CpG-Activated Regulatory B-Cell Progenitors Alleviate Murine Graft-Versus-Host-Disease

Abstract

Development of Graft Versus Host Disease (GVHD) represents a major impediment in allogeneic hematopoietic stem cell transplantation (HSCT). The observation that the presence of bone marrow and circulating hematogones correlated with reduced GVHD risks prompted us to evaluate whether B-cell progenitors, which provide protection in various autoimmune disease models following activation with the TLR-9 agonist CpG (CpG-proBs), could likewise reduce this allogeneic disorder. In a murine model of GVHD that recapitulates an initial phase of acute GVHD followed by a phase of chronic sclerodermatous GVHD, we found that CpG-proBs, adoptively transferred during the initial phase of disease, reduced the diarrhea score and mostly prevented cutaneous fibrosis. Progenitors migrated to the draining lymph nodes and to the skin where they mainly differentiated into follicular B cells. CpG activation and IFN-γ expression were required for the protective effect, which resulted in reduced CD4⁺ T-cell-derived production of critical cytokines such as TGF-β, IL-13 and IL-21. Adoptive transfer of CpG-proBs increased the T follicular regulatory to T follicular helper (Tfr/Tfh) ratio. Moreover, CpG-proBs privileged the accumulation of IL-10-positive CD8⁺ T cells, B cells and dendritic cells in the skin. However, CpG-proBs did not improve survival. Altogether, our findings support the notion that adoptively transferred CpG-proBs exert immunomodulating effect that alleviates symptoms of GVHD but require additional anti-inflammatory strategy to improve survival.

Keywords: Bregs: regulatory B cells; CpG-proBs; allogeneic stem cell transplantation (allo-SCT); cell therapy; fibrosis; graft-versus host disease; regulatory B-cell progenitors.

Figure 1

Effect of adoptively transferred CpG-proBs on GVHD symptoms. Balb/c recipients irradiated at 5.8 Gy on day-0, were reconstituted on day+1 with T- and B-cell depleted BM cells (5x10⁶ cells) and splenocytes (1 x 10⁶ cells) from C57BL/6J donors. CpG-proBs (7.5 x 10⁵cells) or proBs prepared from C57BL/6J donors and expanded in co-culture with OP-9 stromal cells were adoptively transferred on day+2, day+9 or day+23 post-irradiation (DPI) as indicated. Diarrhea, cutaneous scores and survival are shown over a period of 60-80 days. Results are expressed as means ± SEM. Adoptive transfer (or PBS injection in control GVHD mice) was performed on day+2 in GVHD control mice (N=30, black line), CpG-proB recipients (N=19, red line), proB recipients (N=10, blue line); on day+9, in GVHD controls (N=6, black line) and CpG-proB recipients (N=7, red line); on day+23, in GVHD controls (N=7, black line) and CpG-proB recipients (N=6, red line). Statistical analysis was performed with two-way ANOVA with Bonferroni post-tests for diarrhea score and cutaneous score and Kaplan-Meier estimates for survival; p values as indicated; ns=non significant.

