PD-L1/PD-L2-expressing B-1 cells inhibit alloreactive T cells in mice

B cells constitute a complex system of antigen-presenting cells (APCs) and exist as distinct subsets that differ in their lineage affiliation, surface molecule expression, and biological function, thus potentially regulating the immune response. In this study, we investigated the immune-regulatory roles of murine B cell subsets as regulatory APCs targeting alloreactive T cells. Either splenic B cells, peritoneal cavity (PerC) B cells, or non-B cells from Balb/c mice were intravenously injected into B6 mice. Serum levels of anti-Balb/c antibodies in the recipients of PerC B cells were significantly lower than those in the recipients of splenic B cells and PerC non-B cells, as determined over a 4-week period after the injection. Mixed-lymphocyte reaction (MLR) assays using splenocytes from the B6 mice at 2 weeks after the injection revealed the significantly reduced anti-Balb/c T cell-responses in the recipients of PerC B cells, as compared to those in the recipients of splenic B cells or untreated control mice. Since PerC B cells contained MHC class II+ CD80+ CD86+ PD-L1+ PD-L2+ cells among the CD5+ B-1a cell subset, PerC B cells from Balb/c mice were pre-incubated with anti-PD-L1/PD-L2 mAbs prior to injection. This treatment abrogated their immune-regulatory effects on anti-Balb/c T cells in the MLR assays. In addition, the inoculation with Balb/c PerC B cells significantly prolonged the survival of subsequently grafted Balb/c hearts in B6 mouse recipients, whereas that with SPL B cells did not. These findings indicate that the PerC B cells, including PD-L1/PD-L2 B-1a cells, may suppress T cells responding to allostimulation, and thus may be optimal for donor lymphocyte injection.


Introduction
T-cell activation during sensitization or rejection in organ transplant recipients first needs antigen presentation to T cells by antigen-presenting cells (APCs), such as macrophages, dendritic cells, and B cells. B cells are well known to play a role in allostimulation of T cells [1][2][3][4]. B cells not only constitute a complex system of APCs but also exist as distinct subsets that differ in their lineage affiliation, surface molecule expression, and biological function, thus potentially a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 regulating the immune responses. B cells play a central role in humoral immunity, but specific B cell subsets can also regulate T cell responses to foreign and self-antigens. These B cells are called regulatory B cells (Breg), and different types have been reported to date [5][6][7][8][9][10][11][12][13][14][15][16].
As one part of Breg cells, B cells were found to express programmed death-ligand (PD-L) 1 and PD-L2, which are one of the key factors responsible for inhibitory T cell signaling, providing immune homeostasis and mediating the mechanisms of tolerance [17][18][19][20], PD-L1 is constitutively expressed by various immune cells and nonimmune cells, whereas PD-L2 expression is expressed only by APCs such as macrophages and dendritic cells [21]. In addition, constitutive expression of PD-L2 together with PD-L1 is observed on 50-70% of mouse peritoneal CD5 + B-1a cells, a unique B cell subset, that act as efficient APCs [17]. Although the molecular and cellular functions of PD-L1 and PD-L2 in these B-1a cells currently remain unclear, these molecules per se are known to have a physiological role in regulating autologous T cell-immune responses in self-immunity by engaging PD-1.
This knowledge prompted us to investigate if the unique B cell subset that expresses PD-L1 and PD-L2 also inhibits alloimmune T cell responses. The present study has demonstrates that the infusion of allogeneic PD-L1 + /PD-L2 + B-1 cells suppresses T cells responding to cognate allostimulation. Thus, the present study provides a novel approach to induce T-cell hyporesponsiveness against transplanted allografts by infusing donor specific B-1 cells.

Materials and methods
Mice C57BL/6 (B6; H-2D b ) and Balb/c (H-2D d ) mice were purchased from CLEA Japan (Tokyo, Japan). All the mice were housed in the animal facility of Hiroshima University, in a pathogen-free, micro-isolated environment and were used at an age of 8-12 weeks. All experiments were approved by the Institutional Review Board of Hiroshima University and were handled and conducted according to the guidelines of the National Institutes of Health (publication no.86-23, revised 1996). Animal welfare was carefully ensured constantly by experienced operators every day. Mice were euthanized by cervical dislocation after inhalation of isofluran, when indicated. All efforts were made to minimize suffering of the animals along all the duration of their life and during the sacrifice.

