Background
We have focused on the development of effective immunotherapeutic strategies for central nervous system (CNS) tumors, such as glioblastoma multiforme (GBM). Preclinical studies have demonstrated that tumor-specific T helper type-1 (Th1) and T cytotoxic type-1 (Tc1) cells, but not type-2 counterparts, can efficiently traffic into CNS tumor sites and mediate effective therapeutic efficacy, recruited via the type-1 chemokine CXCL10 [
1‐
3] and the integrin receptor, Very Late Antigen (VLA)-4 [
4‐
7]. Despite the importance of the type-1 T cell response, cancers, including GBMs, secrete numerous type-2 cytokines [
8‐
10] that promote tumor proliferation [
11,
12] and immune escape [
13]. Hence, the strategic skewing of existing type-2 to type-1 immunity in glioma patients may be critical for the development of more effective immunotherapy.
MicroRNAs (miRs) are a novel class of endogenous small single-stranded RNA molecules which are 18-24 nucleotides in length [
14]. Mature miRs repress mRNA encoded protein translation and are highly conserved between species, including viruses, plants and animals [
15]. There are over 700 miRs identified in the human genome that collectively are predicted to regulate two-thirds of all mRNA transcripts [
14]. Findings over the past several years strongly support a role for miRs in the regulation of crucial biological processes, such as: cellular proliferation [
16], apoptosis [
17], development [
18], differentiation [
19], metabolism [
20], and immune regulation [
21,
22]. We recently reported that miR-222 and -339 in cancer cells down-regulate the expression of an intercellular cell adhesion molecule (ICAM)-1, thereby regulating the susceptibility of cancer cells to cytotoxic T lymphocytes (CTLs) [
23]. This is among the first reports to demonstrate the role of miR in cancer immunosurveillance.
In the current study, in an effort to understand the potential roles of miRs in anti-tumor immunity, we examined miRs differentially expressed in Th1 and Th2 cells. Our miR microarray and RT-PCR analyses revealed that of all analyzed miRs, members of the miR-17-92 cluster (miR-17-92) are of the most significantly over-expressed miRs in murine Th1 cells when compared with Th2 cells. The miR-17-92 transcript encoded by mouse chromosome14 (and human chromosome 13) is the precursor for 7 mature miRs (miR-17-5p, miR-17-3p, miR-18a, miR-19a, miR-20a, miR-19b and miR-92) [
24,
25]. This cluster is also homologous to the miR-106a-363 cluster on the X chromosome and the miR-106b-25 cluster on chromosome 5. Together, these three clusters contain 15 miR stem-loops, giving rise to 14 distinct mature miRs that fall into 5 miR families. The members in each family have identical seed regions. This genomic organization is highly conserved in all vertebrates for which complete genome sequences are available [
26].
miRs in the miR-17-92 cluster are amplified in various tumor types, including B cell lymphoma and lung cancer, and promote proliferation and confer anti-apoptotic function in tumors, thereby promoting tumor-progression [
27‐
31]. Knockout and transgenic studies of the miR-17-92 cluster in mice have demonstrated the importance of this cluster in mammalian biology [
25]. Transgenic mice with miR-17-92 overexpressed in lymphocytes develop lymphoproliferative disorder and autoimmunity but not cancer [
24]. These findings demonstrate a critical role for miR-17-92 cluster in T cell biology.
We show here that miR-17-92 is up-regulated in Th1 cells when compared with Th2 cells. IL-4 and STAT6 signaling mediate the down-regulation of miR-17-92. Tumor-bearing host conditions also suppress the miR-17-92 cluster expression in T cells, which is associated with a loss in ability to produce IFN-γ. This led us to hypothesize that miR-17-92 cluster overexpression might enhance type-1 responses. Indeed, type-1 T cells derived from miR-17-92 transgenic mice demonstrated a more pronounced type 1 phenotype including enhanced IFN-γ production and increased VLA-4 expression when compared with control type-1 T cells. These findings suggest that miR-17-92 plays a critical role in type-1 adaptive immunity.
