A critical role for regulatory T cells in driving cytokine profiles of Th17 cells and their modulation of glioma microenvironment

Abstract

IL-17A, produced by Th17 cells, may play a dual role in antitumor immunity. Using the GL261-glioma model, we investigated the effects of Th17 cells on tumor growth and microenvironment. Th17 cells infiltrate mouse gliomas, increase significantly in a time-dependent manner similarly to Treg and do not express Foxp3. To characterize the direct effects of Th17 cells on GL261 murine gliomas and on tumor microenvironment, we isolated IL-17-producing cells enriched from splenocytes derived from naïve (nTh17) or glioma-bearing mice (gTh17) and pre-stimulated in vitro with or without TGF-β. Spleen-derived Th17 cells co-expressing IL-17, IFN-γ and IL-10, but not Treg marker Foxp3, were co-injected intracranially with GL261 in immune-competent mice. Mice co-injected with GL261 and nTh17 survived significantly longer than gTh17 (P < 0.006) and gliomas expressed high level of IFN-γ and TNF-α, low levels of IL-10 and TGF-β. In vitro IL-17 per se did not exert effects on GL261 proliferation; in vivo gliomas grew equally well intracranially in IL-17 deficient and wild-type mice. We further analyzed relationship between Th17 cells and Treg. Treg were significantly higher in splenocytes from glioma-bearing than naïve mice (P = 0.01) and gTh17 produced more IL-10 than IFN-γ (P = 0.002). In vitro depletion of Treg using PC61 in splenocytes from glioma-bearing mice causes increased IL-17/IFN-γ cells (P = 0.007) and decreased IL-17/IL-10 cells (P = 0.03). These results suggest that Th17 polarization may be induced by Treg and that Th17 cells in gliomas modulate tumor growth depending on locally produced cytokines.

Fig. 1

Increment of Th17 cells infiltration during glioma progression. a On day 10, only few CD4+ cells are detected inside the tumor area: most of double-positive lymphocytes, recruited from peripheral lymphoid organs, are concentrated around the tumor vessels. On day 20, after GL261 injection, the microenvironment immunocharacterization shows an increment of CD4+ IL-17+ population and double-positive cells are distributed as cellular cluster within the tumor area. Isotype control cells are not detectable in GL261-glioma (data not shown). b Infiltrating Th17 cells investigated by flow cytometry increased from 7.9 ± 1.9 on day 10 to 13.1 ± 5.2 on day 20 (P = 0.03). The panels show a representative dot plot and mean percentage value ± SD obtained from three different evaluations at each time point

Fig. 2

Treg and Th17 cells are differentially distributed into the tumor mass. a Arrows indicate the same blood vessels in adjacent sections of tumors evaluated for the presence of infiltrating Foxp3 and IL-17 positive cells at different time points; 7, 10, 15 and 19 days after tumor implantation. Two representative mice for each time point have been investigated, and representative images are displayed. b Quantitative analysis is summarized in histogram, results are expressed as mean ± SD of positive cells in 5 areas/mouse/time point (number of positive cells/5 different 40× HPF). c Immunofluorescence of GL261-glioma on day 19 showed infiltrating CD4+ IL-17+ population double-positive cells, but negative for Treg marker Foxp3. Foxp3-positive cells do not co-express IL-17

Fig. 3

Spleen-derived Th17 cells express IL-17, IFN-γ and IL-10 and do not express Foxp3. a Schematic experimental procedure. Spleen from naïve C57BL6N and glioma-bearing mice was surgically removed, and splenocytes were co-cultured in vitro with irradiated GL261 tumor cells in the presence or absence of rhTGF-β1. IL-17 secreting cells from different co-culture conditions (see the box below) were isolated and enriched immunomagnetically. Th17 cells were co-injected intracranially (ic) and subcutaneously (sc) with GL261 cells in immune-competent mice. b Th17 cells enriched by immunomagnetic separation were characterized by RT-PCR. All fractions express high levels of IL-17 and different expression of IFN-γ and IL-10 (*P < 0.01; **P < 0.0001; ***P < 0.00005). As shown in histogram, Foxp3 has not been detected

