Specific autoantigens identified by sera obtained from mice that are immunized with testicular germ cells alone

There are various autoimmunogenic antigens (AIs) in testicular germ cells (TGCs) recognized as foreign by the body’s immune system. However, there is little information of TGC-specific AIs being available. The aim of this study is to identify TGC-specific AIs. We have previously established that immunization using viable syngeneic TGC can also induce murine experimental autoimmune orchitis (EAO) without using any adjuvant. This study is to identify TGC-specific AIs by TGC liquid chromatography–tandem mass spectrometry analysis, followed by two-dimensional gel electrophoresis that reacted with serum IgG from EAO mice. In this study, we identified 11 TGC-specific AIs that reacted with serum from EAO mice. Real-time RT-PCR analysis showed that the mRNA expressions of seven TGC-specific AIs were significantly higher in only mature testis compared to other organs. Moreover, the recombinant proteins of identified 10 (except unnamed protein) TGC-specific AIs were created by using human embryonic kidney 293 (HEK293) cells and these antigencities were reconfirmed by Western blot using EAO serum reaction. These results indicated Atp6v1a, Hsc70t, Fbp1 and Dazap1 were candidates for TGC-specific AIs. Identification of these AIs will facilitate new approaches for understanding infertility and cancer pathogenesis and may provide a basis for the development of novel therapies.


Results
Explanation of the experimental design. We induced EAO by immunization using viable syngeneic TGC without using any adjuvant. B cell infiltration and IgG deposit were detected by immunohistochemistry in the testis of EAO. IgG titers in TGC-induced EAO mice were extremely high as detected by enzyme-linked immunosorbent assay (ELISA). Then, we identified TGC-specific AIs by TGC liquid chromatography-tandem mass spectrometry (MS) analysis, followed by two-dimensional gel electrophoresis (2D), which showed serum IgG from EAO mice reaction. The expression pattern of identified TGC-specific AIs was analyzed by the Real time PCR. The recombinant proteins of identified 10 (except unnamed protein) TGC-specific AIs were created by using human embryonic kidney 293 (HEK293) cells and these antigencities were reconfirmed by Western blot using EAO serum reaction.

