| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Ophthalmology, Schepens Eye Research Institute, Harvard Medical School, Boston, MA 02114;
Experimental Immunology Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892; and
Department of Ophthalmology & Visual Science, Tokyo Medical and Dental University Graduate School of Medicine, Tokyo, Japan
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
is required in this process. We report that IPE produces both soluble and membrane-bound active TGF, but that only the latter is actually delivered to CD8+ T cells. In turn, these T cells become IPE Tregs by up-regulating their own expression of B7-1/B7-2 and soluble and membrane-bound TGF. IPE Tregs through their expression of B7 are able to engage CTLA-4+ bystander T cells, and thus precisely, target delivery of membrane-bound TGF. We propose that this mechanism of suppression via TGF ensures that soluble active TGF is not released into the ocular microenvironment where it can have unregulated and deleterious effects, including elevation of intraocular pressure and development of glaucoma. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
2, neuropeptides) (3) and 2) surface molecules expressed on ocular parenchymal cells, especially the pigment epithelium (4, 5, 6, 7, 8), the corneal epithelium (9), and the corneal endothelium (10). Ocular pigment epithelium (PE)4 lines the posterior surface of the iris, the ciliary body, and neural retina, thereby surrounding partially the privileged sites of the anterior chamber, the vitreous cavity, and the subretinal space, respectively. Freshly prepared, as well as cultured, PE cells from iris, ciliary body, and retina share the property of suppressing TCR-dependent activation of naive and primed T cells with which they are cultured (4, 5, 7). Moreover, T cells that have encountered ocular PE in vitro are spared from TCR-induced apoptosis and, instead, differentiate into regulatory T cells (6). We suspect that the ability of ocular PE to suppress T effector cell activity and to convert responding T cells into regulators is an important immune privilege strategy to limit immunogenic inflammation in the eye while promoting immune privilege. Iris PE (IPE) are of particular interest because these cells, unlike ciliary body PE (CBPE) and retina PE (RPE), use a cell surface contact-dependent mechanism exclusively to suppress T cell activation in vitro, i.e., soluble factors released from IPE fail to suppress activation of cocultured T cells, nor do they induce the responding T cells to become regulators (4). We have begun to identify the molecules involved in the cell contact process and to unravel the mechanism of suppression: IPE, both fresh and cultured, constitutively express B7-1 and B7-2 on their surface, and these molecules are required to be expressed by IPE if they are to suppress naive splenic T cells and then convert them into regulators (4). A subpopulation of CD8+ T cells among the splenic T cells in these cultures expresses CTLA-4, and interactions between B7 on IPE and CTLA-4 on CD8+ T cells leads to impaired T cell activation and conversion into IPE T regulators (Tregs) (5).
TGF is an immunomodulatory cytokine (11, 12, 13) that has been found to be constitutively present in ocular fluids (3, 14) and to be associated with regulatory T cells of different types (15, 16, 17, 18, 19, 20, 21, 22, 23). Suspecting that TGF might be involved in the process by which IPE generate Tregs, we have recently reported that as the CTLA-4+CD8+ T cells cultured with IPE become B7-expressing IPE Tregs, they secrete enhanced amounts of both latent and active TGF (5). Moreover, neutralizing anti-TGF Abs permitted naive T cells to be activated by anti-CD3 Abs in cultures containing IPE Tregs (6). Both of these findings suggest that IPE Tregs may use secreted TGF to suppress bystander T cells. However, we have also determined that neutralizing anti-TGF Abs are unable to prevent IPE from suppressing the activation of naive T cells by anti-CD3 Abs (8). It is relevant that there are several recent reports to the effect that a membrane-bound form of active TGF exists, and that this form may be used by other types of regulatory T cells (15).
The present experiments were designed to determine the extent to which IPE and T cells exposed to IPE 1) up-regulate their TGF and TGF receptor genes, 2) convert the latent TGF they produce into the active form, and 3) use membrane-bound or soluble TGF to suppress bystander T cells. The results indicate that both IPE and B7+CTLA-4+CD8+ IPE Tregs 1) produce enhanced amounts of active TGF and 2) use predominantly the membrane-bound form of TGF to suppress T cell activation. The evidence strongly suggests that surface interactions between B7 and CTLA-4 function to target the delivery of membrane-associated TGF to the appropriate T cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Adult C57BL/6 mice, purchased from Taconic Farms, served as donors of ocular PE cells and splenic T cells. Dr. James P. Allison (University of California at Berkeley, Berkeley, CA) provided CTLA-4 heterozygous mice from which we generated CTLA-4 homozygous progeny that were used at 3 wk of age (4, 5). Mice of the C57BL/6 background with disrupted genes for CD28 were purchased from The Jackson Laboratory. Dominant- negative TGF type II receptor (DN TGF RII) transgenic mice were generated by Drs. R. E. Gress and P. J. Lucas (24, 25).
