Tick saliva-induced programmed death-1 and PD-ligand 1 and its related host immunosuppression

The tick Rhipicephalus microplus is a harmful parasite of cattle that causes considerable economic losses to the cattle breeding industry. Although R. microplus saliva (Rm-saliva) contains several immunosuppressants, any association between Rm-saliva and the expression of immunoinhibitory molecules, such as programmed death (PD)-1 and PD-ligand 1 (PD-L1), has not been described. In this study, flow cytometric analyses revealed that Rm-saliva upregulated PD-1 expression in T cells and PD-L1 expression in CD14+ and CD11c+ cells in cattle. Additionally, Rm-saliva decreased CD69 expression in T cells and Th1 cytokine production from peripheral blood mononuclear cells. Furthermore, PD-L1 blockade increased IFN-γ production in the presence of Rm-saliva, suggesting that Rm-saliva suppresses Th1 responses via the PD-1/PD-L1 pathway. To reveal the upregulation mechanism of PD-1/PD-L1 by Rm-saliva, we analyzed the function of prostaglandin E2 (PGE2), which is known as an inducer of PD-L1 expression, in Rm-saliva. We found that Rm-saliva contained a high concentration of PGE2, and PGE2 treatment induced PD-L1 expression in CD14+ cells in vitro. Immunohistochemical analyses revealed that PGE2 and PD-L1 expression was upregulated in tick-attached skin in cattle. These data suggest that PGE2 in Rm-saliva has the potential to induce the expression of immunoinhibitory molecules in host immune cells.

The Rhipicephalus microplus is a one-host tick that feeds on cattle. It is one of the most harmful parasites of cattle in subtropical areas of the world and causes significant economic losses to the cattle industry 1 . The tick causes many deleterious effects including loss of blood, slowed weight gain, and reduced milk production 1 . Additionally, the tick transmits Babesia spp. and Anaplasma spp. to cattle 2,3 . Although chemical acaricides are used to control tick infestation, the continuous use of these acaricides causes many adverse side effects including the selection of resistant tick populations to these chemicals as well as food and environmental contamination 4,5 . Thus, the development of an alternative control strategy, such as a tick vaccine 6 , is strongly needed.
Tick saliva is important to obtain a blood meal and facilitate pathogen transmission to the vertebrate host. Tick saliva contains bioactive factors including a large variety of immunosuppressants 7 , such as cystatins 8 , serpins 9 , metalloproteinases 10 , and lipocalins 11 . For example, salivary cystatins, such as sialostatin L and sialostatin L2, suppress host immune responses, leading to the enhancement of tick-borne pathogen transmission 8,12,13 . Interestingly, Kotsyfakis et al. have reported that guinea pig vaccination against sialostatin L2 leads to a decreased feeding ability and increased rejection rate of Ixodes scapularis nymphs 14 . Hence, salivary immunosuppressants are a potential target for tick control. Previous studies on R. microplus have described that host exposure to salivary gland extracts regulates immune responses such as macrophage activation 15 . Furthermore, isolated salivary proteins modulate the immune response 9,13,[16][17][18] . However, the detailed mechanism of immunomodulation in cattle caused by R. microplus saliva (Rm-saliva) is not well understood.
Here, we analyzed the expression of PD-1 and PD-L1 in bovine immune cells in the presence of Rm-saliva. We then examined PGE 2 concentrations in Rm-saliva by enzyme-linked immunosorbent assay (ELISA) and ultra-performance liquid chromatography (UPLC)-mass spectrometry (MS)/MS system, and we examined the association of PGE 2 in Rm-saliva with PD-L1 expression by in vitro culture and immunohistochemistry.

Results
Upregulation of PD-1/PD-L1 expression by Rm-saliva. To examine whether Rm-saliva induces PD-1 and PD-L1 expression in cattle, peripheral blood mononuclear cells (PBMCs) were cultured with Rmsaliva. PD-1 expression in CD4 + and CD8 + T cells was increased by Rm-saliva ( Fig. 1a-c). Additionally, PD-L1 expression levels in CD14 + and CD11c + cells were also increased by Rm-saliva ( Fig. 1d-f). These results suggest that Rm-saliva is an inducer of immunoinhibitory molecules, such as PD-1 and PD-L1, in bovine immune cells.

Inhibition of Th1 responses by Rm-saliva via PD-1/PD-L1 pathway.
