Molecular Characterization and Serodiagnostic Potential of Two Serpin Proteins in Psoroptes Ovis

Background: Psoroptes ovis a common ectoparasite of wild and domestic animals, leads to cutaneous inammation, extreme pruritus, scaly lesions, and causes an economically devastating loss of animal husbandry and animal welfare issues. Serine proteinase inhibitor (Serpin) is present in almost all organisms that are involved in host-pathogen interactions, inammatory responses, and reproductive development, etc. However, the research on P. ovis serpins is still limited. Methods: In this study, two serpins of P. ovis (PsoSP1 and PsoSP2) were cloned, and the molecular characterization was analyzed by bioinformatics. The transcriptional proles and tissue localization of PsoSP1 and PsoSP2 in P. ovis were investigated by quantitative real-time PCR and immunohistochemistry, respectively. The potential function of recombinant PsoSP1 and PsoSP2 (rPsoSP1 and rPsoSP2) in the serodiagnosis of P. ovis infestation in rabbits were evaluated by indirect enzyme-linked immunosorbent assay (iELISA). Results: Both of the 523 residue PsoSP1 and the 240 residue PsoSP2 proteins contained typical serpin domains and signatures. Both PsoSP1 and PsoSP2 expressed throughout the life cycle, more specically, signicantly higher expression in female mites than the larva, nymph, and male mites (p < 0.001). The native PsoSP1 and PsoSP2 proteins localized in ovary and mouthpart of adult female mites, respectively. Compared to rPsoSP2, the rPsoSP1 displayed better diagnostic eciency with higher values of sensitivity, specicity and the area under the receiver operating characteristic curve (AUC) than rPsoSP1 by iELISA (rPsoSP1 - rPsoSP2: 96.0% - 90.0%; 90.91% - 78.18%; 0.988 - 0.964, respectively). Moreover, the rPsoSP1 showed seropositive in 80% rabbits as early as the 2 nd week post-infestation (p.i.), prior to visible clinical signs and microscopy-positive of skin scrapings. Conclusions: These results suggested that these two serpins may play essential roles in reproductive development, blood-feeding, and pathogenicity of P. ovis. Compared to PsoSP2, PsoSP1 appeared as a potential antigen for serodiagnosis of P. ovis infestation in rabbits, especially at the early stage of infestation.


Introduction
Psoroptes ovis is a common ectoparasite of wild and domestic animals worldwide and is divided into subspecies according to the name of its host, e.g., P. ovis var. cuniculi from rabbits, P. ovis var. bovis from cattle [1]. This mites cause various features of the psoroptic disease, mainly characterized as intense cutaneous in ammation, extreme pruritus and crusted skin lesions [2,3]. Additionally, it causes severe economic losses and animal welfare issues, especially in sheep, rabbits, and cattle [3,4,5].
The obligate ectoparasite P. ovis, as a non-burrowing mite, spends its entire life on the surface of host skin. The life cycle lasts 10 ~ 19 days from egg hatch to egg production in adults [6]. P. ovis feeds on serous uids, lymph, and red blood cells [7]. Consequently, mite produces essential proteins to resist the host hemostatic system for successful blood feeding and self-proliferation. Meanwhile, it excretes allergens to promote the subsequent cutaneous in ammatory response [8,9]. Serine protease inhibitor (serpin), expressed in almost all organisms, has shown a variety of fundamental physiological functions in arthropods including anticoagulation, regulation in ammation response and reproductive development etc. [10]. It also played an essential role in host-pathogen interaction [11]. Besides, serpin may serve as a promising diagnostic antigen or vaccine candidate [12,13].
Recently, genome and transcriptome analyses revealed that serpins existed in P. ovis [14,15], but beyond that, no research has been reported on P. ovis serpins. Therefore, we are highly interested in the function of two serpins of P. ovis from rabbits (also known as Psoroptes cuniculi) (PsoSP1 and PsoSP2) which were identi ed based on our transcriptomic data [14]. In this study, we cloned and expressed the two recombinant PsoSP1 and PsoSP2 in prokaryotic expression vectors, performed the bioinformatics analysis. Additionally, we also investigated the transcriptional pro les as well as tissue localization in mites, and their potential e ciencies in the diagnosis of P. ovis infestation in rabbits were accessed by indirect enzyme-linked immunosorbent assay (iELISA). This is a preliminary study in relevance to the roles of these two proteins in P. ovis, that laying the foundation for further understanding of their functions.