Culture media
Primary cultures of pigmented epithelial cells from the iris (IPE) and ciliary body (CBPE) were cultured in RPMI 1640 complete medium composed of RPMI 1640, 10 mM HEPES, 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, 100 U/ml penicillin, 100 µg/ml streptomycin, 10% FBS (all from BioWhittaker), and 10–5 M 2-ME (Sigma-Aldrich). DMEM complete medium, used for primary cultures of retinal PE (RPE) cells, contained 0.1 mM nonessential amino acids, 100 U/ml penicillin, 100 µg/ml streptomycin, and 20% FBS. Serum-free medium was used in cultures and in assays involving T cells stimulated with anti-CD3 Abs to mimic as closely as possible the intraocular microenvironment behind the blood-ocular barrier. Serum-free medium is composed of RPMI 1640 with 0.1% BSA (Sigma-Aldrich) and 0.2% insulin, transferrin, and selenium culture supplement (ITS+) (Collaborative Biochemical Products).
Preparation of cultured pigment epithelium from iris
PE cells of iris were isolated and cultured as described previously (4, 5, 7). In brief, iris tissue was dissected out of C57BL/6 eyes, incubated (1 h) in PBS containing 1 mg/ml Dispase and 0.05 mg/ml DNase I (both from Boehringer Mannheim). The single-cell suspensions from four iris tissues were cultured for 14 days at 37°C in 5% CO2 and air. The primary cultures were found to be >99% cytokeratin positive (Clone PCK-26; Sigma-Aldrich) by flow cytometry. The CBPE and RPE used as controls were obtained from C57BL/6 mice as described in previous studies (4, 5, 7) and were also found to be 95% cytokeratin positive after 14 days of incubation.
Culture preparation and activation assays of T cells
The CD3+ cells were enriched using an Immulan mouse T cell kit from Biotex Laboratories that yielded >95% CD3-positive cells by flow cytometric analysis. The CD3+ T cells (106 cells/well) were placed into the IPE culture wells for 24–48 h. The level of IPE contamination of the harvested T cells was <0.97% by flow cytometry using anti-cytokeratin Abs. For anti-CD3-driven T cell activation, purified T cells (wild type or DN TGF RII donors) were added (2.5 x 105 cells/well) to culture wells containing gamma-irradiated (2000 rad) IPE or not. Anti-mouse CD3
Ab (clone 2C11; BD Pharmingen) was added to the wells and the cultures were maintained for 72 h. Purified T cells were stimulated with the Abs at different concentrations, 0.1, 0.25, 0.5, and 1.0 µg/ml. After 72 h of incubation, the cultures were assayed for uptake of [3H]thymidine (1 µCi/ml) added during the terminal 8 h of culture of cell proliferation. In some experiments, CD8+ T cells were enriched using MACS beads (MACS cell isolation kits; Miltenyi Biotec) where >95% of cells expressed CD8 by flow cytometric analysis.
ELISA for cytokines and bioassay for TGF
The concentration of IFN-
in supernatants of the T cell cultures was measured by sandwich ELISA according to the manufacturer’s instructions (BD Pharmingen) using a rat mAb to mouse IFN- (clone R4-6A2; BD Pharmingen) as the capture Ab and biotinylated rat mAb to mouse IFN- (XMG1.2; BD Pharmingen) as the detecting Ab. Mouse rIFN- (BD Pharmingen) was used to establish the standard curve for the assay.
