Depletion of CD8⁺ T Cells Exacerbates CD4⁺ T Cell–Induced Monocyte-to-Fibroblast Transition in Renal Fibrosis Free

Fibroblast activation and excessive accumulation of extracellular matrix (ECM) are the hallmark events of chronic kidney diseases, which lead to widespread glomerular sclerosis and interstitial fibrosis (1). Fibroblast activation is regarded as a key event in renal fibrosis, because it is a principal effector cell for ECM production (2–4). Recent evidence indicates that fibroblasts may originate from bone marrow–derived fibroblast progenitor cells (also called fibrocytes) (5–9). Fibroblast progenitor cells could originate from a subpopulation of monocytes via monocyte-to-fibroblast transition (10–13). These cells express not only mesenchymal markers, such as collagen 1 (Col-1) and vimentin, but also hematopoietic markers, such as CD45 and CD11b (10, 14–16).

The inflammatory microenvironment plays a crucial role in the initiation of renal fibrosis after injury (17). Peritubular infiltration of inflammatory cells determines the extent and duration of renal interstitial fibrosis following obstructive injury (18). The differentiation of bone marrow–derived fibroblast progenitor cells is regulated by the profibrotic cytokine IL-4, whereas the antifibrotic cytokine IFN-γ inhibits their differentiation (12, 19). Several studies have shown that bone marrow–derived fibroblast progenitor cells are involved in the pathogenesis of renal fibrosis (3, 6, 9, 20). However, the regulatory mechanism underlying the recruitment and maturation of these cells in the injured kidney microenvironment is not fully understood.

T cells have been detected in the kidneys of patients with chronic kidney disease (21), as well as in models of renal fibrosis such as unilateral ureteric obstruction (UUO) (22–26). After obstructive injury, infiltrated T cells can produce chemokines and cytokines, which induce monocyte recruitment and amplify the inflammatory response (27, 28). Profibrotic IL-4 is mainly produced by CD4⁺ Th2 cells, and antifibrotic IFN-γ is primarily produced by CD4⁺ Th1 cells or CD8⁺ T cells in the regulation of fibrosis after injury (29, 30). Depletion of CD4 protects against renal fibrosis, whereas adoptive transfer of CD4⁺ (but not CD8⁺) T cells into lymphocyte-deficient mice restores fibrosis (11, 26, 31). Moreover, reconstitution of lymphocyte-deficient mice with Th2 cells enhances fibrosis to a greater degree than reconstitution with Th1 cells, identifying Th2 cells as a therapeutic target for renal fibrosis (31). In addition, T cells might participate in renal fibrosis by regulating the differentiation of circulating fibroblast precursor cells derived from the bone marrow (11). A role for CD8⁺ T cells in renal fibrosis has been suggested in the context of interstitial fibrosis or tubular atrophy after transplantation (21, 32). Reconstitution of RAG^−/− mice with CD8⁺ T cells indicates a potential antifibrotic role of CD8⁺ T cells in this model (26). However, it is unknown how inflammatory microenvironmental factors such as T cells regulate the differentiation of bone marrow–derived fibroblast progenitor cells in renal fibrosis.

In this study, we investigated the role of CD8⁺ T cells in the activation of bone marrow–derived fibroblast precursors in the kidney in tubulointerstitial fibrosis. Our results establish that the CD8⁺ T cell–IFN-γ–CD4 T cell axis is an important microenvironment for monocyte-to-fibroblast transition, which negatively regulates renal fibrosis.

Male CD8 knockout (KO) mice (B6.129S2-Cd8a^tm1Mak/J), IFN-γ KO mice (B6.129S7-Ifng^tm1Ts/J), EGFP transgenic mice [C57BL/6-Tg(CAG-EGFP)1Osb/J], and wild-type (WT) littermates on a C57BL/6 background were from The Jackson Laboratory (Bar Harbor, ME). All mice used for experiments were 10–12 wk of age and maintained under specific pathogen-free conditions in the Laboratory of Animal Experiments at Capital Medical University. The mice were given a standard diet. These mice were subjected to UUO as described previously (33). Briefly, UUO was performed under ketamine/xylazine anesthesia in which a midline incision was made, and the left ureter was exposed and tied off. Sham surgery was performed similarly but without ureter ligation. Mice were sacrificed on postoperative days 0, 1, 3, 5, or 7 by cardiac exsanguination. Kidneys were collected for analyses. All animal care and experimental protocols complied with the Animal Management Rule of the Ministry of Health, People’s Republic of China (Documentation no. 55, 2001), and the Guide for the Care and Use of Laboratory Animals published by the United States National Institutes of Health (National Institutes of Health Publication no. 85-23, revised 1996) and were approved by the Animal Care and Use Committee of Capital Medical University.

