Regulator of G Protein Signaling-1 Selectively Regulates Gut T Cell Trafﬁcking and Colitic Potential X Deena L

Cutting Edge: Regulator of G Protein Signaling-1 Selectively Regulates Gut T Cell Trafficking and Colitic Potential x Deena L. Gibbons,*,† Lucie Abeler-Do¨rner,*,†,1 Tim Raine,*,‡,1 Il-Young Hwang, { ,{ ,{ ‡ Anett Jandke, Melanie Wencker,*x Livija Deban,* Christopher{ E. Rudd, Peter M. Irving,† John H. Kehrl, and Adrian C. Hayday*,†, The RGS1 gene is associated with celiac disease, mul- (5). Nonetheless, the strong associations of RGS1 single nu- tiple sclerosis, and type I diabetes, which are all T cell- cleotide polymorphisms (SNPs) with T cell-mediated dis- mediated pathologies, yet there is no reported analysis eases, including type 1 diabetes, celiac disease, and multiple of regulator of G protein signaling (RGS)1 biology in sclerosis (4, 6–8), argue for it being an important T cell human T cells. This study shows that RGS1 expression regulator. Therefore, this study investigates RGS1 biology in 2/2 is substantially higher in T cells from human gut versus human intestinal T cells, with supporting studies of Rgs1 peripheral blood and that this can be exaggerated in mice. The data argue for a pivotal role for RGS1 in T cell intestinal inflammation. Elevated RGS1 levels pro- trafficking and tissue immunopathologies. foundly reduce T cell migration to lymphoid-homing Materials and Methods chemokines, whereas RGS1 depletion selectively en- Human samples hances such chemotaxis in gut T cells and impairs their Mucosal biopsies and paired blood samples were taken from patients, with colitogenic potential. These findings provide a revised their consent, who had ulcerative colitis (UC) or Crohn’s disease (CD) and framework in which to view the linkage of RGS1 to in- were undergoing ileocolonoscopy or from control patients undergoing colo- flammatory disease. The Journal of Immunology,2011, noscopy for polyp follow-up or diagnosis. PBMCs were isolated using Ficoll (GE Healthcare), and gut intraepithelial lymphocytes and lamina propria 187: 2067–2071. lymphocytes were isolated, as described (9). Cells were stained and flow sorted into discrete populations. RNA was isolated; cDNA synthesis was per- formed; and RGS1, RGS2, and RGS10 mRNA were assayed by quanti- or T cells to make regulated responses to infections tative RT-PCR relative to cyclophilin A, using the following primers: RGS1: within tissues, they must enter, reside within, and then 59-TGTGCATTCAGATGCTGCTA-39 and 59-CTCGAGTGCGGAAGT- egress from relevant sites. Chemokines play pivotal CAATA-39; RGS2: 59-AAGATTGGAAGACCCGTTTG-39 and 59-GA- F 9 9 roles in such processes. In chronic inflammation, it may be TGAAAGCTTGCTGTTTGC-3 ;RGS10:5-CCCTGGAGAATCTGCT- GGAAC-39 and 59-TTTCCTGCATCTGAGTCTTA-39; and cyclophilin A: argued that T cell egress in response to lymphoid-homing 59-TGGTGGACCCAACACAAATC-39 and 59-TGCCATCCAACCACTCA- chemokines is impaired; hence the value of elucidating the GTCT-39. factors regulating the selective chemotaxis of T cells. Being G- Mice protein coupled receptors, chemokine receptors may be reg- 2 2 Wild-type (WT) and Rgs1 / C57BL/6 mice have been described (5); ulated by regulator of G-protein signaling (RGS) molecules 2 2 a C57BL/6 Rag2 / mice were from Taconic Farms. Mice were age-matched that bind G proteins, increasing their intrinsic GTPase ac- in all experiments. tivity up to 100-fold and, thereby, attenuating signaling (1). Rgs1 mRNA is highly enriched in murine gut versus lymphoid T cell transfer