To measure the concentration of mature (active) TGF, we used the mink lung assay. The supernatants of T cell-exposed IPE cells were collected and added to wells of cultured Mv1Lu cells (CCL-64 cell line; American Type Culture Collection) as described previously (5). The acidified (1 N HCl) supernatants (pH 2.0) were incubated for 1 h. To measure total TGF in the culture supernatants, the transiently acidified supernatants (100 µl) were added to the wells of a 96-well flat tissue culture plate containing Mv1Lu cells (105 cells) in Eagle’s MEM (100 µl) and cultured for 24 h at 37°C in CO2 and air, we pulsed (mg/ml) with 1 µCi of [3H]thymidine in the last 4 h. The concentration of TGF was calculated by the suppression of Mv1Lu cell proliferation in comparison to the suppression in proliferation by known amounts of TGF1 and TGF2 (R&D Systems).
Detection of TGF and TGF receptor transcripts within ocular PE and in T cells exposed to IPE
Cellular extracts were prepared from the cultured primary ocular PE or from purified T cells exposed to IPE cultured as described above. Enriched naive T cells (column-purified splenic T cells) obtained from wild-type, DN TGF RII, or CTLA-4 knockout (KO) donors were added to cultures of primary ocular IPE as described above, but were incubated for 24 h. The cultured PE and T cells were washed twice with PBS, then treated with RNA STAT60. PCR was conducted by the Hot-start PCR method with AmpliTaq and AmpliWax (Applied Biosystems). The products were subjected to 35–40 cycles of PCR amplification. Primers for TGF1 were 5'-CAAGGAGACGGAATACAGGGCT-3' and 5'-CGCACACAGCAGTTCTTCTCTGT-3', giving an amplification product of 260 bp. Primers for TGF2 were 5'- CACCACAAAGACAGGAACCTG-3' and 5'-GCGAAGGCAGCAATTATCCTGCAC-3', giving an amplification product of 327 bp. Primers for TGF receptor II were 5'-CGCCAACAACATCAACCAC-3' and 5'-CAGGCAACAGGTCAAGTCGT-3', giving an amplification product of 439 bp. The forward and reverse primers used for GAPDH, CD80 (B7-1), and CD86 (B7-2) were the same as described previously (4). The PCR products were electrophoresed in 1% or 1.5% agarose gel and visualized by staining with ethidium bromide. The expression level of mRNA was standardized by the expression of GAPDH as an internal control.
Flow cytometry
CD8+ T cells were enriched by magnetic MACS beads as described above. The purified CD8+ T cells from wild-type, CD28 KO, or DN TGF RII mice were stimulated with anti-CD3 in the presence or absence of IPE and were incubated for 24 h. The T cells were collected and stained with PE-conjugated anti-CD152 mAb (clone UC10-4F10-11; BD Pharmingen) for CTLA-4. Before staining, the cocultured cells were incubated with mouse Fc block (FcIII/II receptor, clone 2.4G2; BD Pharmingen) for 15 min and incubated with intracellular staining materials (BD Cytofix/Cytoperm kits; BD Pharmingen). We used PE-conjugated hamster IgG isotype (BD Pharmingen) as the isotype control for CTLA-4. The expression of CD152 was also analyzed on T cells treated with TGF2 (R&D Systems) at its ocular physiological concentration of 5 ng/ml.
Flow cytometry was also used to analyze the expression of membrane-bound (cell surface) TGF on cultured IPE or CD8+ T cells exposed to IPE. Enriched CD8+ were purified T cells incubated with IPE blocked with mouse Fc block and stained with monoclonal anti-TGF1/2 (R&D Systems) or control mouse IgG isotype at 4°C for 30 min. The bound primary Ab was detected with a biotin-conjugated anti-mouse IgG (BD Pharmingen) Ab and FITC-conjugated streptavidin (BD Pharmingen). The cells were double stained with CyChrome-conjugated anti-CD8 mAbs (BD Pharmingen). Cultured IPE cells were stained with anti-TGF2 Abs (R&D Systems) or control mouse IgG using the same methods. We also examined intracellular TGF2 with the same abs (BD Pharmingen Cytofix/Cytoperm kits).