After anesthesia, mice were perfused with PBS through the left ventricle. The kidneys were removed and processed for cryosection or paraffin section. Sections were stained with Sirius red (34). The severity of interstitial renal fibrosis was evaluated by the calculation of the positive area of the total tissue measured with an NIS-Elements quantitative automatic program (Nikon, Tokyo, Japan). Immunofluorescence was performed using Abs against CD8, CD45, CD206, GFP, E-cadherin, α-smooth muscle actin (α-SMA), or Vimentin (Supplemental Table I). Briefly, sections were permeabilized in 0.3% Triton X-100 in PBS and blocked with protein block (DakoCytomation, Glostrup, Denmark) for 1 h at room temperature. Sections were incubated with a primary Ab mixed in Antibody Dilute (DakoCytomation) followed by detection of the primary Abs using secondary Abs conjugated to Alexa 568 or 488 (Invitrogen, Eugene, OR). Tissues were visualized using a Nikon 80i microscope or confocal microscope and images acquired using DS-cooled camera and NIS-Elements Br 3.0 software (Nikon, Melville, NY). To quantify the number of CD45⁺vimentin⁺ cells or CD206⁺vimentin⁺ cells in obstructed kidney, five fields that included CD45⁺ cells or CD206⁺ cells in cortex renis were imaged for each mouse, and five mice were analyzed in each group. To quantify the levels of vimentin in the cell, optical intensity of Alexa 568 (red) per pixel was analyzed by using ImageJ software (National Institutes of Health). Average of optical intensity in each high-power field (HPF) was taken as the value of vimentin in cell. A minimum of eight HPFs were analyzed per experiment, and experiments were repeated three times.

RNA was extracted by the TRIzol reagent method (Invitrogen). Aliquots of 2 μg total RNA were used for first-strand cDNA synthesis with Moloney murine leukemia virus reverse transcriptase (Promega, Southampton, U.K.). Aliquots of 2 μl reaction mixture were amplified with 10 μl SYBR Green PCR Master Mix and 1 μmol/l primers.

Amplification was at 95°C for 5 min, 95°C for 5 s, and 60°C for 30 s for each step for 45 cycles. Relative gene expression was calculated from cycle threshold (Ct) values using GAPDH as an internal control (relative expression = 2^{[Ct GAPDH − Ct sample]}). All samples were run in duplicate. Primers and their sequences are listed in Supplemental Table II.

Western blot analysis was performed as described elsewhere (34). Briefly, protein extracts were prepared from kidney samples with cell lysis buffer. In total, 50 μg protein lysates was separated by SDS-PAGE before transfer to nitrocellulose membranes (Bio-Rad, Hercules, CA), which were incubated with the primary Abs anti–α-SMA (1:3000), anti-fibronectin (1:1000), anti–Col-1 (1:1000), and anti-GAPDH (1:5000) at 4°C overnight and then with infrared dye-conjugated secondary Abs (1:10000) (Rockland Immunochemicals, Gilbertsville, PA) for 1 h at room temperature. The images were quantified by the use of the Odyssey infrared imaging system (LI-COR Biosciences, Lincoln, NE).

Kidneys and spleens were dissected and ground separately. Then, kidney fragments were digested with 2 ml collagenase type IA (2.5 U ml⁻¹; Sigma-Aldrich, St. Louis, MO) in PBS with 10 mmol CaCl₂ at 37°C for 30 min. After washing, the kidney and spleen slurries were passed separately through a 40-μm strainer (BD Biosciences, Franklin Lakes, NJ) and washed with PBS. Cells were collected by centrifugation at 1500 rpm for 5 min and incubated in PBS containing 2 mmol EDTA and 2% FBS plus primary Abs for 30 min at 4°C. Cells were resuspended at ∼1 × 10⁷ cells/ml before sorting or analysis. Cells were separated and analyzed by the Beijing Institute of Heart, Lung, and Blood Vessel Diseases Cytometry and Cell Sorting Core Facility, using BD FACSAria II or BD FACSCanto II (BD Biosciences), and data were collected using FACSDiva 7.0 software (BD Biosciences).Antibodies used are listed in Supplemental Table I.