model of colitis 2 2 T cells (2–4), suggesting that it might likewise be enriched in A total of 0.5 3 106 naive, CD4+CD45RBhi cells from either WT or Rgs1 / 2 2 human gut T cells where it might selectively regulate che- C57BL/6 mice was injected i.p. into Rag2 / mice. Mice were weighed mokine responses. However, although RGS1 was shown to weekly and culled 7–8 wk postinjection. Colons were removed, measured, and subjected to histology. regulate B cell chemotaxis, there has been no substantive study of it in human T cells. Transient transfections Perhaps discouraging such analysis, splenic and lymph node Jurkat cells were transiently transfected with internal ribosome entry site 2/2 T cells from Rgs1 mice show seemingly normal chemotaxis (IRES)-enhanced GFP (EGFP) (Clontech) or RGS1-EGFP by electroporation,

*Peter Gorer Department of Immunobiology, King’s College London, London SE1 Foundation Trust in partnership with King’s College London and King’s College Hos- 9RT, United Kingdom; †National Institute for Health Research, Biomedical Research pital National Health Service Foundation Trust; The Wellcome Trust; the Marie Curie Centre, King’s Health Partners, Guy’s Hospital, London SE1 9RT, United Kingdom; fellowship programme; and the National Institutes of Health. ‡Department of Pathology, University of Cambridge, Cambridge CB2 1QP, United x { Address correspondence and reprint requests to Prof. Adrian C. Hayday, King’s College Kingdom; National Institutes of Health, Bethesda, MD 20892; and Cancer Research School of Medicine, Second Floor, Borough Wing, Guy’s Hospital, London SE1 9RT, UK, London Research Institute, London WC2A 3LY, United Kingdom United Kingdom. E-mail address: [email protected] 1L.A.-D. and T.R. contributed equally to this work. The online version of this article contains supplemental material. Received for publication April 1, 2011. Accepted for publication June 30, 2011. Abbreviations used in this article: CD, Crohn’s disease; EGFP, enhanced GFP; IRES, This work was supported by the National Institute for Health Research Comprehensive internal ribosome entry site; RGS, regulator of G protein signaling; SNP, single nucle- Biomedical Research Centre award to Guy’s and St. Thomas’ National Health Service otide polymorphism; UC, ulcerative colitis; WT, wild-type. www.jimmunol.org/cgi/doi/10.4049/jimmunol.1100833 2068 CUTTING EDGE: RGS1 SELECTIVELY REGULATES GUT T CELL TRAFFICKING alone or with CCR7-tomato (pMIT-tomato; kind gift of H. Jumaa, Max Planck Institute of Immunobiology and Epigenetics, Freiburg, Germany). Cells were left for 24–48 h and washed before use. Primary human T cell blasts were transfected with IRES-EGFP, RGS1-EGFP, or RGS10-EGFP (B isoform), using Amaxa II nucleofection.

Migration assays Transfected Jurkat cells were placed in Transwell-migration assays (5 mM pore; Corning) for 2–3 h at 37˚C, and migrated cells were enumerated by FACS Canto II (BD) analysis of 150 ml harvested from the bottom chamber. Cells migrated to 200 ng/ml human CXCL12 (ImmunoTools) or 500 ng/ml human CCL19 (R&D Systems). Percentages of migrated cells were calculated by 2 dividing the number of transmigrated EGFPhi, EGFPlow, or EGFP cells by the number of input cells with comparable EGFP levels. Murine intestinal T cells were isolated and placed in Transwell-migration assays, and migrated cells (CD45+, TCR+) were enumerated as above in response to recombinant murine CCL19 (50 ng/ml), CCL20 (50 ng/ml), CXCL12 (100 ng/ml; all from ImmunoTools), or CCL25 (50 ng/ml; R&D Systems).