In cattle, the PD-1/PD-L1 pathway is involved in the inhibition of Th1 immune responses, such as IFN-γ production 37 . Therefore, to analyze whether Rm-saliva inhibits bovine Th1 responses, we cultured bovine PBMCs with Rm-saliva in the presence of T-cell stimulation. Treatment with Rm-saliva downregulated the expression of CD69, an activation marker, in CD4 + and CD8 + cells (Fig. 2a-c). Furthermore, treatment with Rm-saliva decreased IFN-γ and TNF-α production from bovine PBMCs (Fig. 2d,e). Interestingly, the blockade of the PD-1/PD-L1 pathway using anti-PD-L1 antibodies (Abs) enhanced IFN-γ production in the presence of Rm-saliva (Fig. 2f), suggesting that Rm-saliva inhibits bovine Th1 responses, at least in part, via the PD-1/PD-L1 pathway.

Induction of PD-L1 expression via PGE 2 in Rm-saliva.
We then demonstrated the upregulation mechanism of PD-1/PD-L1 expression by Rm-saliva. The concentration of PGE 2 in Rm-saliva determined by ELISA was approximately 200 ng/mL (Fig. 3a). In addition, Rm-saliva was also qualified and quantified by UPLC-MS/MS system. A clear single peak was confirmed in the transition of m/z 351 > 271 (qualifier) and m/z 351 > 315 (quantifier) at the same retention time to the PGE 2 standard (7.1 min) (Fig. 3b). The estimated concentration of PGE 2 from Rm-saliva using an external calibration curve was 330 ng/mL. We previously have reported that PGE 2 induces PD-L1 expression in bovine immune cells 21,33 . In the present study, bovine CD14 + cells were cultured with 100 ng/mL of PGE 2 . The concentration of PGE 2 in the cell culture was similar to that in Rm-saliva. As shown in Fig. 3c, PD-L1 expression in CD14 + cells was significantly upregulated by PGE 2 treatment in vitro (Fig. 3c). In addition, immunohistochemical staining was performed to examine PGE 2 and PD-L1 expression in tick-attached and tick-nonattached sites. We observed the increase in both PGE 2 positive and PD-L1 positive immune cells in tick-attached site, although statistical analysis was not done because of the limited number of samples that were evaluated. (Fig. 4a-d). Thus, PGE 2 in Rm-saliva may be associated with the upregulation of PD-1/PD-L1 expression.

Discussion
The immunomodulatory properties of tick saliva have been investigated in several in vitro and in vivo studies. The saliva compounds such as serpins 9,38 , cystatins 39,40 , PGE 2 26,27 , and other molecules 41,42 can interfere with different host immune cells to modulate the host immune response and to help tick feeding.
In the present study, we demonstrated the existence of PGE 2 in Rm-saliva. PGE 2 concentration in Rm-saliva (200 ng/mL) was much higher (approximately 2740-fold greater) than that in the plasma from healthy cattle (73 pg/mL) as described in a previous paper 43 . We previously showed that PGE 2 treatment in vitro suppressed Th1 cytokine production and T-cell proliferation in cattle 21 . Thus, salivary PGE 2 is identified as one of the immunosuppressants in R. microplus. PGE 2 has immunosuppressive effects not only on T cells, but also on other immune cells, such as macrophages and DCs 20 . PGE 2 in I. scapularis saliva inhibits DC maturation and function 26 , and PGE 2 in Dermacentor variabilis saliva regulates macrophage activity 27 . Additionally, Rm-saliva is involved in immunoregulation of bovine macrophages 15 . Furthermore, the data obtained from the skin biopsies of tick-resistant and tick-susceptible cattle have shown that innate and adaptive immune pathways are activated in tick-resistant animals and therefore could be modulated by tick saliva to the benefit of the feeding ticks 44 . Together, these findings suggest the contribution of the PGE 2 in tick saliva to the parasitism by modulating both the innate immune responses, such as the activities of DCs and macrophages [25][26][27][28] , and adaptive immune responses (present study) in vitro and in vivo. However, additional experiments are necessary to examine the association of PGE 2 in Rm-saliva with other bovine immune cells. www.nature.com/scientificreports/ The PD-1/PD-L1 pathway contributes to the process of T-cell exhaustion 34 . A previous study has shown that a salivary gland protein from R. appendiculatus upregulates PD-L1 expression in human DCs 45 . Carvalho-Costa et al. also have shown that Amblyomma cajennense tick saliva increases PD-L1 expression in murine bone marrow-derived DCs 46 . However, to date, no report demonstrates the association of tick saliva with immunosuppression caused by the PD-1/PD-L1 pathway. In the present study, we found that Rm-saliva induced PD-1 expression in T cells and PD-L1 expression in CD14 + and CD11c + cells in vitro. The blockade of PD-1/PD-L1 interaction using an anti-PD-L1 Ab ameliorated the Rm-saliva-induced Th1 suppression. To the best of our knowledge, this is the first study that shows that tick saliva regulates host immune responses, at least in part, via PD-1/PD-L1 upregulation. Our previous studies have demonstrated that PGE 2 is one of the inducers of PD-L1 expression in bovine CD14 + cells 21,33 . In the present study, we found that Rm-saliva contained a high www.nature.com/scientificreports/ concentration of PGE 2 . In addition, the treatment with commercially available PGE 2 increased PD-L1 expression in CD14 + cells, suggesting that Rm-saliva upregulates PD-L1 expression via PGE 2 . To reveal the direct relationship between PD-L1 upregulation and PGE 2 derived from Rm-saliva, we attempted to purify PGE 2 from Rm-saliva. However, PD-L1 upregulation by PGE 2 derived from Rm-saliva was not confirmed in the present study, presumably due to contamination by an acidic solvent during the purification process (data not shown).