Detection of surface TGF and TGF receptors on IPE and T cells exposed to IPE in immunohistochemistry
Colocalization of both TGF2 and the receptors was done by double immunohistochemical labeling. IPE was harvested from the iris of C57BL/6 mice as described above and cultured on coverslips for 14 days. Anti-CD3-stimulated T cells were either cultured on coverslips or added to the culture. The cultured cells were fixed with acetone directly on the coverslips and then rinsed with PBS treated with Fc blocks for 30 min. The cultures were double labeled with an anti-TGF2 Ab (1/50; Santa Cruz Biotechnology) and an Ab against one of the TGF receptors (I, II, and III) (all Abs 1/100; Santa Cruz Biotechnology). The bound anti-TGF2 Abs were visualized with Cy3-conjugated secondary Abs, anti-rabbit IgG (for receptor III, donkey Abs were used) (Jackson ImmunoResearch Laboratories). The bound anti-TGF receptor (I, II, and III) Abs were visualized with Cy2-conjugated secondary Abs (Jackson ImmunoResearch Laboratories). The images were visualized and photographed on an epifluorescence microscope (Nikon E800).
In vitro assays of Treg cell activity
Proliferation. Naive T cells were exposed to cultured IPE as described above, were harvested, gamma-irradiated (2000 rad), and then added (105 cells/well) to 96-well plates containing fresh T cells (responder T cells; T resp) at 105 cells/well) plus anti-CD3 Ab. Proliferation was assessed by [3H]thymidine incorporation as described above.
Transwell assay to detect soluble factors that suppress bystander T cell activation. The CD8+ IPE Tregs (anti-CD3 pretreated or not) enriched from the T cell IPE cultures described above were harvested, gamma-irradiated, and placed (0.5 x 106 cells/well) into Falcon cell culture inserts (0.4-µm pore size; BD Biosciences) that were inserted into wells containing 0.5 x 106 cells/well naive T cells plus anti-CD3 Abs.
Proliferation when interactions between TGF and TGF receptors are blocked.
Anti-TGF mAbs (anti-TGF1, 2) were added (1 or 10 µg/ml)) to cultures containing anti-CD3-stimulated responder T cells plus gamma-irradiated CD8+ IPE Tregs. As a control, purified mouse IgG (BD Pharmingen) was added to appropriate cultures. In other experiments, purified T cells from wild-type C57BL/6 and from DN TGF RII transgenic mice were added to cultured IPE. After 48 h, the T cells were removed, gamma-irradiated, and added to secondary cultures containing naive responding T cells from wild-type donors plus anti-CD3 Abs. In another set of experiments, purified T cells from wild-type donors were added to cultured IPE. The IPE-exposed T cells were added to secondary cultures containing either naive responding T cells from wild-type or T cells from DN TGF RII donors plus anti-CD3. Proliferation was assessed after 72 h of incubation with [3H]thymidine as described above.
Separation of membrane-bound TGF+ (mTGF+) and TGF– (mTGF–) CD8+ T cells.
Purified enriched CD8+ T cells were cultured for 24 h in the presence of IPE, then harvested and stained as IPE Tregs. Before staining, the T cells were incubated with Fc block. Anti-TGF1, 2 Ab was used to stain IPE Tregs followed by biotin conjugated anti-mouse IgG. MACS beads separated the population into mTGF+ or mTGF– IPE Tregs before staining with FITC-conjugated CD80 (B7-1; BD Pharmingen), CD86 (B7-2; BD Pharmingen), CD152 (CTLA-4) Abs, or control hamster IgG (15). The cells were washed and analyzed by flow cytometry for protein and RT-PCR for mRNA. The level of IPE contamination of harvested T cells was found to be <0.97% with anti-cytokeratin Abs by flow cytometry.
Depletion of mTGF+ T cell population.
The CD8+ IPE Tregs were depleted of mTGF+ T cells by using anti-TGF1/2 and biotin-conjugated anti-mouse IgG and anti-biotin MACS beads as described above. As a control, undepleted CD8+ IPE T regulators were used in these experiments.
Statistical evaluation of results.
Each experiment was repeated at least twice with similar results. All statistical analyses were conducted with Students t test. Values were considered statistically significant if p 
0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
and TGF receptor expression by ocular pigment epithelial cells and tissues
We first examined whether TGF1 and TGF2, as well as TGF RII genes were expressed by cultured PE obtained from normal C57BL/6 mice. Each PE cell type (IPE, CBPE, RPE) expressed easily detectable mRNA for TGF1 and TGF2 (Fig. 1A). We also searched for similar transcripts in excised whole ocular tissues (iris, ciliary body, retina) freshly obtained from eyes of normal C57BL/6 mice (Fig. 1B). In results for TGF RII genes, these tissues all expressed these same transcripts, although expression of the TGF RII gene was barely detected in IPE (Fig. 1C).