Histopathological changes in TGC-immunized EAO mice. No lymphocyte infiltration was observed
in any of the control mice testes (Fig. 1a-c). Conversely, extensive lymphocytic infiltration with spermatogenic disturbances was observed in all TGC-immunized EAO mice testes ( Fig. 1e-g). Numerous lymphocytes surrounded the peripheral seminiferous tubules, resulting in aspermatogenesis ( Fig. 1e-g). No. inflammation was observed in the epididymis of the TGC-immunized EAO mice (data not shown). Immunohistochemical analysis revealed that a portion of the lymphocytes in the interstitium and in the seminiferous tubules were B220-positive cells (Fig. 1g). Moreover, deposits of immunoglobulin (Ig) G were detected a portion of the lymphocytes that accumulated in the interstitium (Fig. 1h). No deposits of IgG and B220 were detected in any of the negative controls (Fig. 1c,d).
ELISA of TGC autoantibodies in TGC-immunized EAO mice. ELISA analysis revealed that levels of TGC-reactive IgG, IgA, and IgM significantly increased substantially in TGC-induced EAO mice compared with those in control mice (Fig. 2). In particular, IgG titers in TGC-induced EAO mice were extremely high.
Identification of AI using 2D in TGC-immunized EAO mice. Because IgG titers in TGC-induced EAO mice were higher than those of IgA and IgM, IgG was used as a secondary antibody for detecting TGC-specific AIs. Spot numbers were only assigned to spots identified by mass spectrometry (MS) using TGC autoantibody as a primary antibody (Table S1). All spots from EAO and control were detected by MS. However, because the spots in which multiple proteins are mixed together, they cannot be identified by MS, thus spot number was not attached. All proteins that visualized on a silver-stained gel were excised and processed for protein identification using MS (Fig. 3a,d). Altogether, isoelectric point (pI) 4-7 set of 29 protein spots ( Fig. 3c and Table S1) and pI Scientific RepoRts | 6:35599 | DOI: 10.1038/srep35599 6-9 set of 10 proteins spots ( Fig. 3f and Table S1) were identified that reacted with EAO serum sample, whereas pI 4-7 set of 23 protein spots ( Fig. 3b and Table S1) and pI 6-9 set of two proteins spots ( Fig. 3e and Table S1) were identified that reacted with control sera. Some spots (spot no. 1, 3-9, 11-14, 17-20, 22-23, 25-26, 28-31, and 33) overlapped between the EAO and control serum. Other spots (spot no. 2, 10, 15, 16, 21, 24, 27, 32, and 34-36) in EAO serum were not identified in control sera. In effect, we have identified the spots which reacted only with the serum IgG of EAO as TGC-specific AIs. These spots were summarized in Table 1. Figure 4 the mRNA expressions of Tubb2c, Pdhb, Hsc70t, Fbp1, Lrrc34, Gapdhs, Pdha2, Dazap1 and the unnamed protein in testes were significantly higher those in the other organs. Atp6v1a mRNA expression in the testis was significantly higher different between the testis and various other organs except for the brain (p = 0.11). Dnpep mRNA expressions in testes were not significantly different from the ones in epididymis (p = 0.57), submaxillary gland (p = 0.33), spleen (p = 0.88), small intestine (p = 0.82), liver (p = 0.28), lung (p = 0.44), while it was significantly higher than heart, pancreas, kidney, muscle, and brain.  . HE staining and B220 and IgG deposit detection on testicular sections from control (a-d) and TGCinduced EAO (e-h) mice. Testes histological sections from control (a,b) and TGC-induced EAO (e,f) mice were stained with hematoxylin and eosin. Additional tissue sections were incubated with specific antibodies to detect B220 (c,g) and IgG deposits (d,h) in testes from control (c,d) and EAO (g,h) mice. The presence of infiltrating B-cells (g) and IgG deposits (h) with disrupted spermatogenesis was observed in the testes from TGC-induced EAO mice. Brown spots indicate positive cells (g,h). Scale bar: 150 μ m (a-f). Comparison of mRNA expressions of AIs between TGC and epididymal spermatozoa. The mRNA expressions of all AIs were significantly different between TGC and epididymal spermatozoa (ES). The mRNA expressions of Tubb2c, Atp6v1a, Pdhb, Hsc70t, Fbp1, Lrrc34, Gapdhs, Pdha2, Dazap1, and the unnamed protein in TGC were significantly higher those in ES. On the contrary, Dnpep in TGC was significantly lower than that in ES (Fig. 6). We decided to include Tubb2c, Atp6v1a, Hsc70t, Fbp1, Lrrc34, Gapdhs, and Dazap1 from the real-time PCR as candidates for TGC-specific AIs.

Discussion
Novelty of identified TGC-specific AIs. This study identified 11 AIs using serum autoantibodies from TGC-induced EAO mice. AIs related to testicular autoimmunity have been previously identified using various methods. Primakoff et al. 38 and Tung et al. 39 identified that immunization of male guinea pigs with the sperm surface protein sperm adhesion molecule 1 (SPAM1, PH-20) reproducibly resulted in infertility. They demonstrated that EAO could be induced in guinea pigs immunized with PH-20 + CFA, but they could not induce EAO by the same method in mice 38,39 . Recently, Fijak et al. reported that disulphide isomerase endoplasmic reticulum 60, heat-shock 70 kDa protein 5, heterogeneous nuclear ribonucleoprotein H1, and sperm outer dense fiber major protein 2 are AIs related to rat EAO in a TH + CFA + BP-induced EAO model 36 . The mice injected with these proteins in CFA had induced EAO at a rate of 25%. PH-20, disulphide isomerase endoplasmic reticulum 60, heatshock 70 kDa protein 5, and sperm outer dense fiber major protein 2 were expressed in the testis and epididymis in humans and mice [40][41][42][43] . Heterogeneous nuclear ribonucleoprotein H1 was expressed at testicular somatic cells, including Sertoli and Leydig cells, at 0 and 1 week old in mice 36 . Additional expression was observed in spermatocytes and spermatids at 12 weeks in humans and mice 44 . Moreover, following mRNA analysis, it was found that sperm outer dense fiber major protein 2 is expressed in the ovary and uterus 45 , and heat-shock 70 kDa protein 5 is expressed in the lung, pancreas islet, and kidney 46 . Ten of our 11 detected proteins detected are novel AIs that have not been reported.