In addition, we also attempted to elucidate the direct relationship by using the blockers of PGE 2 receptors. However, it was not able to demonstrate using the in vitro cultures because treatment with PGE 2 receptor blockers in vitro increased PD-L1 expression presumably due to the induction of cytokine production such as IFN-γ and TNF-α (data not shown). Therefore, to establish an appropriate evaluation system is required, and future studies will examine in details the biological effect of PD-L1 upregulation induced by PGE 2 from Rm-saliva in the host-parasite relationship. Further, recent studies in human research have shown that PGE 2 also induces the expression of other immunoinhibitory molecules, such as PD-1 and T-cell immunoglobulin and mucin domain 3 32,47 . Our data demonstrated that Rm-saliva induced PD-1 expression in bovine T cells in vitro. However, it is still unclear whether PGE 2 in tick saliva is involved in the upregulation of other immunoinhibitory molecules. Additional experiments to address the issue should be conducted in the future research. Previous studies have shown the correlations between the expression of immunoinhibitory receptors in bovine T cells and Th1 responses, such as IFN-γ production. A study showed that PD-1 expression in CD4 + and CD8 + T cells was upregulated in cattle infected with Mycoplasma bovis, and negative correlations were observed between PD-1 expression in T cells and IFN-γ production from PBMCs 48 . Similarly, another study showed that the expression of lymphocyte activation gene-3 (LAG-3), an immunoinhibitory molecule expressed on T cells, was upregulated in T cells of bovine leukemia virus-infected cattle, and the expression levels of LAG-3 in CD4 + T cells were negatively correlated with the expression of IFN-γ gene 49 . Here, by treatment with Rm-saliva, the percentage of PD-1 + CD4 + and PD-1 + CD8 + T cells was increased significantly but not dramatically. However, based on these previous studies, the differences observed in Fig. 1b,c could have influence in the activity of T cells.
Tick saliva potentiates infection with a variety of agents that cause tick-borne diseases 50 , but detailed mechanisms of heightened infectivity have not been fully elucidated. The modulation of host immune systems is speculated as a critical element 51,52 . Th1 responses help to protect the host against tick-borne pathogens, including Anaplasma marginale, which is transmitted to cattle by R. microplus 53,54 . In addition, the PD-1/PD-L1 pathway is used as an immune evasion mechanism of A. marginale from host defense 55 . Interestingly, PD-L1 has been used in a chimeric vaccine against Babesia microti infections in which the extracellular domain of PD-L1 was fused to the C-terminus of B. microti-profilin 56 . Our results demonstrated that Rm-saliva regulates immune responses, especially Th1 responses, through PD-1/PD-L1 upregulation, suggesting that Rm-saliva promotes the transmission of tick-borne pathogens via PD-1/PD-L1-mediated Th1 suppression.
In conclusion, the results of the present study revealed the association of Rm-saliva with PD-1/PD-L1 upregulation in host immune cells. Rm-saliva regulated Th1 immune responses by the PD-1/PD-L1 pathway. Additionally, Rm-saliva contained a high concentration of PGE 2 , and treatment with PGE 2 in vitro induced PD-L1 expression in bovine CD14 + cells. Furthermore, immunohistochemical analyses showed that the expression of both PGE 2 and PD-L1 was upregulated in the tick-attached site, suggesting that PGE 2 in Rm-saliva may increase PD-L1 expression. Further studies will examine the role of Rm-saliva-induced PD-1/PD-L1 upregulation on the transmission of tick-borne pathogens.