Materials and methods
Reagents
RPMI 1640, FBS, L-glutamine, sodium pyruvate, 2-mercaptoethanol, nonessential amino acids, and penicillin/streptomycin were obtained from Invitrogen Life Technologies. Recombinant murine (rm) IL-12 was purchased from Cell Sciences Technologies. RmIL-4, recombinant human (rh) IL-4 and rhIL-2 were purchased from PeproTech. Purified monoclonal antibodies (mAbs) against IL-12 (C15.6), IFN-γ (R4-6A2), IL-4 (11B11), CD3 (145-2C11), CD4 (RM4-5), CD8 (53-6.7) and CD49d (R1-2) were all purchased from BD Pharmingen. Purified mAbs against CD3 (UCHT1) and CD28 (CD28.2) and IL-4 (MP4-25D2) were purchased from Biolegend. RT-PCR reagents and primers were purchased from Applied Biosystems and analyzed on a BioRad IQ5. WST-1 reagent was purchased from Roche. For isolation of T cells, immunomagenic isolation kits from Miltenyi Biotec were used. All reagents and vectors for lentiviral production were purchased from System Biosciences with the exception of Lipofectamine 2000, which was from Invitrogen.
Mice
C57BL/6 mice and C57BL/6 background STAT6 deficient mice (B6.129S2 [C]-Stat6tm 1Gru/J; The Jackson lab) (both 5-9 wk of age) were purchased from The Jackson Laboratory. C57BL/6-background miR-17-92 transgenic (TG) mice (C57BL/6-Gt [ROSA]26Sortm 3(CAG-MIRN 17-92,-EGFP)Rsky/J; The Jackson Lab) were maintained in the Hillman Cancer Center Animal Facility at University of Pittsburgh as breeding colonies and bred to C57BL/6-background mice transgenic for Cre recombinase gene under the control of the Lck promoter (B6.Cg-Tg [Lck-cre]548Jxm/J, the Jackson Lab) to obtain mice, in which T cells expressed miR-17-92 at high levels (miR-17-92 TG/TG). For mouse tumor experiments, C57BL/6 mice and C57BL/6 background STAT6-/- mice received subcutaneous injection of 1 × 106 B16 tumor cells resuspended in PBS into the right flank. On day 15 following tumor inoculation, mice were sacrificed and splenic T cells were isolated. Animals were handled in the Hillman Cancer Center Animal Facility at University of Pittsburgh per an Institutional Animal Care and Use Committee-approved protocol.
T cells from Healthy Donors and Patients with GBM
This study was approved by the local ethical review board of University of Pittsburgh. All healthy donors and patients with GBM signed informed consent before blood samples were obtained. To determine the impact of IL-4, healthy donor-derived CD4+ T cells were isolated with immunomagentic-seperation and stimulated with 100 IU/ml rhIL-2, anti-CD3 and anti-CD28 mAbs (1 μg/ml for each) in the presence or absence of rhIL-4(10 ng/ml). RT-PCR analyses were performed with both healthy donor- and patient-derived T cells to determine the expression of miR-17-92 as described in the relevant section.
Th1 and Th2 Cell Culture
Th1 and Th2 cells were differentiated from immunomagnetically-separated CD4+ splenic T cells. Magnetic activated cell separation (MACS) was carried out using positive selection. Briefly, spleens were minced in complete media, resuspended in red blood cell lysis buffer and stained with immunomagnetically labeled anti-CD4 antibody. Cells were then washed and placed through the magnetic column in 500 μl of MACS buffer. The column was then washed 3 times with buffer and then removed from the magnet and labeled cells were extracted in 3 ml of MACS buffer.
For differentiation of T cells, purified CD4+ cells were stimulated in 48 well plates with anti-CD3 mAb (5 μg/ml) in the presence of irradiated C57BL/6 spleen cells (3000 Rad) as feeder cells. RmIL-12 (4 ng/ml), rmIFN-γ (4 ng/ml), anti-IL-4 (10 μg/ml) mAb and rhIL-2 (100 IU/ml) were added for Th1 development. Th2 cells were generated from the same CD4+ cell precursors stimulated with anti-CD3 mAb and feeder cells in the presence of rmIL-4 (50 ng/ml), two anti-IFN-γ mAbs (10 μg/ml), anti-IL-12 mAb (10 μg/ml) and rhIL-2 (100 IU/mL). After 10 days cells were stained for IL-4 and IFN-γ to confirm differentiation. Neutral cell culture included anti-CD3, feeder cells and rhIL-2. For studies involving IL-4 blockade, 12.5 ng/ml anti-human IL-4 mAb (Biolegend) was used in human experiments and 2.5 μg/ml anti-mouse IL-4 mAb (11B11) in murine studies. IFN-γ and IL-4 in the culture supernatants were measured using specific ELISA kits (R&D Systems). For FACs analysis, cells were incubated with mAb at 4°C for 30 min, washed twice in staining buffer, and fixed in 500 μl of 2% paraformaldehyde in PBS. Cells were stored in the dark at 4°C until analysis. Flow cytometry was carried out on the Coulter XL four-color flow cytometer at the flow cytometry core facility of the University of Pittsburgh Cancer Institute.