Cell preparation
Mononuclear cell suspensions from the spleen (SPL), peritoneal cavity (PerC), liver, bone marrow (BM), cervical lymph node (CLN), and peripheral blood (PB) were prepared. B cells from the mouse SPL or PerC were purified by performing positive selection with CD19 MicroBeads (Miltenyi Biotec, San Diego, CA, USA) and an autoMACS Separator (Miltenyi Biotec), according to the manufacturer's instructions. Non-B cells from the PerC were purified by negative selection at the same time. Purity of the isolated B and non-B cells was assessed by flow cytometry (FCM) and was typically found to be >95%.

Adoptive B cell transfer
B6 recipient mice received a total of 10 × 10 6 of Balb/c B cells isolated from SPL or PerC, or non-B cells from PerC in a volume of 200 μL via tail vein injection. Splenic T cell response was analyzed 14 days after the adoptive transfer by performing CFSE-MLR assay as described below. To assess the allo-Ab response in recipient mice, blood samples of recipient mice were obtained before injecting the cells and weekly thereafter for 4 weeks. Serum was isolated and was frozen for allo-Ab analysis by FCM assay.

CFSE-MLR assay
Splenocytes of B6 recipients at 2 weeks after adoptive transfer were labeled with 5 μM CFSE (Molecular Probes: Eugene, OR) as described previously [22,23], and were resuspended in culture medium for use as responder cells. Fractions of whole splenocytes were prepared from B6 mice that received SPL B cells, PerC B cells, or PerC non-B cells. After each fraction had been irradiated (30 Gy) for use as stimulator cells, 4 × 10 6 stimulator cells and 4 × 10 6 responder cells were cocultured in 24-well, flat-bottom plates at 37˚C in a 5% CO 2 incubator in the dark. After 4 days of incubation, the responder cells were harvested and analyzed by FCM. CD4 + and CD8 + T cells were selected through gating and analyzed for intensity of CFSE fluorescence. Stimulation indexes (SI) and precursor frequencies (PF) were calculated as described previously [22].

Heart transplantation
Two weeks before heart transplantation, recipient B6 mice were intravenously injected with 10 × 10 6 SPL or PerC CD19 + B cells from Balb/c mice. Prior to surgery, mice were anesthetized by intraperitoneal injection of xylazine (5 mg/kg body weight) and ketamine (100 mg/kg body weight). Cervical heterotopic heart transplantation from Balb/c mice into naïve B6 mice or B cell injected B6 mice was performed using a modified cuff technique [24]. Briefly, the right external jugular vein and the right common carotid artery were dissected free and were fixed to the appropriate cuffs composed of polyethylene tubes. For anastomoses, the aorta and the main pulmonary artery of the obtained donor heart were drawn over the external jugular vein, respectively. Graft ischemic time for the transplanted hearts was < 30 min. The function of the grafts was monitored daily through inspection and palpation. Rejection was determined based on the cessation of heart beat and was confirmed by performing histological analysis. Mice were daily monitored for weight and well being. If there was evidence of more than 15% weight loss compared to weight at surgery-date, no food intake, breathing difficulties, apathy, or a hunched position as well as when the heart was rejected, animals were euthanized. Two mice were excluded from the evaluation of graft survival because they suddenly died of unknown cause with beating allograft heart.

Statistical analysis
Data are presented as mean ± standard error of mean (SEM). Statistical analysis was performed using Student's t test (GraphPad Prism, San Diego, USA). Survival of the heart allografts was calculated using Kaplan-Meier/log-rank test. A p-value of < 0.05 was considered statistically significant.

Results
Naïve B-1a cells from PerC highly expressed PD-L1, PD-L2, CD80, CD86, and MHC class II To search for B cells having the potential to induce hyporesponsiveness of T cells to allostimulation, we investigated the phenotypic characteristics of naïve B cells from SPL, PerC, liver, BM, PB, and CLN in an anatomical screening for immune-regulatory and co-stimulatory membrane markers. Expressions of MHC class II, CD80, and CD86-surface molecules necessary for efficient antigen presentation to T cells-and the immune-regulatory molecules anti-PD-L1, PD-L2, FasL, and TRAIL were analyzed. In Balb/c mice, naïve B cells from PerC showed remarkably higher levels of PD-L1, PD-L2, CD80, and CD86 compared to those in other organs, while FasL and TRAIL were rarely expressed on B cells from any organs ( Fig  1A). Mean percentage of PD-L1, PD-L2, CD80, and CD86 expression on B cells from PerC was statistically higher than that from any other organ (Fig 1B. FasL and TRAIL on B cells from each organ showed generally low frequency and no difference was detected in all 6 organs. MHC class II was expressed on almost all B cells. Similar results were obtained for B cells isolated from B6 mice (S1 Fig). We further analyzed the expression of immune-regulatory molecules on the various B cell subsets isolated from the SPL, PerC, and liver. SPL cells were stained with combinations of mAbs directed against IgM and CD21 with a variety of immune-regulatory molecules. Since  [25,26], exclusively expressed a high level of PD -L1, but barely expressed PD-L2. The expression of those PD-L1/L2 was limited on CD21 int IgM int FO B cells, which were positioned adjacent to T cell areas, facilitating their response to T cell-dependent antigens and promoting germinal center formation [26], and CD21 -/low IgM high newly formed B-0 cells, which had recently immigrated from the BM into the SPL.