Confocal imaging Transfected primary human T cell blasts were incubated overnight, washed, and equilibrated on an ICAM-1–precoated mVI slide (Ibidi, Munich, Germany). A total of 2 3 105 cells was added to each slide chamber and imaged every 10 s for 20 min. Prior to image acquisition, 10 ml medium or recombinant human CXCL12 (10 mg/ml; R&D Systems) was applied to filter paper at one end of the chamber. Time-lapse films were analyzed by Volocity (Improvision, Coventry, U.K.), and EGFP+ cells were tracked au- tomatically. Results and Discussion RGS1 overexpression in normal and inflamed gut To determine whether human intestinal T cells, like their mouse counterparts, are enriched in RGS1 expression, RGS1 RNA was quantified in purified TCRab+CD4+ and TCRab+CD8ab+ cells from peripheral blood (light diamonds) or gut biopsies (dark diamonds) (Fig. 1A). Not- withstanding interindividual variation common in human studies, RGS1 was overexpressed in human gut T cells by an FIGURE 1. ab+ + ab+ ∼ A, Paired samples of TCR CD4 or CD8 cells from average of 100-fold, as occurs in mice (2–4). RGS2 was blood or gut of normal patients were assessed for RGS1 mRNA relative to overexpressed less overtly (∼15-fold), and RGS10 was gener- cyclophilin A. Significance (by Student t test) is shown for comparisons of ally underrepresented in gut T cells (Supplemental Fig. 1A, blood and intestinal T cells. Note that RGS1 mRNA levels were not statis- 1B), again as is true in mice. RGS1 can regulate B cell mi- tically different between intestinal intraepithelial lymphocytes and lamina a propria lymphocytes (data not shown). B, RGS1 mRNA levels in gut CD4+, gration to chemokines by targeting G i (10), whereas RGS2 + + a CD8ab , or unconventional CD8aa /double-negative cells from normal preferentially targets G q (11), which is not commonly used donors (NHD) or patients with UC or CD. by chemokine receptors. Hence, RGS1 became the primary focus for further studies. Conventional (CD4+; CD8ab+) and pooled unconven- potential by 13 aa, have been reported (5), Jurkat cells, which 2 2 2 tional (CD4 CD8 double negative and CD4 CD8aa+) do not express RGS1, were independently transfected with ab T cells from normal, CD, and UC gut mucosa (Fig. 1B) each species linked via an IRES to EGFP. Relative to control, revealed consistently high RGS1 levels; although many CD/ EGFP+ empty-vector transfectants, EGFP+ RGS1 trans- UC samples fell within the normal range, some showed highly fectants showed dose-dependent inhibition of migration to elevated expression, particularly for intestinal CD4+ T cells CXCL12, a lymphoid chemokine for the receptor CXCR4. (p , 0.04). This is noteworthy, given that many such cells Moreover, transfectants expressing lower levels of RGS1 will have infiltrated the inflamed gut from the blood and (EGFPlow) were less compromised than were EGFPhi cells lymphoid organs where RGS1 levels are low (Fig. 1A). Al- expressing higher RGS1 (Fig. 2A). From data collated from though such cells will be activated in the gut, activation alone 10 independent experiments, RGS1 significantly decreased of peripheral blood T cells evoked only a transient increase in migration to CXCL12 by .35% (p , 0.01). Primary human RGS1 RNA, with peak levels ∼10-fold less than in gut T cells peripheral blood T cells, which express very little RGS1, were (Supplemental Fig. 1C). Moreover, RGS1 expression was not engineered to overexpress either RGS1 or RGS10, and their significantly different in naive versus memory CD4 T cells migration was assessed by time-lapse microscopy (Fig. 2B). (Supplemental Fig. 1D). Addition of CXCL12 to the bottom of the ICAM-1–coated slide redirected the random migration of empty-vector control RGS1 decreases T cell migration to lymphoid-homing chemokines or RGS10-transfected cells along the chemokine gradient To assess the functional significance of RGS1 overexpression, (Fig. 2B), with coincident decreases in migration velocity Transwells were used to assay T cell migration to chemokines. (p ∼ 0.001; Fig. 2C). Conversely, RGS1-transfected cells did Because two human RGS1 transcripts, differing in coding not align with the chemokine gradient (Fig. 2B,2C). Thus, The Journal of Immunology 2069