miR Microarray
Total RNA was isolated from Th1 and Th2 cells using the Trizol reagent and quality was confirmed with an A260/A280 ratio greater than 1.85. Two μg of total RNA was labeled with either Hy5 (red; Th1) or Hy3 (green; Th2) fluorescent dyes using miRCURY LNA microRNA labeling kit (Exiqon, Woburn, MA) according to manufacturer's protocol. Labeled miR samples in duplicate were cohybridized on to miR array slides, a custom spotted miR array V4P4 containing duplicated 713 human, mammalian and viral mature antisense microRNA species (miRBase:
http://microrna.sanger.ac.uk/, version 9.1) plus 2 internal controls with 7 serial dilutions printed in house (Immunogenetics Laboratory, Department of Transfusion Medicine, Clinical Center, National Institutes of Health) [
32]. After washing, raw intensity data were obtained by scanning the chips with GenePix scanner Pro 4.0 and were normalized by median over entire array. Differentially expressed miRs were defined by mean (n = 2) fold change (Th1/Th2 signal intensity) >2.
Quantitative RT-PCR
Total RNA was extracted using the Qiagen RNeasy kit and quality was confirmed with a A260/A280 ration greater than 1.85. RNA was subjected to RT-PCR analysis using the TaqMan microRNA Reverse Transcription Kit, microRNA Assays (Applied Biosystems), and the Real-Time thermocycler iQ5 (Bio-Rad). The small nucleolar SNO202 was used as the housekeeping small RNA reference gene for all murine samples and RNU43 for human samples. All reactions were done in triplicate and relative expression of RNAs was calculated using the ΔΔ
CT method [
33].
WST-1 Proliferation Assay
For WST-1 proliferation assays, 1 × 104 cells were cultured in a 96 well plate for 24-48 hours in 100 μl of complete media. Then, 10 μl of WST-1 reagent was added to each well. Cells were incubated at 37°C, 5% CO2 for 4 hours, and placed on a shaker for 1 min. The plates were then read on a micro plate reader with a wavelength of 420 nm and a reference at 620 nm.
Assays using Jurkat lymphoma cells transduced with miR-17-92
Jurkat human T cell leukemia cells (American Type Culture Collection) were transduced by either one of the following pseudotype lentiviral vectors: 1) control vector encoding GFP; 2) the 17-92-1 expression vector encoding miR-17 18 and 19a, or 3) the 17-92-2 expression vector encoding miR 20, 19b-1, and 92a-1. All vectors were purchased from SBI. Lentiviral particles were produced by co-transfecting confluent 293TN cells (SBI) with pPACK-H1 Lentivirus Packaging Kit (SBI) and the miR containing expression vectors (SBI) noted above using Lipofectamine 2000 reagent (Invitrogen). Supernatant was collected after 48 hour incubation at 37°C with 5% CO2 and placed at 4°C with PEG-it Virus Concentration Solution (SBI) for 24 hrs. Supernatants/PEG solutions were then centrifuged and the pellet was resuspended in a reduced volume of media as viral stock. Jurkat cells were further resuspended in the viral stock together with polybrene (8 μg/ml) for 24 hrs. Fresh media was then added to the cells and transduction efficiency was evaluated by GFP expressing cells. For IL-2 production, transduced Jurkat cells were stimulated with Phorbol 12-myristate 13-acetate (PMA) (10 ng/ml) and ionomycin (500 nM) for overnight and supernatant was assayed for IL-2 by a human IL-2 ELIZA kit. For activation induced cell death (AICD), cells were treated with 10 μg/ml purified anti-CD3 mAb (UCHT1) from Biolegend for 24 hours and then cell viability was measured using WST-1 reagent.
Statistical Methods
All statistical analyses were carried out on Graphpad Prism software. The statistical significance of differences between groups was determined using student t- test. We considered differences significant when p < 0.05. A post test for linear trend test was used to determine linear trend and we considered p < 0.05 to be significant.
Discussion
Attaining effective tumor immunity is a major goal of modern biologic therapy, limited by the tumor microenvironment and profound regulatory mechanisms limiting T cell and NK cell effectors. Here we show that the type-2-skewing tumor microenvironment induces down-regulation of miR-17-92 expression in T cells, thereby hampering anti-tumor T cell responses. It also suggests that development of immunotherapy using miR-17-92-transduced T cells is warranted based on our findings demonstrating that ectopic expression of miR-17-92 in T cells leads to improved type-1 functions, including increased VLA-4 expression and IFN-γ production.
