h IIL stage.

Kyoto Encyclopedia of Genes and Genome (KEGG) annotation
The biological functions of these 323 differentially expressed T. spiralis proteins was further analyzed with the KEGG database, which mapped them to a total of 160 pathways. The most enriched 3 pathways are shown in Figure 6, including fatty acid biosynthesis, arachidonic acid metabolism and mucin type O-glycan biosynthesis. The pathway map of fatty acid biosynthesis is presented in Figure 7, which consists of several relevant proteins: fatty acid synthase (A0A0V1BWJ4, E5SGN7), mitochondrial malonyl-CoA-acyl carrier protein transacylase (A0A0V1BUL1) and 3-oxoacyl-[acyl-carrierprotein] synthase (E5SA15). Two fatty acid synthase proteins were upregulated, and the other two proteins were downregulated.

Protein interaction analysis
Network analyses of differentially expressed T. spiralis protein were conducted using the STRING database. These differential proteins were associated with each other to generate a network ( Figure 8). Proteins situated in the central area were closely connected with larval molting, and the proteins in the core node had high values. In comparison, proteins such as uncharacterized protein (A0A0V1C2F3), ribosome biogenesis protein WDR12-like protein (A0A0V1ARC8), angiotensin-converting enzyme (A0A0V1C1P4), WD repeat-containing protein 3 (A0A0V1BYS3), 28S ribosomal protein S11, mitochondrial (A0A0V1BAZ1) and uncharacterized protein (A0A0V1BIH0) were situated in the connected nodes of the protein functional interaction network. These proteins are involved in intrinsic components of membranes, metallopeptidase activity, ribosome biogenesis, and serine-type endopeptidase inhibitor activity.

qPCR validation of differentially expressed T. spiralis genes
To further assess the differential expression of the molting-related proteins in ML and 10 h IIL, the transcription levels of eight genes, which we chose based on interest and their different ratios, were determined via qPCR, and the results are shown in Figure 9. Out of these eight T. spiralis genes, five (cuticle collagen 14, putative DOMON    domain-containing protein, glutamine synthetase, cathepsin F and NADP-dependent isocitrate dehydrogenase) had a significantly higher transcriptional levels in 10 h IIL than that at the ML stage (P < 0.05), which were similar to their protein expression levels. However, the transcription levels of three genes (cytochrome P450 4V2, endoplasmic oxidoreductin-1 and RNA-binding protein 47) showed low compliance with their protein expression levels.

Discussion
Trichinella infection begins when infective muscle larvae are ingested. This larval phase has been used as the starting time for investigating larval ecdysis and the development of this nematode. Muscle larvae maintain a relatively stable state in the intracellular microenvironment. Once they enter the intestinal epithelium niche, four molts ensue that the larvae quickly grow to adulthood, within 31 h [7]. The selective pressures favor the larvae's completing the ecdysis during such a short time. However, the first molting times of the IIL larvae vary from 6 to 12 h [16]. The difference in larval molting time might be related to different hosts or/and isolates. Therefore, we observed the molting times of T. spiralis IIL using an individual isolate (ISS534) in the same host (BALB/c mice). Our results revealed that IIL began the first molting at 10 hpi, which was consistent with the results of others [16]. Hence, we selected ML and 10 h IIL to investigate their differentially expressed proteins. Trichinella larvae might require a physical barrier against which to rub for removing their outer cuticle layer and need a complex regulatory system to complete the sloughing of old cuticle and generation of new cuticle. Although complete larval ecdysis has been observed while the larvae were cultured in vitro, the molting mechanism was not clarified in as much detail as, for instance, in certain free-living nematodes such as Caenorhabditis elegans [12,13,31]. Advances in proteomics, genomics and bioinformatics techniques have provided a feasible way to study the molecular basis of nematode larval ecdysis and development [32]. In the present work, proteomic profiles of T. spiralis ML and 10 h IIL were analyzed with LC-MS/MS to identify the molting-related proteins. A total of 2885 proteins of T. spiralis were identified, 323 of which were differentially expressed.