FIGURE 2. A, Jurkat cells transfected with EGFP control vector (green bars), short RGS1-EGFP (brown bars), or long RGS1-EGFP (blue bars) were assessed 2 in duplicate CXCL12-responsive migration assays. Specific migration for EGFP ve, EGFPlow, or EGFPhigh cells is shown from a representative experiment. *p , 0.05, **p , 0.02, Student t test. B and C, Peripheral blood T cells nucleofected with RGS1-EGFP, RGS10-EGFP, or control vector were assessed for CXCL12 responses. B, Migration tracks measured by time-lapse confocal imaging: cells migrated to chemokine added to the base of the slide. C, Velocities of cells transfected with the different vectors in the presence or absence of CXCL12. *p , 0.001, Kolmogorov–Smirnov comparisons. D, CCR7 expression levels on Jurkat cells transfected with control tomato vector and control EGFP vector (red), CCR7-tomato and control EGFP vector (green), or CCR7-tomato cotransfected with either “short-RGS1”–EGFP (brown) or “long-RGS1”–EGFP (blue). E, Specific migration to CCL19 of Jurkat cells transfected with CCR7- tomato in the presence or absence of RGS1 isoforms. increasing RGS1 limited CXCL12 responsiveness without chemokine CCL25 (CCR9 ligand) or to CCL20 (CCR6 li- grossly affecting normal kinesis. gand) was not affected (Fig. 3). Hence, RGS1 specifically To assess the capacity of RGS1 to regulate CCR7-mediated limited gut T cell responsiveness to the lymphoid-homing chemotaxis to CCL19, which promotes T cell egress from chemokines tested. Also, in two animal housing facilities, tissues to lymph nodes, Jurkat cells (which express very low levels of CCR7) were transfected with CCR7 (upstream of a tomato-tracker gene), resulting in CCR7 surface expression (Fig. 2D) and migration to CCL19 (Fig. 2E). Cotransfection with RGS1 did not effect CCR7 levels (Fig. 2D), but across 10 independent experiments it reduced migration to CCL19 by ∼54% (p , 0.003) (Fig. 2E). Having established that RGS1 is elevated in human gut T cells and can limit directional chemotaxis, the requirement for RGS1 in T cell migration was re-examined. Of note, the 2 2 comparable migration to CCL19 reported for WT and Rgs1 / splenic T cells (5) might simply reflect the negligible expression of RGS1 by splenic T cells. Consistent with this, in- 2/2 testinal T cells from Rgs1 mice showed significantly FIGURE 3. Migration of intestinal T cells (mean of three independent 2 2 enhanced migration to CCL19 and CXCL12 relative to WT experiments) isolated from WT or Rgs1 / mice (KO) in response to different intestinal T cells. By contrast, migration to the gut-homing chemokines. 2070 CUTTING EDGE: RGS1 SELECTIVELY REGULATES GUT T CELL TRAFFICKING

contribution to the colitogenic potential of T cells. These findings could be consistent with RGS1 ordinarily repressing T cell egress from the gut, possibly to sustain local immu- noprotection and/or immunoregulation vis-a`-vis commen- sals. Nonetheless, an analogous capacity to limit egress of inflammatory and/or autoimmune cells could clearly promote immunopathology. Such a key contribution of regulated T cell migration to pathology seems consistent with sponta- a neous colitis occurring in mice lacking G i2, which reportedly opposes the effects of RGS1 in B cells (14, 15). Additionally, the selective importance of RGS1 is strongly suggested by evidence