Blockade of endogenous IL-4 by inhibitory mAb or disruption of STAT6 signaling was sufficient to up-regulate miR-17-92 in T cells (Fig
3). These findings suggest that STAT6 may negatively regulate miR-17-92 expression in T cells. Several transcription factors have been identified that regulate expression of this miR cluster, including the E2 transcription factor (E2F) family members [
39,
40], c-Myc [
41], STAT3 [
42], as well as the sonic hedgehog pathway [
43,
44]. How IL-4 and the STAT6 signaling pathway negatively influence miR-17-92 expression at a molecular level remains to be elucidated. With regard to the effects of IL-4/STAT6 signaling on Th1 vs. Th2 functions, we have recently demonstrated that STAT6-/- Th2 cells exhibit Th1 phenotype with increased surface expression of VLA-4 [
45]. These observations have led us to hypothesize that STAT6-regulated miR-17-92 may contribute to the promotion of type-1 T cell functions.
Our findings indicate that the tumor-bearing host down-regulates miR-17-92 in T cells (Fig.
3 and
4). Interestingly, not only are
STAT6-/- T cells resistant to tumor-induced inhibition of miR-17-5p, but CD8
+ T cells in tumor bearing
STAT6-/- mice exhibited higher levels of miR-17-5p when compared with CD8
+ T cells obtained from non-tumor bearing
STAT6-/- mice. In addition to IL-4, other tumor-derived factors are likely to be involved in these events. Further studies are warranted to elucidate the molecular mechanisms underlying the regulation of miR-17-92 in T cells, especially in the tumor microenvironment.
While tumor bearing mice demonstrated decreased levels of miR-17-92 in both CD4
+ and CD8
+ cells, human GBM patients exhibited a statistically significant decrease of miR-17-92 in CD4
+ cells but not in CD8
+ cells (Fig.
4). However, there still appears to be a trend towards lower miR-17-92 expression in GBM patient-derived CD8
+ cells compared with those obtained from healthy donors. The lesser degree of miR-17-92 suppression in CD8
+ cells compared with CD4
+ cells in GBM patients is plausible based on our current understanding of CD4
+ and CD8
+ T cell biology. The type-1 vs. type-2 differentiation appears to be more distinct for CD4
+ T cells than for CD8
+ cells [
46,
47], and this may also be the case for miR-17-92. Another speculation is that CD8
+ T cells may be less sensitive to IL-4 than CD4
+ T cells thereby exhibiting less repression of miR-17-92. Further studies with larger sample size are warranted.
Messages encoding proteins that are targeted by miR-17-92 cluster miRs include: E2F1, E2F2, E2F3 [
40,
41], P21 [
48], anti-angiogenic thrombospondin-1 and connective tissue growth factor [
49], proapoptotic Bim, and phosphatase and tensin homolog (PTEN) [
24]. These proteins are all involved in cell cycle regulation or apoptotic cell death, further supporting the importance of miR-17-92 cluster in T cell biology. In fact, Bim and PTEN are down-regulated in T cells overexpressing miR-17-92 [
24]. Furthermore, TGF-β receptor II (TGFBRII) is one of the established targets of miR-17-92 [
50]. We are currently evaluating whether miR-17-92 transgenic T cells show down-regulated TGFBRII and decreased sensitivity to TGF-β.
In agreement with others [
24], our findings demonstrating increased IFN-γ production from miR-17-92 TG/TG T cells compared with control cells suggest that miR-17-92 may actually promote the type-1 skewing of T cells (Fig.
5 and
6C). As miR-17-92 targets hypoxia-inducible factor (HIF)-1α in lung cancer cells [
51], enhanced miR-17-92 expression in activated T cells may promote the type-1 function of T cells at least partially through down-regulation of HIF-1α. Although HIF-1 expression provides an important adaptation mechanism of cells to low oxygen tension [
52,
53], it does not appear to be critical for survival of T cells, unlike its apparent role in macrophages [
54]. T cells do not depend on HIF-1α for survival to the same degree as macrophages since activated T cells produce ATP by both glycolysis and oxidative phosphorylation [
55]. Rather, HIF-1α in T cells appears to play an anti-inflammatory and tissue-protecting role by negatively regulating T cell functions [
52,
56,
57]. Indeed, T cell-targeted disruption of HIF-1α leads to increased IFN-γ secretion and/or improved effector functions [
58‐
61]. Although available data on gene expression profiles in Th1 and Th2 cells do not suggest differential expression of HIF-1α mRNA between these cell populations [
62], as is often the case in miR-mediated gene expression regulation, miR-17-92 may still regulate HIF-1α protein expression at post-transcriptional levels. These data collectively suggest that miR-17-92 expression in activated T cells may promote the type-1 function of T cells at least partially through down-regulation of HIF-1α.
