Fig 1.
SLC15A4 regulates TLR7 induced inflammatory cytokine secretion in dendritic cells and macrophages.
(A) Secretion of IFNα by bone marrow derived pDC in response to R848 stimulation (B-L) Cytokine/chemokine secretion by bone marrow derived pDC (B-D), myeloid DC (E-H), and macrophages (I-L) in response to stimulation with 1μg/ml of R848, or 5U/ml IFNα, or stimulation with both agents. Cytokine / chemokine levels in supernatants were assessed by Luminex (Bio-Rad platform) after 24 hours of stimulation and data is expressed as mean± SD for triplicate samples (n = 3). White bars, wild-type mice (C57BL/6); black bars, slc15a4-/- mice (on C57BL/6 background). Comparisons were made between wildtype group and knockouts or heterozygous groups *p < 0.05, **p<0.005, ***p<0.0005.
Fig 2.
SLC15A4 regulates cytokine secretion but not NFκB phosphorylation in B cells in response to TLR7 stimulation.
(A-D) Cytokine and chemokine levels in the supernatants of B cells stimulated with 1μg/ml of R848, or 5U/ml IFNα, or both agents for 72 hours were measured by Luminex (Bio-Rad platform). (E) lysosomal pH determined by lysosensor dye and expressed as histogram overlay of wild type B cells in absence (blue) or in presence of chloroquine (filled light blue) and slc15a4-/- B cells in absence (brown) or in presence of chloroquine (filled light orange). (F-G) Splenic B cells were stimulated with 1μg/ml of R848 for indicated times with or without chloroquine, and phosphorylation of signaling molecules was assessed by western blot. (H) IL-6 levels in the supernatants of B cells stimulated with 1μg/ml of R848 with or without chloroquine. Data is expressed as mean± SD for triplicate samples (n = 3). White bars, wild-type mice (C57BL/6); black bars, slc15a4-/- mice (on C57BL/6 background). *p < 0.05, **p<0.005, ***p<0.0005.
Fig 3.
SLC15A4 deficient mice are protected in pristane induced lupus model.
Two cohorts of 11-week-old wild-type mice (C57BL/6) and slc15a4-/- mice (on C57BL/6 background) were injected intraperitoneally with either 500μl PBS or pristane. Two weeks post pristane administration, peritoneal lavage was collected from cohort 1, and expression of Interferon signature genes (A-D), secretion of IL12p40 (E) and CCL-2 (F), and number of inflammatory Ly6Chi monocytes (G-H) were assessed by qRT-PCR, Luminex and flow cytometry, respectively. In cohort 2, serum concentrations of anti-nuclear antibodies were assessed 4 months (I) and 9 months (J) after pristane administration. Renal glomerulonephritis was assessed by histological assessment of H&E stained sections after 9 months of pristane administration (K) and scored as described in the methods section (L), and plotted with group means ±SD. *p < 0.05, **p<0.005, ***p<0.0005, N = 7 animals per group (some mice had to be euthanized during study due to fight injuries).
Fig 4.
Lack of SLC15A4 protects mice from developing spontaneous lupus phenotype in NZB/W F1 mice.
NZB/W F1 slc15a4+/+, slc15a4+/-, and slc15a4-/- mice were enrolled at an age of 13 weeks and monitored every two weeks. Survival (A) and proteinuria (B) were monitored throughout the study. Serum was evaluated at indicated times for the presence of anti-dsDNA antibodies (C), anti-histone antibodies (D), IL12p40 (E), and CCL-2 (F). Comparisons were made between wildtype group and knockouts or heterozygous groups *p < 0.05, **p<0.005, ***p<0.0005, N = 4–10 mice per group.
Fig 5.
Lack of SLC15A4 protects mice from developing spontaneous lupus nephritis in NZB/W F1 mice.
NZB/W F1 slc15a4+/+, slc15a4+/-, and slc15a4-/- mice were euthanized at the age of 36 weeks. H&E and PAS staining was carried out on formalin fixed tissues to assess renal disease severity by comparing severity scores of glomerulonephritis (A, B), tubulointerstitial nephritis (C, D) and periarteritis (E, F). Immune-complex deposition in kidney was assessed in frozen tissue sections by staining for IgG (G,H) and IgM (I,J) and quantified using FITC/AF488 integrated intensity in renal cortex. Examples (A, C, E, G, I) and quantification over several samples (B, D, F, H, J) is shown. Data is plotted with group means ± SD; individual data points represent results from two kidney sections on one slide from an individual mouse. *p < 0.05, **p<0.005, ***p<0.0005.