TGC-specific AIs details.
Relevance to EAO with respect to molecular weight and age of the identified TGC-specific AIs. We previously investigated AIs relevant to TGC-induced EAO through reacting individual immune serum samples with testes from normal mice of various ages using immunoblotting 47 . The results showed that the sera obtained from mice with TGC-induced EAO lesions specifically defined testicular antigens with molecular weights of 15 kDa, 40 kDa, 75 kDa, and > 200 kDa from 4-week-old mice 47 . In addition to these, bands of 24-35 kDa and 60-75 kDa were detected in testicular proteins at 8 weeks of age 47 . Moreover, we incubated immunoblots with EAO serum from various weeks after TGC immunization 32 . At 0 week of age, a band corresponding to approximately 40 kDa was detected in TGC proteins, indicating the presence of natural autoantibodies against TGC. At 4 weeks of age, in addition to this band, a band of approximately 42 kDa was detected in TGC proteins. At 8 weeks of age, three TGC protein bands (approximately 20, 28, and 60 kDa) were also detected 32 . Considering the EAO-inducing factors and expression levels in testis at 2 weeks and 8 weeks, the EAO-related proteins were 24-35, 42, and 60-75 kDa. In this study, we identified 11 AIs recognized by serum antibodies from TGC-induced EAO in mice extracted Tubb2c (50 kDa), Atp6v1a (69 kDa), Hsc70t (71 kDa), Fbp1 (37 kDa), Lrrc34 (47 kDa), Gapdhs (48 kDa), Pdha2 (44 kDa), Dazap1 (43 kDa), and the unnamed protein product (55 kDa) from mRNA expression analysis at 2 weeks and 8 weeks of age. Therefore, Atp6v1a, Hsc70t, Fbp1 Lrrc34, Gapdhs, and Dazap1 can reasonably be categorized as EAO-related proteins.
Focus of candidate TGC-specific AIs. We have examined the candidates for TGC-specific AIs from the literature and our experiment. We decided to include Tubb2c, Atp6v1a, Hsc70t, Fbp1, Lrrc34, Gapdhs and Dazap1 as the candidates for TGC-specific AIs from the results of real-time PCR (tissue specificity, age, cell type) in this study. In addition, we selected to include Atp6v1a, Hsc70t, Fbp1 Lrrc34, Gapdhs, and Dazap1 based on molecular weight and age from the literature as potential candidate TGC-specific AIs. Furthermore, we have decided to include Lrrc34, Hsc70t and Dazap1 from testicular localizations, as described in the literature as candidate TGC-specific AIs. Finally, the recombinant proteins of identified 10 (except unnamed protein) TGC-specific AIs were created by using human embryonic kidney 293 (HEK293) cells and these antigencities were reconfirmed by Western blot using EAO serum reaction. The results indicated Atp6v1a, Hsc70t, Fbp1 and Dazap1 were candidates for TGC-specific AIs. Usefulness of the information of identified TGC-specific AIs. Information on TGC-specific AIs identified in germ cell developmental biology. In male germ cells, many marked morphological changes occur during spermatogenesis, particularly in haploid spermatids after meiotic division 64 . In various mammals, male germ cell differentiations proceed actively and continuously in the testis after puberty, and sperms are produced throughout adulthood 64 . Mice require 1 month for completion of spermatogonial stem cell proliferation and differentiation, meiosis, generation of haploid germ cells, and morphogenesis of the developing testicular spermatozoa in seminiferous tubules 64 . After meiotic division (during the process of haploid germ cell differentiation, or spermiogenesis), the rounded spermatids undergo marked morphological changes to become testicular spermatozoa: the nucleus assumes a compact shape, the mitochondria are rearranged, the flagellum forms, and the acrosome is generated. During this period of differentiation, which occurs within 5-6 weeks in humans 65,66 and 2-3 weeks in mice 67 , haploid germ cells do not divide but morphogenesis occurs, indicating that some regulatory mechanism arrests the cell cycle. Searching for functional changes in genes and gene products involved in male infertility would increase our understanding of the causes of this condition and perhaps lead to new treatments for some cases. The simplest strategy for elucidating the mechanism of spermatogenesis is to identify and characterize differentiation-specific molecules and their associated genes in germ cells. In this study, the mRNA expressions of 10 AIs, excluding Dnpep, were significantly higher in mice at 8 weeks of age compared with those at 2 weeks and in TGC compared with epididymal spermatozoa (ES). Thus, these proteins may be involved in the development of testicular spermatozoa.
Information on identified TGC-specific AIs as cancer/testis antigens (CTs). CTs, also known as cancer germline antigens, refer to a growing list of antigens that were initially discovered in the 1980 s-1990 s, which are specifically expressed in various tumor types 68,69 . At present, more than 70 CTs are present families encompassing more than 140 individual members with largely unknown functions 69,70 . Because each cancer type is associated with multiple highly expressed CTs, an effective vaccine may require the presence of multiple CTs. CTs are absent in normal humans and rodent somatic cells and are expressed only in male TGC 34,[71][72][73][74][75] . In this study, the mRNA expressions of seven TGC-specific AIs (including Tubb2c, Atp6v1a, Hsc70t, Fbp1, Lrrc34, Gapdhs and Dazap1) were significantly higher in testis, particularly in male germ cells, compared with that in other organs in 8-week-old mice than in 2-week-old mice. Generally, CTs can be grouped into two classes based on their chromosomal location: CTs-X are located on the X chromosome and non-X CTs are located on the autosomes. Most CTs-X are unique to primates and constitute several subfamilies of homologous genes, organized in discrete clusters along the X chromosome 76 . As a result of their restricted expression in an immune-privileged organ, both CTs groups represent attractive immunotherapy targets 34 . Antigenic CTs-derived peptides that are presented to the immune system with different human leukocyte antigen allospecificities elicit both humoral and cellular immune responses. Spontaneous humoral and cell-mediated immune responses have been demonstrated for several CTs and patients with good antibody titers often present with a better prognosis 71 . TGC-induced EAO involves both cellular and humoral immune responses to the autoantigen-containing AIs. Therefore, we expect that cellular and humoral immunoreactions to AIs can be identified in TGC-induced EAO. Although emerging evidence has clarified the functions of a few CTs-X in cancer, the majority remains poorly understood. In contrast, non-X CTs are fairly well conserved throughout evolution, with established roles in processes such as transformation 76 , chromatin remodeling 34 , transcriptional regulation, and cell signaling 72 . Thus, non-X CTs are particularly promising targets for the development of small-molecule therapeutics 77 . In fact, we examined whether the candidate TGC-specific AIs are upregulated in cancer cells or tumors from previous reports (Table 3). These studies suggested that Tubb2c, Pdhb, Hsc70t, Fbp1 and Gapdhs are upregulated in some cancer cells or tumors (Table 3). A list of CTs has been