Intravenous injection of allogeneic PerC B cells prevented sensitization
To investigate the immunogenicity of SPL and PerC B cells, 10 × 10 6 B cells from SPL, PerC, or non-B cells isolated from PerC of naïve Balb/c mice were injected to B6 mice through the tail vein and production of anti-Balb/c Abs in the sera of recipient B6 mice was serially evaluated. In B6 mice receiving SPL B cells, serum levels of anti-Balb/c IgM were transiently elevated 1 week after the inoculation and declined thereafter (Fig 3A; left). However, the sera of B6 mice receiving PerC B or non-B cells did not show similar increase in anti-Balb/c IgM levels. In B6 mice receiving SPL B cells or PerC non-B cells, serum levels of anti-Balb/c IgG increased over time and reached a plateau 2 weeks after the inoculation (Fig 3A; right). However, this increase in anti-Balb/c IgG level was not observed in B6 mice receiving PerC B cells. Subclass analyses of anti-Balb/c IgGs showed that IgG2a/b was predominant in the recipients of PerC non-B cells, whereas no particular uneven class-switching was observed in recipients of SPL or PerC B cells (Fig 3B). Thus, intravenous injection of allogeneic PerC B cells prevents significant sensitization.

Intravenous injection of allogeneic PerC B cells inhibited T cell-immune responses to cognate allostimulation
Next, we investigated the immune-regulatory effects of SPL and PerC B cells on allogeneic T cells by performing the CFSE-MLR assays. Two weeks after the injection, splenocytes of B6 mice were harvested and assayed by performing the CFSE-MLR. The SI of anti-Balb/c CD4 + T cells was significantly lower in mice receiving PerC B cells than in mice receiving SPL B cells

T-cell allo-response after the intravenous injection of allogeneic B cells, as evaluated by performing CFSE-MLR assay. SPL or PerC B cells, or PerC non-B cells from
Balb/c mice were isolated by magnetic sorting with CD19 microbeads, and were injected into B6 mice through the tail vein. T-cell alloreactivity was determined by performing CFSE-MLR assay 2 weeks after the adoptive transfer. The CFSE-MLR assay was performed using B6 mice injected with Balb/c SPL B cells (n = 6), Balb/c PerC B cells (n = 6), and Balb/c PerC non-B cells (n = 6); control B6 mice were not injected with any cells (n = 7), as responders. SIs and PFs were calculated as described previously [22]. (A) CFSE-PerC B cells pre-incubated with blocking anti-PD-L1/PD-L2 mAbs abrogated immune-regulatory effects on alloreactive T cells To address the functional significance of PD-L1 and PD-L2 expression on PerC B cells in the hyporesponsiveness of alloreactive T cells, we performed the CFSE-MLR assay by using the splenocytes isolated from B6 mice injected with B cells pre-incubated with blocking anti-PD-L1/PD-L2 mAbs at 2 weeks after inoculation. Prior to inoculation, blockade of PD-L1/ PD-L2 expression was confirmed by reanalysis with FCM ( S3 Fig). The SI of anti-Balb/c CD4 + and CD8 + T cells was significantly higher in mice receiving PerC B cells preincubated with anti-PD-L1 mAb than in mice receiving PerC B cells preincubated with isotype-matched control Abs and in uninoculated control mice (Fig 5). Similar results were obtained for mice receiving PerC B cells preincubated with both anti-PD-L1 and anti-PD-L2 mAbs, but any additive effects of the combination of these mAbs were noted. The SI of anti-Balb/c CD4 + and CD8 + T cells was the highest in mice receiving PerC B cells preincubated with anti-PD-L2 mAb, although the difference in SIs between mice receiving PerC B cells preincubated with anti-PD-L1 mAb and those receiving PerC B cells preincubated with anti-PD-L2 mAb was not statistically significant. Consistently, the PFs of anti-Balb/c CD4 + and CD8 + T cells was higher in mice receiving PerC B cells preincubated with anti-PD-L1 and/or anti-PD-L2 mAbs than in mice receiving PerC B cells preincubated with isotype-matched control Abs. This may be because expression of both PD-L1 and PD-L2 on allogeneic PerC B cells is required for inhibiting T cell-responses to the cognate allostimulation.