Fig. 4

Th17 cells play a role on GL261-glioma growth. a MRI performed 16 days after tumor implantation shows that lesion derived from GL261 + PBS was large. b nTh17 cells provide a significant delay on tumor formation. T2 (b left, 20 slices; thickness 0.8 mm, no gap; TR: 2,658 ms/TE: 54 ms/RARE Factor 8/FOV: 2.20 × 2.20 cm²; averages 6; image matrix 256 × 256; in plane resolution 0.086 mm; acquisition time 8 min 30 s) and T1 with contrast medium representative images show the total absence of mass lesion compared to the other groups of co-injection (b right 18 slices; thickness 0.8 mm, no gap; TR: 435 ms/TE: 11 ms/FOV: 2.20 × 2.20 cm²; averages 4; image matrix 256 × 256: in plane resolution 0.086 mm; acquisition time 7 min 26 s). T2 and T1 images of gliomas from GL261 plus gTh17 (d) showed expanded and lesions compared to nTh17 or TGF-nTh17 (c). Lesion from co-injection with TGF-gTh17 (e) is shown both in T1 as a homogeneous enhancing mass and in T2 images as a strong mass effect toward lateral ventricle. Images refer to one representative animal for each group (five mice for each group). f Kaplan–Meier analysis curves represent the survival time (in days) of five mice for each group of co-injection. Survival rate of mice injected with nTh17 cells was significantly higher than that of control mice (GL261 plus PBS P = 0.005) and of mice injected with TGF-gTh17 cells (P = 0.0067). Mice injected with TGF-gTh17 cells survived significantly lower than controls (P = 0.02)

Fig. 5

Th17 cells modulate tumor microenvironment. RT-PCR analysis of different gene expression performed on paraffin-embedded ic-gliomas. mRNA levels in each group were normalized to the relative quantity of b2m; relative expression of different cytokines was evaluated versus the GL261 plus PBS. a and c nTh17 tumors show an enhanced expression in IFN-γ and TNF-α compared to GL261 and TGF-gTh17 tumors. On the contrary, TGF-β and IL-10 (b and d) are strongly increased in gTh17-treated mice (P value: *P = 0.01; **P < 0.001; ***P < 0.0001)

Fig. 6

IL-17 has no effect on the GL261-glioma formation. a MRI 7 T images on a frontal plane. T2 (Left) and T1 after contrast medium (Right) images on day 8 after GL261 implantation performed on a wild-type (WT) and IL-17 KO (KO) mouse show a similar tumor engraftment. b Brain pictures of wt and IL-17 KO 12 days after tumor implantation (Upper) and Ki-67 staining (magnification 40×, Lower) show a similar proliferation ratio in wt and IL-17 KO mice. c MRI 7 T images on a frontal plane. T2 (Left) and T1 after contrast medium (Right) images 16 after GL261 implantation performed on a wild-type and IL-17 KO mouse. d Kaplan–Meier analysis curves represent the survival time (in days) of IL-17 KO compared to WT mice injected ic with GL261 cells. All mice died by day 29. The median survival of IL-17 KO was 27 days (n = 5, mean ± SD is 26.8 ± 2.2); the median of WT mice was 26.5 (n = 5, mean ± SD is 26.8 ± 1.7) (P = 0.9)

Fig. 7

CD4+ CD25+ Foxp3+ cells (Treg) modulate the Th17 cytokine profile. a Intracellular FACS analysis of naïve and tumor bearing CD4+ lymphocytes. Treg are significantly higher in glioma-bearing mice versus naïve splenocytes (P = 0.01), and they increase significantly in the presence of TGF-β (P = 0.001 in gl-splenocytes; P = 0.02 in n-splenocytes). nTh17 produced higher levels of IFN-γ than IL-10 (P = 0.001) that appeared reduced by TGF-β (P = 0.01). On the contrary, gTh17 produced more IL-10 than IFN-γ (P = 0.002). The presence of TGF-β promoted a relevant decrease in IFN-γ (P = 0.01) and a significant increase in IL-10-producing T cells (P = 0.03). Notably, the absolute number of Th17 cells obtained from the same number of splenocytes did not change significantly (nTh17 8.1 × 10⁵; TGF-nTh17 8 × 10⁵; gTh17 7.9 × 10⁵; TGF-gTh17 9.2 × 10⁵). Data are relative of three independent experiments. b Intracellular FACS analysis of tumor bearing CD4+ lymphocytes. PC61 antibody (0.5 μg/1 × 10⁶ cells) decreased Treg population from gl- and TGF-gl-splenocytes (P < 0.001), increasing IL-17/IFN-γ production (gTh17 P = 0.05; TGF-gTh17 P = 0.01) and decreasing IL-17/IL-10 secretion (gTh17 P = 0.0004; TGF-gTh17, P = 0.003). Data are statistically significant when they are compared to the same culture condition without PC61 (Fig. 7a). No significant differences (P > 0.05) were found in the same experiments performed with IL-6. In vitro data are represented as mean ± SD obtained combining percentage obtained from 24 to 48 h pre-stimulation of splenocytes