CD8⁺ T cells were isolated from the spleens of EGFP-transgenic mice, WT or IFN-γ KO mice by cell sorting and used as donor cells for transferring into CD8 KO mice. Diluted 2 × 10⁶ CD8⁺ T cells in 100 μl DMEM were adoptively transferred into CD8 KO mice by tail vein injection, which was performed at the time of UUO surgery (35).

CD4 T cells were depleted by a single i.p. injection of CD4 mAb (RM4-5, BioLegend, San Diego, CA) as described (31). Briefly, RM4-5 (3 mg/kg) was administered for 1 d before and on the day of UUO and tissues harvested on day 7. Mouse IgG isotype control Abs were used as controls. FACS analysis was used to examine the depletion effect on CD4⁺ T cells. CD4⁺ T cells were stained with anti–CD45-PerCP–Cy5.5, anti–CD3e-PE–Cf594, and anti–CD4-PE (GK1.5) in day 3 and 7 blood (Ab information in Supplemental Table I).

Mouse bone marrow–derived monocytes were isolated and cultured according to a published protocol (36). Briefly, bone marrow–derived monocytes were isolated from the femur and tibia of 8–12-wk-old EGFP-transgenic mice, passed through a 40-μm cell strainer (BD Falcon), and cultured in RPMI 1640 medium containing 10% FBS, 10% L929 conditioned medium, 1% glutamine, 1% MEM vitamins, and 1% penicillin/streptomycin in a humid incubator at 37°C and 5% CO₂. Fresh medium was replaced every 2 to 3 d. More than 98% of the cells were stained positive for CD11b and F4/80. These cell were then cocultured with equal CD4⁺ T cells in RPMI medium containing 1% FBS and 1% penicillin/streptomycin for 2 d.

Results are expressed as mean ± SEM. Data were compared across different mouse strains and time points using two-way ANOVA. Significance testing was performed using one-way ANOVA followed by pairwise comparisons using the Student–Newman–Keuls test. Statistical significance was set at p < 0.05. A minimum of five replicates was performed for each experimental condition.

To investigate whether CD8⁺ T cell infiltration is linked to renal fibrosis, we performed UUO on C57BL/6 mice and collected kidneys at days 0, 1, 3, 5, and 7. CD8⁺ T cell infiltration into the kidney was examined by FACS on each day (Fig. 1A, 1B). E-cadherin staining was performed to show renal tubules. CD8⁺ cells increased gradually in the tubular interstitium following UUO, reaching a peak at d 5 (Fig. 1A, 1C). Previous studies have shown that CD8⁺ NK cells are present in some diseases (37). Thus, we tested whether CD8 KO mice were also depleted of NK cells by staining with the NK cell marker NK1.1 (Fig. 1D). The results showed the percentage of CD8⁺ NK cells in CD8⁺ cells was 3.1 ± 1.1% in WT mice, and the percentages of both NK and NKT in CD45⁺ cells in WT mice were not significantly different compared with CD8 KO mice (n = 6; p > 0.05) (Fig. 1E).

To examine whether CD8⁺ T cells play a role in the accumulation of bone marrow–derived fibroblasts in obstructed kidneys, WT and CD8 KO mice were subjected to UUO injury. CD45⁺Col-1⁺ cells in obstructed kidney were analyzed by FACS. Analysis indicated that the percentage of Col-1⁺ CD45⁺ cells increased in CD8 KO injured kidneys compared with those of WT mice (Fig. 2A, 2B). To confirm this result, WT or CD8 KO mice were subjected to UUO for 5 d, kidney sections stained for CD45, a leukocyte marker, and vimentin (38, 39), a mesenchymal marker, and examined by confocal microscopy. The number of CD45⁺ and vimentin⁺ cells increased in CD8 KO injured kidneys compared with those of WT mice (Fig. 2C). To determine whether kidney CD8 deficiency influences development of bone marrow–derived myofibroblasts, WT and CD8 KO mice were subjected to UUO for 7 d. Kidney sections were stained for α-SMA and examined by fluorescence microscopy. The results revealed that targeted deletion of CD8⁺ T cells increased the number of α-SMA⁺ myofibroblasts in obstructed kidneys compared with those of WT mice (Fig. 2D).