|
2, the dominate isoform of TGF in the eye, using immunostaining and bright field microscopy or flow cytometry. Cultured IPE were placed on coverslips and immunostained with anti-TGF2 Abs. Cultured IPE were readily identified on bright field microscopy by the presence of cytoplasmic melanin granules (Fig. 1D). By fluorescence microscopy, these same cells expressed membrane-associated TGF2. The pattern of TGF2 expression was punctate, suggesting either that the cytokine was present in relatively large cell surface patches or perhaps in cytoplasmic granules subjacent to the cell surface. Similarly, cultured IPE expressed surface TGF (46.4% positive; Fig. 1E) and intracellular TGF (89.7% positive; Fig. 1E). Together these results imply that IPE under the culture conditions we used produced active TGF2 in two phases: membrane-associated and soluble.
Detection of TGF and TGF receptor expression by T cells in contact with IPE
Our next goal was to determine the extent to which naive T cells, T cells activated by anti-CD3, and T cells exposed to IPE expressed TGF and its receptors. As revealed in Fig. 2A, uncultured naive T cells as well as anti-CD3-stimulated T cells after 72 h expressed the TGF RII gene. Similarly, T cells exposed to IPE clearly expressed the TGF RII gene (data not shown). Naive splenic T cells contained no identifiable transcripts for either TGF1 or TGF2 (data not shown). Anti-CD3-stimulated T cells expressed easily detectable transcripts for TGF1 but little, if any, TGF2 transcripts (Fig. 2B). In companion experiments, naive T cells were cultured with IPE for 24 h; transcripts of TGF1 and TGF2 were readily detected in these T cells (IPE Tregs; Fig. 2C), as were transcripts of the TGF RII gene (data not shown).
|
protein. No TGF was detected in supernatants of Cont T cells; by contrast, we readily detected significant amounts of TGF in the culture supernatants of IPE T cells (Fig. 2D). Similarly, CD8+ IPE T cells expressed more intracellular TGF (92.9%) than CD8+ control T cells (12.2%; Fig. 2E). Neither T cell population bound the control mouse IgG (<5%; data not shown). Together these results indicate that T cells exposed to IPE acquire the capacity to secrete active TGF, unlike T cells cultured in the absence of IPE. Moreover, IPE Tregs express receptors for TGF, as do conventional T cells.
Patterns of expression of TGF and TGF RII on cocultured IPE cells and T cells
We next examined surface expression of TGF and TGF receptors (RI, RII, RIII) on IPE and T cells that were cultured together for 24 h. Glass coverslips on which these cells were cultured were immunostained with Abs to TGF RI, RII, and RIII and TGF2 and examined by both bright field and fluorescence microscopy. Representative photomicrographs are presented in Fig. 3. Melanin-positive cells (IPE) were readily identified by bright field examination (Fig. 3, A and E); the IPE expressed identifiable TGF2 (Fig. 3, B and F), no evidence of TGF RII (Fig. 3G), and only trace amount of TGF RI (Fig. 3C). TGF RI was barely visible on the IPE cell surface and slightly more on the nuclear membrane (Fig. 3, C and D).
|
RI was well expressed (Fig. 3, C and D), and expression of TGF RII was even more intense (Fig. 3, G and H). The pattern of TGF RII expression on T cells was punctate rather than diffuse, implying that this receptor is distributed unevenly, perhaps in surface patches. The "patches" seemed to localize to sites of contact between T cells and IPE. Moreover, T cell-T cell aggregates were evident in these cultures (Fig. 3H), and punctate staining for TGF RII also marked the points of contact between T cells. It is worth mentioning that T cells exposed to IPE in these cultures expressed TGF RIII, although no evidence of expression of this receptor on IPE was detected (data not shown). Together, these results indicate that T cells destined to become IPE Tregs up-regulate expression of TGF and its signaling receptors. Among the receptors, only TGF RII appeared to localize to points of intercellular contact, between T cells and IPE, and between T cells themselves.