Materials And Methods
Mite collection and RNA extraction P. ovis var. cuniculi were harvested from an infested New Zealand White rabbit maintained at the Department of Parasitology, Sichuan Agricultural University (Sichuan, China). About 300 mites, a pool of larvae, nymphs and adults, were collected and processed for the total RNA extraction using a MiniBest universal RNA extraction kit (TaKaRa, Dalian, China).

Cloning of two PsoSP genes
Total RNA was converted into cDNA following the protocols of the PrimeScript RT reagent Kit with gDNA Eraser (TaKaRa, Dalian, China). The PsoSP1 and PsoSP2 genes were ampli ed from cDNA by using the following primers: 5'-CGGGATCCGCTCATGTTGGTCAACATC-3' (forward) and 5' CCAAGCTTTTAAAAATCATGAATTTCACC-3' (reverse) for PsoSP1 with underlined restriction enzymes of BamH I and Hind III, and 5'-CGGGATCCTGAATGCGAATTCATTGCTG-3' (forward) and 5'-CCCTCGAG TCAAAATCCATGCATTTCACC-3' (reverse) for PsoSP2 with underlined restriction enzymes of BamH I and Xho I. The DNA fragments were puri ed and ligated into the pMD19-T (TaKaRa, Dalian, China). Then, the plasmids were transformed into Escherichia coli strain DH5α (TIANGEN Biotech Co. Ltd., Beijing, China), and sequences were con rmed by Invitrogen Biotechnology Company (Shanghai, China).

Expression and puri cation of PsoSP
The correct plasmids of two PsoSP genes were digested and ligated into the pET32a (+) expression vector (Invitrogen, Beijing, China), respectively. Then, the resulting constructs were transfected into E. coli BL21 (DE3) and induced by 0.5 mM isopropyl-β-D-thiogalactoside (IPTG) at 20 ℃ for 10 h. The recombinant PsoSP1 (rPsoSP1) and PsoSP2 (rPsoSP2) were harvested from the supernatant and inclusion body, respectively, and further puri ed by a Bio-scale TM Mini Nuvia TM IMAC Ni-Charged column (Bio-Rad, California, USA) using a step-wise elution with 50, 100, 200, 300 and 400 mM imidazole. The eluted fractions were concentrated by Amicon Ultra Centrifugal Filter Devices (Millipore, Billerica, MA, USA).

Rabbit sera
Fifty P. ovis var. cuniculi-positive rabbit sera were collected from a farm located in Chengdu, Sichuan, China. All rabbits were con rmed P. ovis var. cuniculi-positive by observation of ear scab and skin scrapings by microscopy (Ochs et al., 2001). Twenty-ve negative sera from P. ovis var. cuniculi-free rabbits were obtained from a farm without a history of psoroptic mange. For cross-reaction testing, another 30 sera included Sarcoptes scabiei-positive sera, Eimeria spp.-positive sera, and Cysticercus pisiformis-positive sera (ten/group) were provided by the Department of Parasitology, Sichuan Agricultural University.