Figure 2

Migration, differentiation and cytokine expression of CpG-proBs in GVHD mice. CpG-proBs, isolated from the BM of actin-GFP-KI C57BL/6J donors, were adoptively transferred on day+2 post-irradiation. (A) Gating FACS procedure of B220⁺ GFP⁺ cells, shown on day+15 in mesenteric lymph nodes (mLN), in controls with GVHD and in CpG-proB recipients, isotype antibody controls being used to define positivity. (B) The migration of B220⁺GFP⁺ cells was traced and analyzed by FACS on day+15 in peripheral and mesenteric lymph nodes (pLN), mLN) and skin. Indicated are percentages of B220⁺GFP⁺ cells among all recovered cells. In the skin, percentages and counts of B220⁺GFP⁺ cells are indicated per 1 cm² of skin surface. (C, D) Differentiation of CpG-proBs (C) and phenotype of the B220⁺GFP⁺ progeny assessed on day+15 in mLN. Isotype antibody controls were used to define positivity. The various B-cell subfractions were defined as FoB (CD21^loIgM^-CD93^-), T3 (CD21^loIgM^-CD93⁺), MZ (CD21⁺IgM⁺CD23^-), T2-MZP (CD21⁺IgM⁺CD23⁺) and T1+T2 (CD21^-IgM⁺) cells. (D) CpG-proB differentiation on day+15 in mLN, pLN and skin. (E, F) Cytokine expression by the CpG-proB progeny on day+15. (E) FACS profiles of cytokine (IL-10, TGF-β, IFN-γ, GM-CSF, TNF-α and IL-27) expression by CpG-proB-derived B220⁺GFP⁺ and non-CpG-proB-derived B220⁺GFP^- cells in the mLN. (F) Percent cytokine expressing B220⁺GFP⁺ (red) and B220⁺GFP^- (blue) cells in mLN, pLN and skin. Statistical analysis was performed with two-way ANOVA with Bonferroni multiple comparisons. (B, D, F) Results are expressed as mean ± SEM of 3-9 mice per group.

Figure 3

Characterization of two phases of cytokine expression by mLN CD4+ T-cell in controls with GVHD. (A) CD4⁺ T cells were stimulated by PMA + ionomycin in the presence of brefeldin and their intracellular cytokine expression was analyzed by FACS in mLN of GVHD controls at day+15 (black) and day+25 (blue). Data are expressed as means ± SEM of 5 mice per group. Statistical analysis was performed with two-way ANOVA with Bonferroni multiple comparisons. p values as indicated, n.s., non significant. (B) Heatmap representation of the mean of percentages of CD4⁺ T-cell expression of indicated cytokines in mLN of control mice with GVHD, at day+15 and day+25. Right: Colour scale of intensity of percentages.

Figure 4

T-cell subset analysis in mLN and pLN of CpG-proB recipients and GVHD controls. (A, B) Quantification by FACS analysis on day+25 of CD4⁺, CD8⁺ (% and cell counts) and CD4⁺Foxp3⁺ (%) in mLN (A) and pLN (B) of GVHD controls (black) and CpG-proB recipients (red). (C, D) Cytokine expression by CD4⁺ T cells in mLN (C) and pLN (D) of GVHD controls (black) and CpG-proB recipients (red). Data are expressed as means ± SEM of 5 mice per group. Statistical analysis was performed with unpaired Students’t- test (A, B) and two-way ANOVA with Bonferroni multiple comparisons (C–F). p values as indicated, n.s., non significant.

Figure 5

Follicular T-cell (Tf) analysis. (A, B) Percentages and counts of Tfh (CD4⁺CXCR5⁺Foxp3^-) and Tfr (CD4⁺CXCR5⁺Foxp3⁺) cells as well as Tfr/Tfh ratios on day+25 in mLN (A) and pLN (B) of mice, either CpG-proB recipients (red) or GVHD controls (black), were established by FACS analysis. (C, D) Cytokine expression by CD4⁺CXCR5⁺ cells assessed by FACS analysis. Percent IL-21- and IL-10-expressing cells in mLN (C) and pLN (D) of GVHD controls (black) and CpG-proB recipients (red). Results are expressed as means ± SEM from 5 mice per group. Statistical analysis was performed with unpaired Students’t- test (A, B) and two-way ANOVA with Bonferroni multiple comparisons (C, D). p values as indicated, ns= non significant.