Influence of signaling via TGF RII on CTLA-4 expression by T cells exposed to IPE
We postulated that IPE-derived TGF was important in inducing cocultured T cells to up-regulate CTLA-4. To test our hypothesis, we took advantage of genetically manipulated mice in which a DN TGF RII expressed under a human CD2 promoter/enhancer is present only on T cells (24, 25) where it successfully competes with the T cells’ own signaling TGF RII. Purified CD8+ T cells were obtained from wild-type and DN TGF RII donors, then cultured with or without IPE in the presence of anti-CD3. In control experiments, CD8+ T cells from CD28 KO mice were used in place of the DN TGF RII donors. CTLA-4 expression was detected by flow cytometry. CTLA-4 was up-regulated on anti-CD3-stimulated T cells in the time course (24-h culture, 12%; 48-h culture, 50%), whereas little up-regulation of CTLA-4 was observed if the responding T cells in these cultures were derived from DN TGF RII mice (24-h culture, 2%; 48-h culture, 14%; Fig. 4A). Similarly, CTLA-4 was up-regulated on anti-CD3-stimulated T cells in the absence of IPE (8%); when T cells were similarly stimulated in the presence of IPE, the proportion of T cells expressing CTLA-4 rose dramatically to 51% (Fig. 4B). When the responding T cells in these cultures were derived from DN TGF RII mice, very little up-regulation of CTLA-4 was observed (Fig. 4B), implying that the capacity of anti-CD3 to induce CTLA-4 expression and the capacity of IPE to amplify that expression is dependent on TGF signaling. We observed that the up-regulation of CTLA-4 on T cells obtained from CD28 KO mice was comparable to that achieved by wild-type T cells stimulated with anti-CD3 in the presence of IPE. To confirm this point, expression of CTLA-4 was examined by flow cytometry on T cells stimulated with anti-CD3 in the presence of recombinant TGF. A large proportion (48%) of anti-CD3-stimulated T cells cultured for 24 h in the presence of recombinant TGF expressed CTLA-4, compared with anti-CD3-stimulated T cells cultured without recombinant TGF (6%) (data not shown). These results indicate that up-regulation of CTLA-4 expression on anti-CD3-stimulated T cells is enhanced in the presence of active TGF, further implying that active TGF is produced by cultured IPE.
|
gene in IPE-exposed T cells from DN TGF RII and CTLA-4 KO mice
Many laboratories have reported that a functional link exists between CTLA-4 signaling and T cell production of TGF (15, 17, 18, 23); i.e., T cells fail to express CTLA-4 in the absence of TGF signaling. To explore whether a similar link prevails when IPE induce T cells to convert into IPE Tregs, T cells were obtained from wild-type and genetically manipulated mice: DN TGF RII, CTLA-4 KO. These cells were cultured for 24 h in the presence (or absence) of IPE, then examined for expression of mRNA for TGF1 and TGF2. IPE Tregs obtained from wild-type donors up-regulated mRNA expression of both TGF1 and TGF2 (Fig. 5, A and B). By contrast, neither IPE-exposed T cells from DN TGF RII donors (Fig. 5A) nor CTLA-4 KO donors (Fig. 5B) contained significant levels of transcripts of either TGF isoform. Importantly, the CTLA-4 KO mice were unable to express the TGF2 isoform as shown in Fig. 5B. These results support the postulate that T cells destined to become IPE Tregs receive at least two obligatory signals from IPE: one signal is delivered via TGF RII (and is presumably triggered by TGF derived from the IPE); the other signal is delivered via CTLA-4 and is triggered by B7 molecules expressed constitutively by IPE.
|
RII in enabling IPE to suppress anti-CD3-driven T cell activation in coculture
To test the hypothesis that signaling through TGF RII was required for IPE suppression of T cell activation by anti-CD3, we used T cells that lacked the ability to receive a signal via TGF RII. Purified splenic T cells obtained from DN TGF RII and from wild-type mice were added to wells containing cultured IPE. Mitogenic anti-CD3 Abs were then added (or not) at four different concentrations (0.1, 0.25, 0.5, and 1.0 µg/ml). Proliferation was assessed after 72 h by adding [3H]thymidine, and supernatants were harvested from companion cultures and analyzed by ELISA for content of IFN-. The results indicate that IPE failed to suppress proliferation of DN TGF RII T cells across a wide dose range of anti-CD3, although there was profound suppression of wild-type T cells at all doses (Fig. 6, A and B). Similarly, the results revealed that IPE failed to suppress IFN- production by DN TGF RII T cells stimulated with anti-CD3 (Fig. 6C). Thus, signaling of responding T cells via TGF RII is required for IPE suppression of T cell activation; and TGF, presumably derived from IPE, is essential.