To elucidate the factors responsible for the increase in CD45⁺ myeloid fibroblasts in CD8 KO mice, we examined the effect of CD8 deficiency on kidney chemokine and cytokine expression. We have previously demonstrated that fibrosis is associated with CCL2 and CCL5 induction (35, 40, 41). Thus, we measured CCL2, CCL3, CCL4, and CCL5 mRNA levels in kidney of WT or CD8 KO mice. CD8 deficiency did not affect these chemokines in the obstructed kidney (Fig. 3A–D). IFN-γ, produced by CD8⁺ T and CD4⁺ Th1 cells, has been shown to inhibit fibroblast activation and fibrosis; this prompts us to test whether CD8 deficiency affects IFN-γ mRNA expression. IFN-γ was reduced in CD8 KO obstructed kidney (Fig. 3E). CD4⁺ Th2 cytokines such as IL-4 have been shown to promote M2 macrophage polarization and a monocyte-to-fibroblast transition (12, 17, 19, 42). Therefore, we examined the effect of CD8 deficiency on production of IL-4 and CD206, a M2 macrophage marker in the kidney. Notably, IL-4 and CD206 mRNA levels were markedly increased in CD8 KO obstructed kidneys compared with those of WT mice (Fig. 3F, 3H). These data indicate that CD8⁺ T cells may impair myeloid fibroblast accumulation and M2 macrophage polarization by limiting Th2 cytokine expression or activity. We next determined whether CD8 deficiency affected macrophage polarization and myeloid fibroblast formation by subjecting WT and CD8 KO mice to UUO for 5 d. To identify vimentin-producing M2 macrophages, kidney sections were stained for CD206 and vimentin. Results revealed that the number of CD206 and vimentin dual-positive cells increased in CD8 KO UUO kidneys compared with WT UUO kidneys (Fig. 3G, 3I). These results also suggest that myeloid fibroblasts are derived from monocytes through M2 macrophage polarization.

Because CD8⁺ T cells regulate kidney accumulation of bone marrow–derived fibroblasts in response to obstructive injury, we next examined the effect of CD8 deficiency on renal fibrosis development. To investigate whether CD8⁺ infiltration affects renal fibrosis, we examined fibrosis in CD8 KO UUO model. CD8 KO mice had no CD8⁺ cells in their blood, whereas WT mice did (Fig. 4A). Mice were then subjected to UUO surgery, kidneys were collected at day 7, and fibrosis was analyzed. Interstitial collagen deposition was elevated in CD8 KO obstructed kidneys, with 21.6 ± 3.0% of the cortex occupied by the interstitium versus 11.3 ± 2.1% in WT mice (p < 0.05; Fig. 4B). We next examined fibrotic markers fibronectin and α-SMA protein expression by immunoblotting (Fig. 4C). This revealed that both were elevated in CD8 KO UUO kidneys compared with WT mice (p < 0.05; Fig. 4D, 4E). These results indicate that CD8 deficiency increases renal fibrosis.

We adoptively transferred EGFP⁺ CD8⁺ T cells into CD8 KO mice to determine whether CD8⁺ T cells are responsible for the reduced fibrosis in obstructed kidneys. EGFP⁺CD8⁺ T cells were used to confirm that CD8⁺ cells in CD8 KO mice were donor derived. Mice were then subjected to UUO after successful adoptive transfer of CD8⁺ T cells (Fig. 5A), which suppressed interstitial fibrosis (p < 0.05; Fig. 5B). Western blotting revealed that fibronectin and Col-1 protein levels were decreased (p < 0.05; Fig. 5C, 5D). These results indicate that CD8⁺ T cells reduce renal fibrosis in obstructed kidneys.