Conclusion
Using serum autoantibodies from mice immunized with only syngeneic TGCs alone, we have identified 11 proteins as respective testicular AIs. Real-time RT-PCR analysis showed that the mRNA expressions of seven TGC-specific AIs (Tubb2c, Atp6v1a, Hsc70t, Fbp1, Lrrc34, Gapdhs, Dazap1) were significantly higher in only mature testis compared to other organs. Three TGC-specific AIs (Hsc70t, Lrrc34, Dazap1) were shown as EAO-related proteins in previous reports and the present study also. Seven TGC-specific AIs (including Tubb2c, Atp6v1a, Hsc70t, Fbp1, Lrrc34, Gapdhs and Dazap1) show human homology. Finally, the recombinant proteins of identified 10 (except unnamed protein) TGC-specific AIs were created by using human embryonic kidney 293 (HEK293) cells and these antigencities were reconfirmed by Western blot using EAO serum reaction. These results indicated Atp6v1a, Hsc70t, Fbp1 and Dazap1 were candidates for TGC-specific AIs. Information of AIs have also received considerable attention because of their role as cancer/testis antigens (CTs). Identification of these AIs will facilitate new approaches for understanding infertility and cancer pathogenesis and may provide a basis for the development of novel therapies. In the future, we are planning to generate some recombinant proteins of TGC-specific AIs. The TGC suspension contained more than 99% germ cells at various stages of spermatogenesis; the remaining < 1% consisted of Sertoli and interstitial cells 21 . The ratio of prepared cells was determined by visual microscopic inspection using a hemocytometer. In addition, prepared cells have been confirmed by immunohistochemical method and real-time RT-PCR ( Figure S1).

Induction of EAO by TGC.
At 10 weeks of age, male mice were subcutaneously injected with 1 × 10 7 TGC/mouse once on day 0 and again on day 14 (i.e., at a 2-week interval) for induction of active EAO. Male mice (aging 10 weeks) injected with HBSS alone were used as controls. At 120 days after the first immunization, the mice were deeply anesthetized with pentobarbital (65 mg/kg body weight) and their testes were removed. Histopathological samples were taken from the right testes, and immunohistochemical samples were collected from the left testes of the mice (n = 10 for both the TGC-immunized mice and control mice). Blood samples were collected from the mice by cardiac puncture (n = 10 for each group).