The human Jurkat T cell line with ecotopic expression of miR-17-92 cluster members demonstrate increased IL-2 production and improved viability following treatment with the AICD condition (Fig.
6). The Jurkat cell line was established from the peripheral blood of a T cell leukemia patient in the 1970s. This cell line is often used to recapitulate what would happen in humans T cells as the line retains many T cell properties, such as CD4, a T cell receptor, and ability to produce IL-2 [
63]. For these reasons, we chose to use Jurkat cells in our experiments. We recognize, on the other hand, that this cell line has pitfalls since this is a tumor cell line with enhanced survival compared to normal T cells due to their intrinsic biology. Thus, continued work with human T cells is clearly warranted.
miRs in the miR-17-92 clusters are amplified in various tumor types including B cell lymphoma and lung cancer, and promote proliferation and confer anti-apoptotic function in tumors, thereby promoting tumor-progression and functioning as oncogenes [
27‐
31]. However, miR-17-92 by itself may not be responsible for oncogenesis as transgenic mice with miR-17-92 overexpressed in lymphocytes develop lymphoproliferative disorder and autoimmunity but not cancer [
24]. miR-17-92 may cooperate with other oncogenes to promote the oncogenic process. Transgenic mice overexpressing both miR-17-92 and c-Myc in lymphocytes develop early onset lymphomagenesis disorders [
27]. On the other hand, knockout studies of the miR-17-92 cluster in mice have demonstrated the importance of this cluster in mammalian biology. While knockout of the miR-17-92 cluster results in immediate post-natal death of all progeny, knockout of either or both the miR-106a or miR-106b clusters are viable without an apparent phenotype [
64]. However knock out of the miR-17-92 cluster together with miR-106a or 106b cluster results in embryonic lethality [
25].
During lymphocyte development, miR-17-92 miRs are highly expressed in progenitor cells, with the expression level decreasing 2- to 3-fold following maturation [
24]. In addition, we have evaluated relative expression of miR-17-92 in a variety of Th cells as well as naïve CD4
+ cells. Naïve CD4
+ cells express miR-17-92 at the highest level among the cell populations examined. Albeit lower than that in naïve CD4
+ cells, Th1 cells express miR-17-92 at higher levels than T neutral (anti-CD3, feeder cells and IL-2) and Th17 cells, and Th2 cells consistently exhibit the lowest levels of miR-17-92 among the populations tested (data not shown). More studies are warranted on the specific role of miR-17-92 during differentiation.
These studies reviewed above provide us with critical insights as to what has to be expected if we develop therapeutic strategies by modulating miR-17-92 expression. One major barrier for successful T cell-based cancer immunotherapy is the low persistence of tumor antigen (TA)-specific T cells in tumor-bearing hosts [
65,
66]. It seems promising to generate genetically modified TA-specific T cells
ex vivo that are resistant to tumor-mediated immune suppression and mediate robust and long-lived anti-tumor responses. miR-17-92 cluster has the potential to confer resistance to tumor-derived immunosuppressive factors and to improve type-1 reactivity. Further characterization of the role of miR-17-92 cluster in tumor antigen (TA)-specific CTLs is clearly warranted and may provide us with ability to develop novel immunotherapy strategies with genetically engineered T cells. Additionally, identification of diminished miR-17-92 expression in the peripheral blood may emerge as an important biomarker in patients with malignancy.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
GK participated in the conception of the study, experimental design, performed in vivo and in vitro assays, and was one of the two primary writers of the paper. KS was involved in the conception of the study, further designing of the experiments, and took a primary role in culturing the differentiated T cells. AH performed studies using Jurkat cells. MF participated in experimental design, troubleshooting, editing the manuscript and statistical analysis. RU assisted in microRNA array and expression studies and analysis. HM helped with the ELISA and technical editing of the manuscript. TAR, JM and MTL participated in the design of experiments and interpretation of data. EW and FMM performed the miR microarray and assisted with analysis. HO conceived the study, mentored primary authors, was one of the two primary writers of the paper, and heavily participated in experimental design and data analysis. All authors read and approved the final manuscript.