|
and TGF RII in enabling IPE Tregs to suppress bystander T cell activation
In an attempt to validate the hypothesis that IPE Tregs use cell contact and a membrane-associated form of active TGF to suppress bystander T cells, we made use of Transwell inserts that permit soluble molecules, but not cells, to pass between segregated cell populations. CD8+ IPE Tregs (anti-CD3 pretreated or not) were seeded into cell inserts, and bystander naive T cells were seeded below the cell inserts onto the bottoms of the culture wells in the presence of anti-CD3. In control cultures, IPE Tregs were seeded directly into the wells containing naive T cells and anti-CD3. The results of this experiment demonstrated that the efficiency with which IPE Tregs suppressed bystander T cell proliferation was significantly impaired if the Tregs and the bystander T cells were separated by a Transwell membrane (Fig. 7A). When IPE Tregs in the Transwell were treated with anti-CD3 Ab, they suppressed activation of bystander T cells; again the efficiency was more significantly impaired if the Tregs and bystander T cells were separated by a Transwell membrane. These data indicate that direct cell-cell contact optimizes the capacity of IPE Tregs to suppress bystander T cells. We next compared the capacity of IPE Tregs and control T cells to suppress proliferation when added to secondary cultures containing naive T cells, anti-CD3 Abs, and neutralizing anti-TGF Abs (or not). As expected, IPE Tregs profoundly suppressed T cell proliferation, whereas control T cells did not (Fig. 7B). This suppression was partially relieved in a dose-dependent manner in the presence of anti-TGF Abs. These results are compatible with the hypothesis that IPE Tregs bearing membrane-associated TGF lose their regulatory effect if the TGF is neutralized.
|
RII made it possible to determine whether IPE can convert naive T cells into IPE Tregs if TGF RII signaling is curtailed and whether IPE Tregs derived from wild-type donors were capable of inhibiting bystander T cells if the latter were from DN TGF RII donors. T cells from DN TGF RII and wild-type donors were cultured with IPE and added to secondary cultures containing naive T cells plus anti-CD3. When proliferation of bystander T cells was assessed (Fig. 7C), only regulatory T cells created by exposing naive T cells to IPE were inhibitory. In fact, bystander T cells in cultures containing IPE-exposed DN TGF RII T cells displayed enhanced proliferation in response to anti-CD3. We then prepared IPE Tregs, using wild-type T cells, and added these cells (after gamma-irradiation) to secondary cultures containing anti-CD3 plus naive T cells from wild-type or DN TGF RII donors. The results indicated that wild-type IPE Tregs were capable of suppressing bystander T cell activation only when the T cells could accept a TGF signal (Fig. 7D). Bystander T cells bearing DN TGF RII were not suppressed and actually out performed the positive controls. Together these results indicate that T cell signaling via TGF RII is required 1) when naive T cells are converted into regulators by IPE and 2) when bystander T cells are inhibited by IPE Tregs.
Evidence that active TGF is membrane associated in CD8+B7+CTLA-4+ IPE Tregs
IPE achieve suppression of T cell activation, and convert the T cells into regulators of bystander T cells, exclusively via a cell-cell contact mechanism in which interaction of costimulation molecules (B7 and CTLA-4) is required (4, 5). Because supernatants of cultured IPE and IPE Tregs contained active TGF, we wondered whether soluble active TGF or its membrane-bound form was the more effective mechanism of signaling via TGF RII. We analyzed the expression of membrane-associated TGF on IPE Tregs using flow cytometry. The results indicated that CD8+ IPE Tregs expressed TGF on their cell surfaces, whereas control T cells (cultured in the absence of IPE cells) expressed little or no surface TGF (Fig. 8A). We next determined whether membrane-bound TGF-positive IPE Tregs express B7-1, B7-2, and CTLA-4 costimulatory molecules. After being cocultured with IPE, CD8+ IPE Tregs were stained with anti-TGF Abs, then separated into mTGF+ or mTGF– T cell subpopulation. Flow cytometry analyses showed that mTGF+CD8+ IPE Tregs expressed greater levels of CD80 (B7-1-positive cells, 69.8%), CD86 (B7-2-positive cells, 59.7%), and CD152 (CTLA-4-positive cells, 32.1%), whereas mTGF– CD8+ IPE Tregs poorly expressed these molecules on their surface (Fig. 8B). Similarly, mTGF+CD8+ IPE Tregs expressed greater levels of mRNA for CD80 and CD86, whereas mTGF– subsets slightly expressed mRNA for these molecules (data not shown). Thus, IPE-exposed CD8+ T cells up-regulate the expression of B7-1, B7-2, CTLA-4, and mTGF.