Methods
Ticks and saliva collection. Rhipicephalus microplus ticks (Porto Alegre strain, free from Babesia spp. and Anaplasma spp.) were obtained from a laboratory colony maintained as previously described 57 . Hereford calves were acquired from a tick-free area and infested with 10-day-old R. microplus larvae. Fully engorged female ticks were obtained after the spontaneous detachment from the calves. Ticks were rinsed with sterile distilled water, and salivation was induced by dorsal injection of 5 µL of 2% pilocarpine hydrochloride (Sigma-Aldrich, St. Louis, MO, USA) in ethanol (Merck KGaA, Darmstadt, Germany) 58,59 . Rm-saliva was collected using a pipette tip from ticks maintained at 28 °C in a humid chamber for approximately 3 h. Collected Rm-saliva was stored at − 80 °C upon use in the experiments. Animals used in these experiments were housed in Faculdade de Veterinária, Universidade Federal do Rio Grande do Sul (UFRGS), Brazil. Experiments were conducted considering ethical and methodological guidelines in agreement with the Norms by the Animal Experimentation Ethics Committee of the UFRGS. The protocols were approved by the Comissão de Ética no Uso de Animais (CEUA)-UFRGS. These experiments were carried out in compliance with the ARRIVE guidelines (http://www.nc3rs .org. uk/page.asp?id=1357).
PBMC culture. PBMCs were cultured with 1 µL of Rm-saliva in 100 µL of culture medium for 24 or 72 h.
Cultures were stimulated with or without 1 μg/mL of anti-CD3 mAb (MM1A; WSU Monoclonal Antibody Center) and 1 μg/mL of anti-CD28 mAb (CC220; Bio-Rad, Hercules, CA, USA). After 24 h of incubation with the stimulation, PBMCs were collected, and CD69 expression was examined by flow cytometry as described below. After 72 h of incubation with the stimulation, culture supernatants were collected and the concentrations of IFN-γ and TNF-α were measured by ELISA as described below. After 24 h of incubation without the stimulation, the cells were collected, and the expression of PD-1 and PD-L1 was measured by flow cytometry as described below.
To examine whether the inhibition of the PD-1/PD-L1 pathway using a specific antibody (Ab) rescues IFN-γ production in the presence of Rm-saliva, PBMCs were cultured with 1 µL of Rm-saliva and 10 µg/mL of anti-PD-L1 Ab (Boch4G12) 60 in the presence of 1 μg/mL of anti-CD3 mAb and 1 μg/mL of anti-CD28 mAb in 100 µL of culture medium for 72 h. Bovine IgG (Sigma-Aldrich) was used as a negative control. After incubation, culture supernatants were collected, and IFN-γ concentrations were measured by ELISA as described below.
CD14 + cell culture. Isolated CD14 + cells were cultured with 100 ng/mL of PGE 2 (Cayman Chemical, Ann Arbor, MI, USA) for 24 h. Dimethyl sulfoxide (DMSO; Nacalai Tesque, Kyoto, Japan) was used as a vehicle control. After incubation, cells were collected and PD-L1 expression was measured by flow cytometry as described in the following section. ELISA. IFN-γ and TNF-α concentrations in culture supernatants collected from PBMC cultures were determined by Bovine IFN-γ ELISA Development Kit (Mabtech, Nacka Strand, Sweden) and Bovine TNF-α Do-It-Yourself ELISA (Kingfisher Biotech), respectively, according to the manufacturers' instructions. The PGE 2 concentration in Rm-saliva was measured by Prostaglandin E 2 Express ELISA Kit (Cayman Chemical) per the manufacturer's instruction. PGE 2 purification. PGE 2 purification from Rm-saliva was conducted by the combination with solid phase extraction (SPE) and the fractionation by using high-performance liquid chromatography (HPLC). The SPE method was performed as previously described 63 . Briefly, 2 mL of a water/acetic acid mixture (99:1) was added to 0.5 mL of Rm-saliva. The sample was then loaded onto an SPE column (Bond Elut LRC-C18, Agilent Technologies, Santa Clara, CA, USA) conditioned in advance with 2 mL of ethyl acetate, methanol, and the water/ acetic acid mixture (99:1). The column was washed with 2 mL of water and hexane. Next, the Bond Elut DEA (Agilent Technologies) was conditioned with 2 mL of ethyl acetate, and the Bond Elut LRC-C18 described above was connected in series. PGE 2 was eluted from Bond Elut LRC-C18 with 4 mL of ethyl acetate, and the Bond Elut LRC-C18 was