Preparation of polyclonal antibodies and Western blotting
Polyclonal antibodies of anti-rPsoSP1 and anti-rPsoSP2 from rabbits were obtained by experimentally immunized with puri ed rPsoSP1 and rPsoSP2, respectively. The rabbits were immunized with about 1 mg puri ed recombinant protein mixed with an equal volume of Freund's complete adjuvant (Sigma-Aldrich) by subcutaneous injection at day 0. The next two immunizations were performed with mixture 1 mg protein and the same amounts of Freund's incomplete adjuvant at days 7 and 14, respectively. At 7 days after the nal immunization, sera were collected from rabbits via the marginal ear vein, and then puri ed by HiTrap Protein A a nity chromatography (Bio-scale TM Mini UNOsphere SUPrA TM Cartridge) to obtain the IgG of anti-rPsoSP1 and anti-rPsoSP2.
Two puri ed rPsoSP proteins were separated by 12% SDS-PAGE and transferred to the nitrocellulose membranes using Trans-Blot SD Semi-Dry Transfer Cell (Bio-Rad). The membranes were blocked using 5% skim milk powder for 2 h. After three times of 5-min washes with TBST (0.02 M Tris-HCl, pH7.6, 0.15 M NaCl, 0.05% Tween-20), membranes were incubated with rabbit anti-P. ovis var. cuniculi antibody or anti-rPsoSP1 IgG or anti-rPsoSP2 IgG (1:150 v/v) overnight at 4 °C. Non-infested rabbit serum was used as a negative control. After washing three times with TBST, membranes were incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit antibody (1:1000 dilution; Boster Bio-project Co. Dalian, China) for 1 h at room temperature. Following three washes with TBST, the signal was detected using an Enhanced HRP-DAB Chromogenic Substrate Kit (Tiangen, Beijing, China).
Immunolocalisation of PsoSP1 and PsoSP2 in adult female P. ovis var. cuniculi Adult female mites were xed in 0.8% (w/v) agarose, and then soaked in 4% paraformaldehyde in phosphate-buffered saline (PBS) for 24 h to investigate the tissue localization of PsoSP1 and PsoSP2. The mites were processed according to the para n-embedded tissue technique and sliced into 5 µm histological section using a rotary microtome (Leica, Frankfurt, Germany). After dewaxing and rehydration, parts were treated with 0.01 M citrate buffer and incubated at 37 °C with 5% (w/v) skimmed milk for 4 h, then incubated with puri ed rabbit anti-rPsoSP1 IgG or anti-rPsoSP2 IgG or pre-immune IgG Transcriptional pro les of PsoSP at different life stages of P. ovis var. cuniculi To analyze stage-speci c gene expression pro les of PsoSP1 and PsoSP2 in P. ovis var. cuniculi, the total RNA was extracted from larvae, nymph, and adult (male/female) mites using a MiniBest universal RNA extraction kit (TaKaRa), resptectively. Relative gene expression was evaluated by two-step qRT-PCR with the following primers: PsoSP1, 5'-TGGCAGCAGTGGATCAGAATCATC-3' (forward) and 5'-AATGCAACAGCAACACTGTATGGC-3' (reverse); PsoSP2, 5'-TCCTACATACACGTCCATCAACA-3' (forward) and 5'-TGGTACAATAGCGACGGCTG-3' (reverse). The β-actin gene was used as a housekeeping control to correct the relative uorescence signal of the target genes using primers 5'-TGAATTGCCTGATGGTCAAG-3 (forward) and 5'-TGGCGAACAAGTCTTTACGG-3 (reverse Establishment of an Indirect ELISA (iELISA) The establishment of iELISA was performed as described by Crowther and Walker [17]. The concentrations of antigen and primary serum samples were determined by the checkerboard titration tests. Brie y, the puri ed proteins were diluted two-fold in 0. PA, USA). The optimal working conditions were determined with the highest P/N (positive/negative serum) value. The cut-off value of iELISA was determined as the mean OD450 value plus three standard deviations (SD) using 25 negative serum samples from naïve rabbits [1] .
To further evaluate the feasibility of the iELISA, 50 P. oivs var. cuniculi-positive serum samples were assessed by the iELISA, and the sensitivity was calculated as (ELISA positive × 100) / true P. ovis var.
cuniculi-positive [1] . Thirty serum samples from rabbit infected with S. scabiei, Eimeria spp and C. pisiformis (10 samples for each species) were used to evaluate the cross-activity of the iELISA. Twentyve negative serum samples from naïve rabbits and 30 serum samples in the cross-activity assay were used to determine the speci city of the iELISA, which was calculated as (ELISA negative × 100) / real P.
The repeatability (intra-assay variability) and reproducibility (inter-assay variability) of the iELISA were evaluated using three P. ovis var. cuniculi-positive serum samples, substantially as previously described [19].
The experimental infestation of rabbits with P. ovis var. cuniculi and serological testing using the established iELISA Rabbits infected with P. ovis var. cuniculi were performed strictly as previously described [3]. Brie y, ten 3month-old naive New Zealand rabbits (5 females and 5 males) were infested on the aural region with approximately 200 mites of mixed stages of P. ovis var. cuniculi. Three non-infested rabbits were applied as controls. Serum samples from 13 rabbits were collected from the marginal ear vein at weeks 0, 1, 2, 3, and 4. Afterward, in a total of 65 serum samples (50 from the P. ovis var. cuniculi infestation rabbits and 15 from the non-infestation rabbits) were examined by the establishment optimal iELISA method. Each serum sample in triplicate was analyzed in one ELISA plate, and positive and negative controls were contained in the dishes.

Statistical analysis
All data are represented as mean ± standard deviation (SD), and statistical differences between groups were evaluated using Mann-Whitney U tests in SPSS software v.17.0. P-values < 0.05 were considered as statistically signi cant.

Bioinformatics analyses of two PsoSP
The PsoSP1and PsoSP2 shared 31.33% ~ 50.85% and 28.99% ~ 69.92% amino acid sequence identity with orthologs in Dermatophagoides farinae, D. pteronyssinus, Euroglyphus maynei, and Sarcoptes scabiei (Fig. 1). Secondary structure prediction showed that PsoSP1 protein was comprised of 12 helix and 3sheets, and PsoSP2 protein contained 5 helices and 3-sheets (Fig. 1a). A serpin domain was identi ed in the amino acid sequence from Arg78 to His493 for PsoSP1 and Asn4 to Met237 for PsoSP2 [20]. Moreover, both two proteins appeared to possess the speci c serpin signature at the deduced amino acid sequences from 496 to 506 (LRFDHPFLYFV) for PsoSP1 and from 213 to 223 (LSFDHPFLYFL) for PsoSP2, respectively ( Fig. 1a and 1b). The NJ tree revealed that PsoSP1 had a close relationship with D.