that mice transgenic for Rgs16, which does not affect responses to CCL19 (16, 17) and which is not substantively expressed in the gut (data not shown), display T cell hyper- a responsiveness (as do G i2 mutant mice) but no colon pathology. Notwithstanding practical difficulties, it would seem ap- propriate to assess whether RGS1 is elevated in relevant tissue- associated T cells in type 1 diabetes, multiple sclerosis, and celiac disease, all of which are associated with RGS1 SNPs (4, 6–8). That such SNPs have not been associated with IBD may merely reflect the complex, multifactorial nature of IBD and should not diminish the prospect of RGS1 regulating disease initiation and/or progression. Indeed, RGS1 has been high- lighted as a biomarker for undifferentiated spondyloarthritis (18), often associated with gastrointestinal lesions (19). Clearly, RGS1 seems to be a valid, tissue-associated target for further

2 2 biological and clinical investigation. FIGURE 4. Rag2 / mice were injected in three separate experiments with either PBS (n = 6) or CD4+CD45RBhi T cells extracted from WT (n = 13) or 2 2 Rgs1 / mice (n = 11). Colons were removed 7–8 wk later, weighed, and Acknowledgments sectioned. Mean colon weight (A) and medial colon thickness (B) in recipients We thank T. Chan for sample collection, R. Ellis and T. Hayday for flow 2/2 of WT or Rgs1 T cells or PBS. Statistics were calculated using the Student cytometry, N. Powell for assistance with colitis experiments, and the Experi- t test. C, Representative mid-colon histologies for recipients of WT cells (left mental Histopathology Unit at Cancer Research UK. 2 2 panels)orRgs1 / cells (right panels). One colon for each group is shown from each of three independent experiments, stained with H&E. Scale bars, 500 mm. Disclosures The authors have no financial conflicts of interest. the gut-associated integrin, CD103, was found on significantly more CD4+ T cells (p , 0.008) (Supplemental Fig. References 1. Willars, G. B. 2006. Mammalian RGS proteins: multifunctional regulators of cel- 2A)atsignificantlyhigherlevels(p , 0.037) in the spleens of lular signalling. Semin. Cell Dev. Biol. 17: 363–376. 2 2 Rgs1 / mice versus WT mice, perhaps reflecting enhanced 2. Shires, J., E. Theodoridis, and A. C. Hayday. 2001. Biological insights into TCRgammadelta+ and TCRalphabeta+ intraepithelial lymphocytes provided by intestinal egress. serial analysis of gene expression (SAGE). Immunity 15: 419–434. To determine whether these effects extrapolated to a role for 3. Pennington, D. J., D. Vermijlen, E. L. Wise, S. L. Clarke, R. E. Tigelaar, and A. C. Hayday. 2005. The integration of conventional and unconventional T cells that Rgs1 in intestinal immunopathology, three experiments in two characterizes cell-mediated responses. Adv. Immunol. 87: 27–59. animal housing facilities on two continents compared the 4. Hunt, K. A., A. Zhernakova, G. Turner, G. A. Heap, L. Franke, M. Bruinenberg, 2 2 colitogenic potentials of Rgs1 / versus WT T cells after J. Romanos, L. C. Dinesen, A. W. Ryan, D. Panesar, et al. 2008. Newly identified 2/2 genetic risk variants for celiac disease related to the immune response. Nat. Genet. transfer to Rag2 mice (12). Invariably, colitis was amelio- 40: 395–402. rated by Rgs1 deficiency, as judged by a significantly reduced 5. Moratz, C., J. R. Hayman, H. Gu, and J. H. Kehrl. 2004. Abnormal B-cell responses to chemokines, disturbed plasma cell localization, and distorted im- increase in colon weight/thickness and by histopathology (Fig. 2/2 2 2 mune tissue architecture in Rgs1 mice. Mol. Cell. Biol. 24: 5767–5775. 