Our results showed that some peptidases were upregulated in 10 h IIL, including cathepsin (E5SFB3, S5M797), metalloprotease (E5S4Y1, A0A0V1C237), and carboxypeptidase (E5SP83). However, we did not know whether these peptidases were molting-related enzymes. Previous studies showed that during the development of L3 to L4 larvae of lymphatic filarial Brugia pahangi, larval molting depended on the cathepsin L enzymatic activity, which was identified in excreted/secreted proteins of molting L3 larvae [33]. In C. elegans, an elevation of cathepsin L transcription level was observed in the intermolt period approximately 4 h before each molting. RNA interference with the enzymatic activity of cathepsin L brought about embryonic fatality and a prolonged development of larvae to adulthood, demonstrating a major function of this protease for embryogenesis and larva growth [34]. A previous study indicated that cathepsin L and Z were indispensable for ecdysis of Onchocerca volvulus L3-L4 larvae by immune-electron microscopy [35]. Similar to cathepsin, metalloproteases play an important role in cuticle ecdysis. The C. elegans nas-36 and nas-37 genes encode astacin metalloproteases. When they are mutated, larva molting defects result, e.g., the incapacity of the larvae to shed previous old cuticle [36]. In addition, carboxypeptidase is necessary for new cuticle generation and body morphogenesis in free-living and parasitic nematodes, and it is located at the hypodermis and interacts with other collagen-modifying enzymes [37]. Therefore, hydrolases play an important role in the molting process.
There are other types of proteases that play important regulatory roles in molting, such as E3 ligase (A0A0V1B0D3) and serine/threonine-protein kinase RIO1 (A0A0V1BAH6). Similarly, our results indicated that the expression of these two proteases was elevated in the 10 h IIL. In C. elegans, the expression level of E3-ligase RNF-5 is raised, especially during larval molting, and this enzyme ubiquitinates the dense body protein so that the old cuticle sheds [38]. NEKs are serine-threonine protein kinases that participate in cellular mitosis.
Using a genetic approach, a previous study revealed that NEKL-3 was indispensable for completing larval molting. A hypomorphic mutation of NEKL-3 and sv3 in C. elegans resulted in the inability of larvae to escape from the old cuticle. Inhibition of NEKL-2 produced the mislocalization of leucine-responsive regulatory protein-1 (LRP-1)/megalin, a cell surface receptor of low-density lipoprotein binding protein [39]. The LRP mutations of  C. elegans resulted in an obvious defect, an inability to remove all old cuticles at each larval molting stage [40]. Genetic mosaic analysis suggests that the LRP-1 gene plays a role in the major epidermal syncytium hyp7, a polarized epithelium that secretes cuticle from its apical surface. Consistent with this, NEKL-2 and NEKL-3 are expressed at the apical surface of epidermal syncytium hyp7 [39,40]. In our study, the expression of T. spiralis low-density lipoprotein receptor-related protein (A0A0V1AUS8) was consistent with serine/threonineprotein kinase.