Acquisition of recombinant PsoSP1 and PsoSP2
The rPsoSP1 were mainly present in the supernatant with a single band of ~ 75 kDa, whereas rPsoSP2 principally present in insoluble inclusion bodies with the single bands of ~ 46 kDa (including ~ 18 kDa His-tag fusion peptide from pET-32a) (Fig. 3).

Western blotting
The rPsoSP1 and rPsoSP2 were reacted with P. ovis var. cuniculi-positive sera and the correspondent antiserum IgG from rabbits, but not negative sera, revealing the favorable reactivity and antigenicity (Fig. 3).
Tissue localization of PsoSP1 and PsoSP2 in adult female P. ovis var. cuniculi Immuno uorescence histochemistry indicated that native PsoSP1 and PsoSP2 were located in ovary and mouthpart of female mites, respectively ( Fig. 4b and 4c). No uorescence signal was observed in adult female mites using pre-immunized rabbit IgG antibodies (Fig. 4a).
Transcriptional pro les of PsoSP1 and PsoSP2 in P. ovis var. cuniculi Transcriptional pro les of PsoSP1 and PsoSP2 in different development stages of P. ovis var. cuniculi were investigated by qRT-PCR (Fig. 5). Both PsoSP1 and PsoSP2 were expressed throughout the life cycle of mites, more speci cally, signi cantly higher expression in female mites than larva, nymph, and male mites (p < 0.001) (Fig. 5).

Serodiagnosis potential of rPsoSP
By checkerboard titration, the optimal working conditions of iELISA were 46.0 µg/mL of rPsoSP1, 64.5 µg/mL of rPsoSP2 for coated antigens and a 1:100 dilution for rabbit sera. The cut-off values were 0.633 of rPsoSP1and 0.490 of rPsoSP2, respectively.
Serodiagnostic test of rabbits experimentally infested with P. ovis var. cuniculi After 4 weeks of post-infestation (p.i.), all infested rabbits were observed with the visable ear scabs. Meanwhile, P. ovis var. cuniculi showed positive in skin scrapings. By rPsoSP1-iELISA, the mean value of the anti-rPsoSP1 level from the infestation group revealed an increase from 1 to 4 weeks p.i (Fig. 8). The positive anti-rPsoSP1 above cut-off value was rstly detected with 2/10 serum samples at 1-week p.i. in the infestation group. Afterward, the rate of positive serum gradually increased to 80% (8/10) at 2 and 3 weeks p.i., then up to 100% (10/10) at 4 weeks p.i. (Fig. 8). In the non-infestation group, the anti-rPsoSP1 antibody appeared below the cut-off value throughout the experiments.