Fig 6.
Lack of SLC15A4 causes a defect in development of germinal centers and differentiation of antibody secreting cells in NZB/W F1 mice.
NZB/W F1 slc15a4+/+, slc15a4+/-, and slc15a4-/- mice were euthanized at the age of 36 weeks. Development of splenic germinal centers was assessed by PNA-lectin staining and immunohistochemistry in formalin fixed spleens (A) and quantified for all mice in the experiment (B). The phenotype of splenic lymphocyte population from individual mice was determined by flow cytometry and expressed as mean ± SD. Symbols represent individual mice and depict frequencies of proliferating B cells (B220+Ki67+) (C), germinal center B cells (GL7+Fas+) (D), splenic plasma cells (CD138+CD38lo) (E) and follicular T helper cells (PD1+CXCR5+) (F). Gene expression changes in spleens were assessed by RNA sequencing. Specifically, selected transcripts associated with plasma cell generation and induced upon disease development (log2-fold change > 2, p < 0.05) were significantly reduced in heterozygous and slc15a4-/- mice (G). Representative box plots of genes expressed by antibody secreting cells including J chain (H), tnfrsf17 (I) and txndc5 (J) are shown as mean ± SD. Comparisons were made between wildtype group and knockouts or heterozygous groups *p < 0.05, **p<0.005, ***p<0.0005, N = 7 per group.
Fig 7.
SLC15A4 deficient mice are protected from IFNα accelerated lupus disease development in NZB/W F1 mice.
14–15 week-old NZB/W F1 slc15a4+/+, slc15a4+/-and slc15a4-/- mice were injected intravenously with 1.2x108 pfu IFNα rAd5- IFNα. Three weeks post rAd-IFNα administration, serum IFNα levels were assessed by ELISA (A) and interferon gene signature score was assessed in spleen after 7 weeks (B). Mice were regularly monitored for survival (C) and proteinuria (D) rAd-IFNα administration. Serum anti-histone antibody (E) and anti-dsDNA antibody (F) titers were assessed by ELISA 3 weeks and 6 weeks post rAd-IFNα administration. Data is expressed as mean ± SD and each symbol represents an individual mouse in respective groups. Germinal center formation was assessed by PNA staining in spleen after 7 weeks; representative immunohistochemistry sections (G) and numbers of PNA+ germinal centers expressed as mean ± SD (H) are shown. Gene set enrichment analysis was performed using pre-defined immune gene modules after 7 weeks. Gene sets that showed significant enrichment (p<0.01) are shown with log fold change and FDR (I). Transcripts associated with antibody-secreting cell generation were significantly reduced in the absence of slc15a4 and visualized here as heatmap of select antibody-secreting cell-specific transcripts that were induced with disease (log2-fold change > 2, p < 0.05) (J). Comparisons were made between wildtype group and knockouts or heterozygous groups *p < 0.05, **p<0.005, ***p<0.0005, N = 5 per group.
Fig 8.
SLC15a4 deficient NZB/W F1 mice are protected from IFNα accelerated lupus nephritis.
7 weeks post rAd-IFNα administration, NZB/W F1 slc15a4+/+, slc15a4+/-and slc15a4-/- mice were euthanized and kidneys were collected to assess renal disease severity by histology and RNA sequencing. (A-D) Representative images and quantification of histological scoring for glomerulonephritis (A, B) and periarteritis (C, D) are shown. (E-F) Quantification of IgG (E) and IgM (F) immune complex deposition. Data is expressed as mean± SD. (G) RNA sequencing analysis was performed on kidney RNA. Fold changes between KO and WT at the disease timepoint are compared to those between naïve WT and disease WT animals. Each data point represents a gene. Red, green and blue dots represent genes that are significant only in naive versus diseased condition in wildtype (red), only in knockout versus diseased wildtype (green) or both (blue). (H) Gene set enrichment analysis of pre-defined immune and human LN ortholog gene modules are shown. (I-J) Effect of slc15a4 on human LN ortholog genes previously identified to be up-regulated in the glomerulus (I) or tubulointerstitium (J) of nephritic kidneys from human LN patients (see Material and Methods for details). Genes that showed individual significant changes in absence of slc15a4 are indicated (≥1.5 fold difference between groups with an adjusted p-value <0.01). Each column in the heatmap represents one animal within each group. *p < 0.05, **p<0.005, ***p<0.0005. N = 5–7 per group.