We have shown that CD8 deficiency decreases antifibrotic IFN-γ secretion and increases profibrotic IL-4 production in UUO kidneys. Given that Th1 cells produce IFN-γ and Th2 cells produce IL-4, we tested whether CD8 deficiency affects CD4⁺ T cell differentiation. We isolated CD4⁺ T cells from the obstructed kidneys of WT and CD8 KO mice (Fig. 6A) and examined mRNA expression of Th1 markers IFN-γ and T-bet and of Th2 markers IL-4 and GATA3. IFN-γ and T-bet mRNA expression decreased, whereas IL-4 and GATA3 expression increased in CD4⁺ T cells isolated from CD8 KO UUO kidneys compared with WT CD4⁺ T cells (Fig. 6B–E). These data indicate that depletion of CD8⁺ T cells promotes differentiation of CD4⁺ T cells to a Th2 phenotype. To confirm that CD8 deficiency–induced fibrosis was due to CD4⁺ T cell differentiation, we depleted CD4⁺ T cells using a monoclonal CD4 Ab in CD8 KO mice (Fig. 6F) or CD8⁺ T cell–reconstituted CD8 KO mice (Fig. 6G) and examined fibrotic areas and fibrosis-related protein levels (Fig. 6H) in UUO kidneys. Depletion of CD4⁺ T cells reduced fibrosis in the obstructed kidney, regardless of the presence or absence of CD8⁺ T cells (Fig. 6I, 6J).

We have shown that depletion of CD8⁺ T cells promotes CD4⁺ T cell differentiation to Th2 cells and increases monocyte-to-fibroblast transition in vivo. In this study, we aimed to determine whether these CD4⁺ T cells regulate the monocyte-to-fibroblast transition in vitro. EGFP⁺ murine bone marrow–derived monocytes were cocultured with CD4⁺ T cells isolated from obstructed kidneys of CD8 KO or WT mice. Results showed that CD4⁺ T cells isolated from CD8 KO-obstructed kidney induce monocytes to produce a higher density of vimentin compared with WT mice (Fig. 7A, 7B). To confirm this result, we also determined collagen 1 and vimentin mRNA expression (Fig. 7C). These results indicate that CD8⁺ T cells reduce the CD4⁺ T cell–induced monocyte-to-fibroblast transition.

To examine IFN-γ and IL-4 mRNA expression in CD8⁺ or CD4⁺ T cells in UUO kidneys, we isolated CD8⁺ and CD4⁺ T cells from obstructed kidneys (Fig. 8A). IFN-γ expression in CD8⁺ T cells was higher than that in CD4⁺ T cells, whereas IL-4 expression was higher in CD4⁺ T cells (Fig. 8B, 8C). To examine whether IFN-γ–CD8⁺ T cells impair differentiation of CD4⁺ T to Th2 cells and renal fibrosis, we isolated CD8⁺ T cells from IFN-γ KO mice and used them or WT CD8⁺ T cells to reconstitute CD8 KO mice. The adoptive transfer efficiency of WT CD8⁺ T cells (CD45⁺CD3⁺CD8⁺IFN-γ⁺ cells) and IFN-γ KO CD8⁺ T cells (CD45⁺CD3⁺CD8⁺IFN-γ⁻ cells) in UUO kidney at day 7 were examined by FACS analysis (Fig. 8D). The results indicated that adoptive transfer efficiency of WT CD8⁺ T cells were similar to IFN-γ KO CD8⁺ T cells. Notably, reconstitution by IFN-γ KO CD8⁺ T cells did not reduce fibrosis or impaired CD4⁺ T cell differentiation to a Th2 phenotype. By contrast, reconstitution with WT CD8⁺ T cells reduced fibrosis and impaired CD4⁺ T cell differentiation to a Th2 phenotype (Fig. 8E–I).

Renal fibrosis is a common endpoint of numerous progressive kidney diseases. Recent evidence indicates that fibroblasts may originate from bone marrow–derived fibroblast progenitor cells (i.e., fibrocyte) (5–9), which are derived from a subpopulation of monocytes via monocyte-to-fibroblast transition (10–13). It is unknown how the inflammatory microenvironment regulates this transition. In this study, we identify the CD8⁺ T cell–IFN-γ–CD4 T cell axis as an important microenvironment for monocyte-to-fibroblast transition, which plays an antifibrotic role in renal fibrosis.