Figure 6

Role of IFN-γ in the protective properties of CpG-proBs against GVHD. CpG-proBs were prepared from either WT or IFN-γ deficient C57BL/6J donors and adoptively transferred (7.5 x 10⁵ cells/recipient) on day+2 post-irradiation (DPI). (A) Diarrhea, cutaneous scores and survival of GVHD controls (injected with PBS, black, n=10), WT CpG-proB recipients (red, n=9) and IFN-γ deficient CpG-proBs (blue, n=9). Statistical analysis was performed with two-way ANOVA with Bonferroni multiple comparisons for diarrhea and cutaneous scores. Results are expressed as means ± SEM. P values as indicated. ns= non significant. (B) CpG-proB progeny, derived from either WT (red) or IFN-γ deficient (blue) C57BL/6J CpG-proBs, was gated as CD45.2 H2Kb⁺ cells in mLN of GVHD Balb/c (H2Kd) recipients of CD45.1 TBCD-BM and splenocytes from CD45.1 C57BL/6J donors and their cytokine expression analyzed by FACS as in Figure 2E on day+15 after adoptive transfer. N=3 mice per group. (C) Lymph node cells from naive C57BL/6J mice were co-cultured at a 1:1 ratio with WT or IFN-γ KO CpG-proBs at 5 x 10⁵ cells/ml for 3 days in RMPI 1640 medium, 10% FCS, 1% antibiotics, 0.1% β-mercaptoethanol in the presence of anti-CD3 (200 ng/ml) and analyzed by FACS for IL-10 and IL-21 expression in gated CD4⁺CXCR5⁺ Tfh cells. One experiment out of two. Statistical analysis was performed with one-way ANOVA for (B, C) Results are expressed as means ± SEM. p values as indicated.

Figure 7

Analysis of skin histological modifications, gene expression and cellular infiltrate in GVHD controls versus CpG-proB recipients. (A) CpG-proB recipients were mostly protected from alopecia and skin damage induced by GVHD in Balb/c recipients. Picture at day+70 of one representative mouse per group. (B) H&E staining of representative skin sections at day+70 in GVHD controls versus CpG-proB recipients. Scale bar = 100 μm. Red arrows indicate the epidermal thickness. Forty measures were taken per skin section. Right: Histogram representation of epidermal thickness in GVHD controls (black) and CpG-proB recipients (red). Results are expressed as means ± SEM from 6 mice/group. p value as indicated. Analysis was performed with unpaired Student’s t-test. (C) Heatmap showing significant fold-change expression of genes as measured by qRT-PCR microarray in skin fragments (2 cm²) isolated at day+70 from GVHD controls (right) and CpG-proB recipients (left). N = 3 animals per group. Analysis was performed with Qlucore. Listed are genes showing ≥1.4 expression fold change with p ≤ 0.05, considered significant. Right: Color scale of positive and negative fold-change gene expression. (D) Change fold of Col3A1 mRNA expression measured by qRT-PCR in skin samples recovered at day+70 from n=3 animals per group. (E) Flow cytometry analysis on day+15 and day+42 of total immune cell infiltrates as well as T-cell (CD4⁺ and CD8⁺) infiltrates in skin samples of GVHD controls (black) and CpG-proB recipients (red). Results are expressed as means ± SEM from 4 mice per group. Statistical analysis performed with unpaired Student’s t-test, ns, non significant, p values as indicated.

Figure 8

Flow cytometry analysis of skin infiltrates on day+15 and day+42 post-irradiation. CD4⁺, CD4⁺Foxp3⁺, CD8⁺, B220⁺PDCA-1^- B and CD11c⁺CD11b⁺ dendritic cell percentages and cell counts are shown in GVHD controls (black) and CpG-proB recipients (red). IL-10-expressing fraction of CD4⁺, CD8⁺, B220⁺ and CD11c⁺ cells on day+15 and day+42 in the skin of GVHD controls (black) and CpG-proB recipients (red). Results are expressed as means ± SEM for 5 mice per group. Statistical analysis was performed with unpaired Student’s t-test, ns, non significant. p values as indicated.

Figure 9

Graphical summary of the protective effects of CpG-proBs against GVHD. Adoptive transfer of CpG-proBs at the early phase of GVHD alleviated disease symptoms, in particular skin fibrosis. Following their migration into lymph nodes and skin, these progenitors produced many cytokines but depended on IFN-γ production for their protective effect. CpG-proB transfer reduced the CD4⁺ T-cell production of profibrotic cytokines TGF-β, IL-21 and IL-13 and enhanced the Tfr/Tfh T-cell ratio in lymph nodes. They also promoted the accumulation of IL-10-producing CD8⁺ T-cells, B-cells and dendritic cells in the skin. Figure created using BioRender.com.