4A–C). Recipients of Rgs1 / cells usually lost less weight 6.Smyth,D.J.,V.Plagnol,N.M.Walker,J.D.Cooper,K.Downes,J.H.Yang, J.M.Howson,H.Stevens,R.McManus,C.Wijmenga,etal.2008.Sharedand (Supplemental Fig 2B), although, as noted previously (13), distinct genetic variants in type 1 diabetes and celiac disease. N. Engl. J. Med. overall weight loss does not correlate well with colitis in this 359: 2767–2777. 7. Johnson, B. A., J. Wang, E. M. Taylor, S. J. Caillier, J. Herbert, O. A. Khan, A. H. system. Reduced disease did not reflect generally impaired Cross, P. L. De Jager, P. A. Gourraud, B. C. Cree, et al. 2010. Multiple sclerosis engraftment, because in contrast to the gut, spleens of recipient susceptibility alleles in African Americans. Genes Immun. 11: 343–350. 2 2 mice showed comparable accumulation of Rgs1 / and WT 8. International Multiple Sclerosis Genetics Conssortium (IMSGC). 2010. IL12A, MPHOSPH9/CDK2AP1 and RGS1 are novel multiple sclerosis susceptibility loci. T cells (Supplemental Fig. 2C). Genes Immun. 11: 397–405. In sum, this first study of RGS1 biology in human T cells 9. Lundqvist, C., M. L. Hammarstro¨m, L. Athlin, and S. Hammarstro¨m. 1992. Iso- lation of functionally active intraepithelial lymphocytes and enterocytes from human identified overtly high expression in the gut; exaggerated ex- small and large intestine. J. Immunol. Methods 152: 253–263. pression in some cases of inflammatory bowel disease (IBD); 10. Moratz, C., V. H. Kang, K. M. Druey, C. S. Shi, A. Scheschonka, P. M. Murphy, T. Kozasa, and J. H. Kehrl. 2000. Regulator of G protein signaling 1 (RGS1) a significant, nonredundant capacity to limit gut T cell markedly impairs Gi alpha signaling responses of B lymphocytes. J. Immunol. 164: responses to lymphoid-homing chemokines; and a profound 1829–1838. The Journal of Immunology 2071

11. Heximer, S. P., N. Watson, M. E. Linder, K. J. Blumer, and J. R. Hepler. 1997. 16. Estes, J. D., T. C. Thacker, D. L. Hampton, S. A. Kell, B. F. Keele, E. A. Palenske, RGS2/G0S8 is a selective inhibitor of Gqalpha function. Proc. Natl. Acad. Sci. USA K. M. Druey, and G. F. Burton. 2004. Follicular dendritic cell regulation of 94: 14389–14393. CXCR4-mediated germinal center CD4 T cell migration. J. Immunol. 173: 6169– 12. Powrie, F., M. W. Leach, S. Mauze, S. Menon, L. B. Caddle, and R. L. Coffman. 6178. 1994. Inhibition of Th1 responses prevents inflammatory bowel disease in scid mice 17. Lippert, E., D. L. Yowe, J. A. Gonzalo, J. P. Justice, J. M. Webster, E. R. Fedyk, reconstituted with CD45RBhi CD4+ T cells. Immunity 1: 553–562. M. Hodge, C. Miller, J. C. Gutierrez-Ramos, F. Borrego, et al. 2003. Role of 13. Ahern, P. P., C. Schiering, S. Buonocore, M. J. McGeachy, D. J. Cua, K. J. Maloy, regulator of G protein signaling 16 in inflammation-induced T lymphocyte mi- and F. Powrie. 2010. Interleukin-23 drives intestinal inflammation through direct gration and activation. J. Immunol. 171: 1542–1555. activity on T cells. Immunity 33: 279–288. 18. Gu, J., Y. L. Wei, J. C. Wei, F. Huang, M. S. Jan, M. Centola, M. B. Frank, and 14. Rudolph, U., M. J. Finegold, S. S. Rich, G. R. Harriman, Y. Srinivasan, P. Brabet, D. Yu. 2009. Identification of RGS1 as a candidate biomarker for undifferentiated G. Boulay, A. Bradley, and L. Birnbaumer. 1995. Ulcerative colitis and adenocar- spondylarthritis by genome-wide expression profiling and real-time polymerase cinoma of the colon in G alpha i2-deficient mice. Nat. Genet. 10: 143–150. chain reaction. Arthritis Rheum. 60: 3269–3279. 15. Han, S. B., C. Moratz, N. N. Huang, B. Kelsall, H. Cho, C. S. Shi, O. Schwartz, and 19. Orlando, A., S. Renna, G. Perricone, and M. Cottone. 2009. Gastrointestinal J. H. Kehrl. 2005. Rgs1 and Gnai2 regulate the entrance of B lymphocytes into lesions associated with spondyloarthropathies. World J. Gastroenterol. 15: 2443– lymph nodes and B cell motility within lymph node follicles. Immunity 22: 343–354. 2448.