Discussion
In the present study, two P. ovis serpins were characterized, and the potential of the recombinant proteins was evaluated for serodiagnosis of P. ovis infestation in rabbits. The predicted amino acid sequence showed the low overall identity of serpins compared to other mites (31.33% ~ 50.85% of PsoSP1and 28.99% ~ 69.92% of PsoSP2), however, these two target proteins were identi ed as typical serpins due to the presence of the features such as serpin domain and serpin signature in C-terminal end [21]. PsoSP1 shared the highest amino acid sequence identity (50.85%) with the newly characterized D. farinae Der f 27 allergens, which has been proven to orchestrate the pulmonary in ammatory response and mediate Th2 type response in mouse [22]. Besides, NJ analysis revealed that PsoSP1 yielded a close relationship with Der f 27. In combination with the homology and the genetic relationship between Pso SP1 and Der f 27, PsoSP1 may be considered as an allergen of P. ovis, which was possibly associated with the instigation of the host cutaneous pro-in ammatory response [23]. Additionally, this cutaneous in ammation resulted in serum extravasation to provide su cient food for mites population growth and cause aggravation of scabby lesions [7,24]. The expression of PsoSP1 and PsoSP2 in all stages of mites indicated that PsoSP1 and PsoSP2 possibly play an essential role in the development of P. ovis. However, signi cant differences were seen for the transcription of PsoSP1 in female mites being the highest level of expression with a 347-fold change. In addition, the native protein was located in the ovary of female mites, indicating that PsoSP1 possibly was essential in vitellogenesis [15,25]. This role of serpin being involved in vitellogenesis has been proven in a recent research, which indicated RNAi of serpin gene resulted in a reduction of yolk granule accumulation in Rhipicephalus haemaphysaloides [26]. In addition, previous studies have shown that serpins of ticks took part in anti-coagulation [27,28], and were considered as the vaccine candidate against tick infestation [27] [13]. In this study, the localization of native PsoSP2 in mouthpart of female mites and the expression of PsoSP2 throughout the life stages of mites suggested that PsoSP2 may appear to be vital in mites for anti-coagulation to successful bloodfeeding [7,11]. The similar obligate hematophagous ectoparasites of ticks and mites indicate that PsoSP2 could serve as a potential vaccine candidate or therapeutic target as the alternative for current chemical acaricides.
Psoroptic mange spreads rapidly under crowded conditions and causes major morbidity in the rabbit breeding industry in China [29], resulting in an economically devastating loss. Thus, timely diagnosis and treatment of P. ovis infestation in rabbits are paramountly important to reduce the risk of disease transmission and improve pro tability. In China, the current microscopic diagnosis for this disease is extremely time-consuming and ine cient in the low mite carriers and sub-clinical infestations in rabbits. Thus, it is imperative to seek for the effectively immunoreactive antigens for rapid and accurate diagnosis of the P. ovis infestation in rabbits. Furthermore, animals infested P. ovis could evoke serospeci c antibody [2,30], and this sero-speci c antibody was induced at the early phase of parasite infestation when animals appeared asymptomatic [2,3,31]. Thus, enzyme-linked immunosorbent assay (ELISA) can be considered as an accurate method compared to the microscopy of skin scrapings at the low mite carriers and sub-clinical infestations because of its speedy, high sensitivity, and handling convenience. In a previous study, serpin of Schistosoma mansoni was considered as a promising diagnostic antigen in human schistosomiasis [12]. Therefore, in this study, we evaluated the serodiagnostic potential of rPsoSP1 and rPsoSP2 by the establishment of the iELISA. Compared to rPsoSP2-iELISA, the rPsoSP1-iELISA displayed better diagnostic e ciency with higher values of sensitivity, speci city and AUC (rPsoSP1 -rPsoSP2: 96.0% − 90.0%; 90.91% − 78.18%; 0.988 − 0.964, respectively). Although rPsoSP1 showed cross-reaction with sera from 3/10 S. scabiei-infestation, the cross-reaction between these two ectoparasites have been proved commonly in other studies [3,32,33].
Fortunately, these two mite species were effectively treated with the same acaricide [34,35]. Besides, 1/10 rabbits infested with C. pisiformis and Eimeria spp. showed sero-reaction with rPsoSP1, however, their OD values were close to the cut-off value and appeared markedly lower than those rabbits infested with P. ovis var. cuniculi (p < 0.001). Moreover, rPsoSP1-iELISA can detect seropositive in 80% (8/10) rabbits as early as the 2nd-week p.i., prior to visible clinical signs and microscopy-positive of skin scrapings. Regarding the high sensitivity and speci city, PsoSP1 was more suitable as a candidate antigen for serodiagnosis of P. ovis infestation in rabbits, especially at the early stage of infestation.

Conclusions
In conclusion, PsoSP1 and PsoSP2 displayed the typical characterization of serpin superfamily with the regular serpin domain and signature. The expression of PsoSP1 and PsoSP2 were found in all life stages of mites, with signi cantly high expressions in adult female mites. Compared to rPsoSP2, rPsoSP1 seemed to display a better diagnostic e ciency than PsoSP2 by iELISA, suggesting that PsoSP1 could be developed as a potential antigen for serological diagnosis of P. ovis infestation in rabbits, especially at the early stage of infestation.  The neighbor-joining (NJ) tree was constructed based on the deduced amino acid sequence of serpin.

Declarations
The number beyond every branch is the bootstrap value of 1000 replications (%).  Relative transcriptional pro les of PsoSP1 (a) and PsoSP2 (b). The internal reference gene was β-actin in the study. Data are represented as the mean with standard deviation (SD) in triplicate (*** indicates the statistical signi cance of p < 0.001).  The Receiver operating characteristic (ROC) curves of the rPsoSP1-iELISA and rPsoSP2-iELISA for the detection of antibodies against P. ovis var. cuniculi. The ordinate represents the sensitivity of the iELISA.
The abscissa represents the 1-speci city of the iELISA. The purple line shows the mean area under the curve (AUC) plot of rPsoSP1-iELISA, and the green line shows the mean area under the curve (AUC) plot of rPsoSP2-iELISA.