Examination of inflammatory factor by B220 and IgG deposit stain. Isolated left testes obtained
from EAO-affected mice and control mice were placed in OCT compound (Miles Laboratories, IL, USA), frozen in liquid nitrogen, and stored at − 80 °C until use. Five-micrometer-thick sections were cut with a cryostat (CM1900; Leica, Wetzlar, Germany) and then fixed in ethanol for 10 min at − 20 °C. The sections were then rinsed in phosphate-buffered saline (PBS) and incubated with Block Ace (Yukijirushi, Hokkaido, Japan) for 20 min at room temperature to inactivate endogenous peroxidase activity. After rinsing in PBS, the sections were incubated with a rat anti-mouse B220 (clone: RA3-6B2, × 200; BD Biosciences) monoclonal antibody, followed by incubation with rabbit anti-rat IgG (Vector Labs, CA, USA) at room temperature. Bound antibodies were detected by incubation with horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (ZyMax, South San Francisco, CA) at room temperature. Immunoreactive cells were visualized using a Vectastain ABC Kit (Vector Labs, CA, USA) with 3,3′ -diaminobenzidine (DAB) as the chromogen. Bound HRP was detected using 0.05% DAB and 0.01% H 2 O 2 . Sections processed with rabbit serum instead of the primary antibodies were used as negative controls. The stained sections were counterstained with methylgreen (Vector Laboratories, CA, USA).

Examination of autoantibody (IgG, IgM, IgA) titer by ELISA.
ELlSA was performed using the method described by Hirai et al. 32 . For preparation of antigens, TGC obtained from 10-week-old normal mice (n = 3) was homogenized in carbonate-bicarbonate buffer (Sigma-Aldrich, MO, USA). Antigen concentrations were determined using the Bradford method with bovine serum albumin as the standard. Antigen was then added to each well of a microtiter plate (Nunc 96-well plate; Thermo Fisher Scientific, Kanagawa, Japan) and incubated at 37 °C for 30 min. After removing the coating solution, the wells were washed three times with PBS-Tween 20; Then, 200 μ l of a rabbit anti-mouse monoclonal glyceraldehyde 3-phosphate dehydrogenase (GAPDH) antibody (housekeeping gene; Bethyl Laboratories, Inc., TX, USA; 1/200 and 1/1000 dilution) or experimental mouse serum samples serially diluted with 1% goat serum in PBS (each sample run in duplicate) were added to microE-LISA wells and incubated for 2-h at room temperature. The wells were then washed five times with PBS-Tween 20 and incubated for 2-h at RT with 50 μ l of HRP-conjugated goat anti-rabbit IgG (Cappel, PA, USA; 1:1,000), anti-GAPDH antibody, or HRP-conjugated goat anti-mouse IgG (Cappel; 1:1,000). A solution of one Alkaline Phosphatase Yellow (pNPP) tablet and one Tris Buffer tablet (Sigma-Aldrich) dissolved in 5 ml of water was used as a soluble substrate to detect alkaline phosphatase activity. After the wells were again washed five times with PBS-Tween 20, 200 μ l of freshly prepared pNPP solution was added to each well. Thirty minutes later, the absorbance at 492 nm was determined using a microtiter plate reader (MPR-A4, Toyo Soda, Tokyo, Japan).