removed. Then, the Bond Elut DEA was washed with 2 mL of the water/acetic acid mixture (99:1). Next, absolute NEXUS (Agilent Technologies) was conditioned with 2 mL of ethyl acetate, methanol, and the water/acetic acid mixture (99:1), and the Bond Elut DEA holding PGE 2 was connected in series. Elution from Bond Elut DEA was performed with 4 mL of the water/acetic acid mixture (99:1)/methanol (3:2), and this eluate was directly retained in absolute NEXUS. After removing Bond Elut DEA, the absolute NEXUS was centrifuged at about 1500×g (3000 rpm) for 10 min for dehydration and eluted with 4 mL of ethyl acetate. The ethyl acetate was evaporated under a gentle stream of nitrogen gas, and the residue was re-dissolved in 200 µL of an ammonium (pH 4.5)/methanol mixture (3:2). www.nature.com/scientificreports/ Rm-saliva purified by SPE was further purified by two-step fractionation by HPLC. HPLC (LC 20 series, Shimadzu, Kyoto, Japan) coupled with the fraction collector (FRC-10A, Shimadzu) was used for fractionation. The first step was conducted by size exclusion chromatography using the BioSEC-3 column (3 µm, 100 Å, 4.6 mm × 300 mm, Agilent Technologies). The mobile phases A and B were distilled water and 0.1% formic acid containing acetonitrile (5/95, isocratic mode), and the flow rate was 0.5 mL/min. The SPE purified sample (40 µL) was injected into the HPLC, fractionated every minute, and divided into a total of 10 fractions. The 8-9 min fraction, in which PGE 2 was eluted, proceeded to the next step. The second step was conducted by reverse-phase chromatography using Cadentza CD-C18 (2.7 µm, 3 × 150 mm; Imtact, Kyoto, Japan). Aliquots of 20 µL of the sample were injected and separated using a linear gradient of distilled water and acetonitrile at a flow rate of 0.3 mL/min. The gradient was programmed as follows: t = 0-2 min, 10% acetonitrile; t = 28 min, 60% acetonitrile; t = 32 min, 95% acetonitrile; and t = 32-35 min, 95% acetonitrile. A fraction was taken every 1 min, and PGE 2 was eluted in the 23-24 min fraction. The final fraction was dried under a gentle stream of nitrogen gas and re-dissolved in 10 µL of acetonitrile.
Quantification of PGE 2 by UPLC-MS/MS. The PGE 2 concentration in Rm-saliva was quantified by the Agilent 6495B UPLC-MS/MS system (Agilent Technologies). The separation of PGE 2 was performed on a Poreshell EC-C18 column (2.7 µm, 3 mm, 100 mm; Agilent Technologies) at a flow rate of 1.0 mL/min at 60 °C. Mobile phase A was distilled water, and mobile phase B was methanol containing 0.1% formic acid. The following gradient program was applied: t = 0-1 min, 5% B; t = 10 min, 95% B; t = 11 min, 95% B. The injection volume was 1 µL. Detection was performed in electrospray ionization-positive mode. The triple quadrupole was operated in multiple-reaction monitoring mode by monitoring a quantifier (m/z 351 > 271) and qualifier (m/z 351 > 315) transition for PGE 2 . Quantitation was achieved by external calibration using a standard curve built from calibration points at 2.5, 5, 10, 25, 50, and 100 ng/mL. Immunohistochemical staining. Immunohistochemical assays were performed as previously described with slight modifications 21,33 . Briefly, skin biopsies of a tick bite site (partially engorged R. microplus female) and a sample of normal skin as a control were performed using a 6-mm biopsy punch, fixed in 10% buffered formalin, and processed routinely by embedding in paraffin wax. The skin sections were then immunohistochemically stained for PGE 2 and PD-L1 using anti-PGE 2 polyclonal Ab (ab2318, Abcam, Cambridge, England, UK) and anti-PD-L1 mAb (6C11-3A11, Rat IgG 2a ) 62 . Histological slides were scanned with Nano Zoomer 2.0-RS (Hamamatsu Photonics, Hamamatsu, Japan) and processed in QuPath ver 0.2.1. Scanned slides were opened as Brightfield (H-DAB) in QuPath, DAB and hematoxylin stains were adjusted using the Estimate Stain Vectors function. Immune cells were annotated based on their morphologies and locations to allow QuPath to automatically classify each cell type using the Create Detection Classifiers function.
Statistics. Statistical significance was analyzed using the Wilcoxon signed-rank test. A p value of < 0.05 was considered statistically significant.