Inoculation of allogeneic PerC B cells significantly prolonged the survival of subsequently grafted cognate allogeneic hearts
To determine whether allogeneic PerC B cells exerted immune-regulatory effects on subsequently transplanted cognate organs, allogeneic hearts of Balb/c mice were transplanted into B6 mice intravenously injected with Balb/c SPL or PerC B cells (10 x 10 6 cells/mouse) at two weeks injection. Survival curves of the grafted hearts are shown in Fig 6. Untreated B6 mice rejected the hearts of Balb/c within 1-week. B6 mice injected with Balb/c PerC B cells showed significantly prolonged survival of the hearts of Balb/c mice. However, the allografted hearts were eventually rejected within 4-weeks after grafting in these mice. In contrast, B6 mice injected with SPL B cells did not show the prolonged survival of the allografted hearts. Thus, the inoculation of allogeneic PerC B cells could prolong the survival of subsequently grafted cognate allogeneic hearts.

Discussion
Numerous reports have demonstrated that pretransplant exposure to allogeneic donor antigens through donor leukocyte injection (DLI) or blood transfusion, either alone or in combination with other treatments, can facilitate donor-specific hyporesponsiveness. Some studies have investigated the subsets of allogeneic donor cells that may be important for inhibiting anti-donor immune responses, but the results remain controversial [37][38][39][40][41][42]. A previous study demonstrated that a single injection of naïve donor SPL B cells prior to transplantation resulted in significantly prolonged donor-specific skin allograft survival compared to that of labeled CD4 + (upper row) and CD8 + (lower row) T cell division in the MLR assay was quantified by performing FCM analysis. Representative plots and images are shown. CFSE-labeled splenocytes from B6 mice were used as responders, and irradiated (30 Gy) splenocytes from Balb/c mice were used as stimulators. (B) The SIs and PFs of splenocytes from B6 mice in the CFSE-MLR are shown. *p < 0.05. Data are presented as mean ± SEM, and are representative of two independent experiments with six to seven mice per group.  Isotype-matched control IgGs were used instead of anti-PD-L1 or anti-PD-L2 mAb. Rat IgG2b was used instead of anti-PD-L1 mAb, and Rat IgG2a was used instead of anti-PD-L2 mAb. Mice in the control group did not receive adoptive transfer of cells or mAbs. T-cell alloreactivity was determined by performing the CFSE-MLR assay 2 weeks after the intravenous injection. Upper row shows the mean SIs of CD4 + and CD8 + T cells, and lower row shows the mean PFs of CD4 + and CD8 + T cells; *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.00001. Data are presented as mean ± SEM, and are representative of two independent experiments with four to five mice per group.
https://doi.org/10.1371/journal.pone.0178765.g005 PD-L1/PD-L2 B cells inhibit alloreactive T cells PD-L1 and PD-L2 coexpression on B-1a cells potentially has important regulatory functions in alloimmune reponses. Since PD-L1 + PD-L2 + B-1a cells share most of the surface markers with PD-L1 + PD-L2 -B-1a cells, it is very difficult to sort PD-L2 + B-1a cells without using anti-PD-L2 mAb, which blocks and interferes with PD-L2 physiological function. To avoid this problem, an endogenous indicator system for tracking and sorting PD-L1 + PD-L2 + B-1a cells is highly desirable for applying this particular subset of B cells to DLI prior to allografting.
Even without sorting PD-L1 + PD-L2 + B-1a cells, the injection of unfractionated PerC B cells resulted in the avoidance of allo-sensitization, and even induced T cell-hyporesponsiveness in response to allostimulation. Inoculation of allogeneic PerC B cells significantly prolonged the survival of subsequently grafted cognate allogeneic hearts; however, the heart allografts were eventually rejected. One possible explanation for the limited prolongation of allograft survival by allogeneic PerC B cells is that these cells may exert efficient suppressive effects on T cells with direct allospecificity, but not on T cells with indirect allospecificity, which are also involved in rejecting allografted organs. To address this issue, it is necessary to determine the mechanisms underlying the role of PD-L1 and PD-L2 expressed on allogeneic PerC B cells in inhibiting T cells that directly recognize alloantigens.
In conclusion, the allogeneic PerC B cells, including the MHC class II + CD80 + CD86 + PD-L1 + PD-L2 + B-1a cell subset, have the potential to suppress T cells responding to the cognate allostimulation.