In a GO enrichment analysis, the structural constituent of the IIL cuticle was the major category of molecular function. We found that some collagen proteins (E5RZT0, E5SQP6) were upregulated at 10 h IIL, and many cuticle collagen proteins (E5SQP9, A0A0V1BMR6, E5SR67, E5SAG9, A0A0V1BL19 and E5S488) were expressed only in 10 h IIL. Collagen is a structural protein of extracellular matrices. Nematodes can alter cuticle surface protein components between two molting stages or when the ambient environment changes; surface protein alteration can allow parasitic nematodes to escape from the host's immune defenses in the process of infection [41]. Previous studies revealed that schistosome tegument plays an important role in Schistosoma infection, such as by promoting juvenile worm growth, nutrition and immune evasion and modulation [42]. Other studies have indicated that the C. elegans cuticle consists of a complex collagen matrix. A single gene mutation of cuticle collagen could result in a cuticle defect that changes the nematode's morphology [43]. Nuclear hormone receptors (NHRs) play a vital role in collagen synthesis. A putative nuclear hormone receptor, NHR3 (A0A0V1AYJ5), was expressed only at the 10 h IIL stage. In C. elegans, CHR3 is crucial for cuticle formation and molting, while the NHR-23 expression and decreased NHR3 function during the intermolt time lead to a subsequent molt defect [44,45]. Carboxylic acid binding proteins (A0A0V1B0P6, A0A0V1B0I4 and A0A0V1B0M0) occupied the greatest proportion of binding activity proteins, and all these proteins were upregulated in 10 h IIL. However, the function of these proteins in the process of molting is unclear and needs to be further studied. T. spiralis undergoes a process of molting 4 times, which involves an interplay between various intracellular pathways. KEGG pathway analysis indicated that differentially expressed proteins were involved in three signaling pathways: fatty acid biosynthesis, mucin type O-glycan biosynthesis and arachidonic acid metabolism. Knockdown of the genes that encode fatty acid biosynthesis proteins with RNAi resulted in serious defects in triglyceride production and C. elegans larval moltings. Downregulation of these protein's gene expression impaired the new cuticle generation and destroyed the larval cuticle integrity [46]. Other studies demonstrated that a lipoxygenase pathway product is necessary for ecdysis of infective larvae of filarial parasites. When inhibitors of arachidonate metabolism were added to in vitro cultures containing B. malayi L3, they were capable of preventing larval development [47]. The relationship between mucin type O-glycan biosynthesis and ecdysis is unclear. Trypanosoma cruzi is capped with a dense coat of mucin-type glycoproteins, which is vital to facilitate this protozoon's intrusion and parasitism in the host cell [48]. Our results also showed that the protein interaction network center contained ribosome biogenesis protein WDR12-like protein (A0A0V1ARC8), Sel1 repeat family protein (E5S6N7), U3 small nucleolar RNA-interacting protein 2 (A0A0V1B0C4), 28S ribosomal protein, and mitochondrial (S11A0A0V1BAZ1). These proteins might be key for ecdysis and the development of IIL.
In this work, the qPCR results demonstrated that the transcription levels of five T. spiralis genes (cuticle collagen 14, putative DOMON domain-containing protein, glutamine synthetase, cathepsin F, and NADP-dependent isocitrate dehydrogenase) were significantly higher at the IIL stage than at the ML stage, which was consistent with the quantitative proteomic analysis results. However, the protein and mRNA levels of the other three genes (cytochrome P450 4V2, endoplasmic oxidoreductin-1, RNA-binding protein 47) were not consistent. The contradiction might be due to post-translational control, in which negative regulatory factors are possibly activated in the translation process, or the intrinsic mRNA is regulated by specific molecules such as microRNA [23]. The proteomic information obtained from this study is important for T. spiralis molting and development. These proteins are comparable to the homologous proteins of other nematodes and might be valuable as molecular targets to exploit vaccines and new drugs against this parasitic nematode. Furthermore, studies on molting-related protein function will further clarify the mechanism and control of molting in parasitic nematodes [31].
Our results are the first to establish extensive proteomic information on the molting larvae of T. spiralis. A total of 323 differentially expressed T. spiralis proteins were identified by LC-MS/MS analysis combined with bioinformatics, and their molecular functions (binding, catalytic and transporter activity) were annotated. Some proteins are primarily involved in cuticle structural elements, cuticle synthesis, maintenance, remodeling and degradation, hormonal regulation of molting and intracellular trafficking components and regulators. Five T. spiralis genes (cuticle collagen 14, putative DOMON domain-containing protein, glutamine synthetase, cathepsin F and NADP-dependent isocitrate dehydrogenase) had significantly higher transcriptional levels at the 10 h IIL than at the ML stage (P < 0.05), which were similar to their protein expression levels, suggesting that they might be T. spiralis molting-related genes. Further understanding of the regulatory mechanism of larval ecdysis and the function of ecdysis-related proteins will offer an important foundation to develop new vaccines and drugs against T. spiralis infection.