First, we showed that UUO stimulates infiltration of CD8⁺ T cells. We demonstrated that infiltrating CD8⁺ T cells were responsible for suppressed renal fibrosis. Kidney fibrosis was more severe in CD8 KO mice than in WT mice and less severe in CD8 KO mice transplanted with CD8 T cells (Figs. 4, 5). Tapmeier et al. (26) demonstrated a role for CD8⁺ T cells in renal fibrosis by reconstituting RAG^−/− mice with CD8⁺ T cells. In the current study, we revealed that CD8⁺ T cells play an antifibrotic role in the UUO model using CD8 KO mice. Consistent with the results of Tapmeier et al. (26), we showed that CD8⁺ T cells did not significantly influence renal fibrosis when CD4⁺ T cells were absent (Fig. 6). Recently, some evidence has indicated that the differentiation of CD4⁺ T cells to Th2 cells is a key event in CD4-induced renal fibrosis (31). Our present results also showed that absence of CD8⁺ T cells increased the differentiation of CD4⁺ T cells to Th2 cells as well as fibrosis, suggesting CD8⁺ T cells play an antifibrotic role in renal fibrosis via inhibition of the differentiation of CD4⁺ T cells to Th2 cells.

Tissue fibroblasts play a key role in pathologic fibrotic processes. Studies have shown that the bone marrow contributes to the expansion of the fibroblast population in multiple organs and tissues, especially in renal fibrosis (6, 9, 11). We found that depletion of CD8⁺ T cells increased the number of these cells in the obstructed kidney compared with WT mice (Fig. 2).

It is generally thought that M2 macrophage display a Th2-like phenotype, promoting fibroblast activation (42). We have reported that CCL2 and CCL5 are important for macrophage infiltration in renal fibrosis and other models (35, 40, 41), thus we tested CCL2, CCL3, CCL4, and CCL5 production in the obstructed kidney following UUO. The results showed that the absence of CD8⁺ T cells did not influence these chemokines, but increased the mRNA expression of CD206, an M2 macrophage marker. IFN-γ secreted from Th1 cells activates macrophages to an inflammatory phenotype. Alternatively, they can be activated by IL-4 and IL-13, both Th2 cytokines, to facilitate M2 macrophage differentiation and the repair processes (42). We also found that absence of CD8⁺ T cells caused a decrease in IFN-γ and an increase in IL-4 production. These results indicate that depletion of CD8⁺ T cells increases M2 macrophage polarization in renal fibrosis.

Recent data have shown myeloid fibroblasts are derived from monocytes through M2 macrophage polarization in the obstructed kidney (20). We also tested the number of CD206⁺vimentin⁺ cells in the obstructed kidney of CD8 KO and WT mice. The results showed that CD206⁺vimentin⁺ cells were increased in CD8 KO mice compared with WT mice. These data indicate that CD8⁺ T cells play an important role in M2 macrophage polarization and development of bone marrow–derived fibroblasts in the kidney.

We found that absence of CD8⁺ T cells promoted the differentiation of CD4⁺ T cells to Th2 cells and that CD4⁺ T cells, which were isolated from the obstructed kidneys of CD8 KO mice, promoted monocyte-to-fibrocyte differentiation. It has been reported that Th2 cytokines (IL-4 and IL-13) induce, whereas Th1 cytokines (IFN-γ and IL-12) inhibit, monocyte-to-fibrocyte differentiation and fibrosis. Taken together, profibrocyte activities of IL-4 and IL-13 and fibrocyte-inhibitory activities of IFN-γ and IL-12 counteract each other in a concentration-dependent manner (12). Thus, we isolated CD8⁺ T cells from IFN-γ KO mice and then reconstituted these cells into CD8 KO mice. The results showed that these cells neither significantly reverse fibrosis nor markedly influence the differentiation of CD4 to Th2 cells, whereas those from the WT mice do. These data indicate that IFN-γ plays a key role in the CD8⁺ T cell regulatory mechanism in renal fibrosis.

In summary, the current study demonstrates that renal injury with subsequent fibrosis is likely an interactive process among inflammatory cells, with different arms of the immune system involved at different stages. In this UUO model, we defined a novel role for CD8⁺ T lymphocytes that are induced to contribute to decrease renal fibrosis. CD8⁺ T cells negatively regulate M2 macrophage polarization and monocyte-to-fibroblast transition via inhibition of the differentiation of CD4⁺ T cells to Th2 cells, which contribute significantly to the pathogenesis of renal interstitial fibrosis in the initial process of renal fibrosis.

The authors have no financial conflicts of interest.