AI protein identification by Western blot and MS. In mice immunized with isolated TGC alone, only
EAO is inducible without epididymo-vasitis 21 . This EAO model has been found to involve Th1-cell-dependent autoimmunity; however, B-cell lineages were also related to the development of EAO in the present study. Several IgG immune deposits are present in EAO lesions. Various immunoglobulins (IgG, IgM, and IgA) against TGC were detected in TGC-induced EAO mice sera in the present study. TGC-induced EAO is unique as serum autoantibodies are only produced against the acrosomes of ES and round spermatids 25,26,32 . Previously, we had observed that there were some molecular differences between TGC and ES antigens that reacted with autoantibodies using Western blot 32 . Considering that epididymitis does not occur in TGC-induced EAO, the AI analysis was performed using TGC rather than ES. After gel electrophoresis, proteins were transferred onto a nitrocellulose membrane. Membranes were blocked in 4% skimmed milk in TBST for 1 h, followed by serial incubation with EAO and normal serum samples (1:50) 4 °C overnight, a biotinylated goat anti-mouse whole IgG antibody (Amersham Biosciences, Freiburg, Germany; 1:10000), and finally, a streptavidin-HRP conjugate for 30 min (Amersham Biosciences). Bound antibodies were visualized using the ECL Plus Detection Reagent (Amersham Biosciences). Spots corresponding to the Western blot membrane were cut from the silver stained 2D gel for identification by MS. Peptidase protein digestion was performed as follows: A piece of a gel spot stained with silver was placed in a sampling tube, dried with a vacuum concentrator, rehydrated with a trypsin solution (10 μ m/ml in 50 mM ammonium bicarbonate) and then placed in a small volume of 50 mM ammonium bicarbonate. After the incubation at 37 °C under vigorous shaking for 12-16 h, the supernatant solution was removed and stored for further use. The gel piece was successively extracted once with 50 μ l and again with 25 μ l of 5% trifluoroacetic acid/50% acetonitrile with shaking for 30 min each time. The supernatant and extracts were mixed and concentrated to a volume of 5-10 μ l.
Examination of expression of TGC-specific AIs gene (tissue specificity, age, cell type) by gene expression analysis. Testes, epididymis, submaxillary gland, spleen, heart, kidney, muscle, small intestine, liver, brain, lung, and pancreas from normal control A/J mice were evaluated at 8 weeks of age (n = 4) for expression analysis in the various organs. Testes from normal A/J mice were evaluated at 2 weeks of age (n = 4) for analysis of expression that was related to spermatogenesis. TGC and ES were excised from 10-week-old mice (n = 4), minced with scissors in cold HBSS, and passed through a stainless steel mesh. TGC and ES were harvested by centrifugation at 400 g for 15 min and washed three times in cold HBSS.
For analysis, total RNA was isolated from the entire above organs and cells using the TRIzol RNA Extraction Kit (Invitrogen, CA, USA), according to the manufacturer's instructions, and the RNA pellets were dissolved in 10 ml of RNase-free distilled water. Total RNA was measured at 260/280 nm using a UV spectrophotometer and was stored at − 80 °C prior to use. cDNAs were prepared from 10 μ g of total RNA in a 100 μ l reaction mixture using random primers according to a standard protocol (high capacity cDNA archive kit; PE Applied Biosystems,  Foster City, CA, USA). The PCR reactions were conducted using an iCycler thermal cycler (Bio-Rad, Hercules, CA, USA), and the mixtures were stored at − 80 °C until analysis. Real-time RT-PCR was performed using 3 ng of cDNA and the validated SYBR Green gene expression assay in combination with the SYBR Premix Ex Taq II (TaKaRa, Bio Inc., Ohtsu, Japan) for measuring Tubb2c, Atp6v1a, Pdhb, Hsc70t, Fbp1, Lrrc34, Tyrp1, Gapdhs, Pdha2, Dazap1, Fh1, and GAPDH. All of the primers used in this analysis are listed in Table 4. Quantitative real-time PCR was performed in duplicate in a thermal cycler dice real time system TP800 (TaKaRa). The data were analyzed using thermal cycler dice real time system software (TaKaRa), and the comparative C t method (2∆ ∆ C t ) was used to quantify gene expression levels. Real-time RT-PCR data were standardized to GAPDH, which was used as an internal control. To confirm the specific amplification of the target genes, each gene product was further separated on a 1.5% agarose gel to detect any single bands at the theoretical product sizes, and the dissociation curves were analyzed to detect any single peaks.

Examination of expression of TGC-specific AIs gene by Human embryonic kidney 293 (HEK293) cells.
Plasmid vectors (pCMV6) of TGC-specific AIs and Native were purchased from Origene (Table S2). All proteins were expressed in a suspension-adapted HEK293 cell line (ExPi293F, Thermo Fisher). Cells, growing in deep 6-well blocks, were transfected in triplicate at a density of 1.0 × 107/ml using ExpiFectamine ™ 293 according to the manufacturer's instructions. Following transfection, the cells were grown at 37 °C by shaking in 5% CO 2 , for 48 hours. The conditioned media, containing secreted protein, were harvested by centrifugation at 1000 rpm for 5 minutes. The clarified media from each of the three replicates for each expression vector were pooled and stored at − 20 °C. Each transfection was performed in duplicate, on separate days, and the conditioned media was analyzed separately by Western blot analysis. Data analyses. Data are expressed as the mean ± standard devtion (SD). ANOVA and the Tukey's multiple comparison tests were employed for statistical analysis. All tests were performed as two-sided test and a p value of < 0.05 was accepted as significant.