|
-depleted CD8+ IPE Tregs were able to suppress bystander T cell activation. Similar to previous experiments, CD8+ IPE Tregs significantly suppressed bystander T cell activation, whereas mTGF-depleted CD8+ IPE Tregs failed to suppress the T cell activation (Fig. 8C). Together these results indicate that B7+CTLA-4+CD8+ IPE T regulators predominantly use the mTGF to achieve the suppression of bystander T cells. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
and TGF RII, the signaling receptor. Based on the evidence presented here, as well as our recently published data, we have formulated the following hypothesis: IPE make use of one of the immune system’s powerful costimulation strategies to target T cells (B7-1 and B7-2 on IPE, CTLA-4 on a subpopulation of CD8+ T cells), and when thus engaged, IPE membrane-associated active TGF is delivered precisely to these T cells. In turn, these CD8+ T cells (IPE Tregs) up-regulate expression of B7-1/B7-2 and TGF1/TGF2. This enables the IPE Tregs to target bystander T cells, again, via direct cell contact, to deliver a membrane-associated TGF signal that suppresses bystander cell activation. We suggest that the capacity of IPE to contribute to immune privilege within the eye depends on their using a novel and highly targeted mechanism to suppress T cell activation, and that the responding T cells themselves adopt this mechanism to suppress other T cells. We speculate that modifying T cell behavior through membrane-associated, rather than soluble, active TGF is important for maintaining the integrity of the eye from the deleterious nonimmune consequences of active, soluble TGF.
TGF is a pleiotropic cytokine/growth factor and its capacity to suppress aspects of immunity is only one dimension of its activities (11, 12, 13). In the absence of TGF, mice develop a massive, multifocal inflammatory disease, suggesting a major role for this factor in regulation of both adaptive and innate immunity (26). Mice in which T cells express a DN TGF RII develop hyperproliferation of CD8+ T cells, implying a role for TGF in maintaining CD8+ T cell homeostasis. CD8+ T cells that express CTLA-4 have regulatory properties and secrete TGF as one of the mediators of suppression (5). It is within this context that the capacity of IPE to suppress T cell activation and convert T cells into regulators takes on meaning. The evidence presented here demonstrates that cultured IPE, along with their counterpart PE from ciliary body and retina, secrete soluble active TGF. In the case of IPE, some of this active molecule was displayed on the cell surface as membrane associated. Cultured IPE expressed very low levels of TGF RII on their surface, implying that autocrine effects are minimal. Moreover, IPE-derived TGF proved to be essential in order for IPE to suppress T cell activation triggered by anti-CD3 Abs. That is, T cells incapable of receiving a TGF signal (DN TGF RII) were not suppressed when exposed to IPE. This latter finding is important in the context of histologic evidence that distribution of surface TGF on IPE was uneven and was localized to punctate areas on adjacent T cells where expression of TGF RII was also colocalized. This provides circumstantial evidence that the TGF delivered to T cells by IPE may be largely, if not exclusively, membrane-associated and that the delivery of TGF is precisely targeted.
Our results suggest that IPE-exposed T cells, become regulatory, use a similar mechanism to mediate suppression of bystander T cells in secondary cultures. CD8+ T cells exposed to IPE up-regulated expression of TGF1 and TGF2, and these cells proved capable of converting latent TGF into the active forms. Moreover, T cell:T cell contact was observed by microscopic examination of primary cultures comprised of IPE and T cells. In these T cell aggregates, membrane-associated TGF was unevenly distributed on one T cell and localized to the point of contact with an adjacent T cell; moreover, the adjacent T cell membrane was enriched for TGF RII expression at the same point of contact. Since IPE Tregs proved incapable of suppressing the activation of bystander T cells that expressed the DN TGF RII and since the efficiency with IPE Tregs suppressed bystander T cells was markedly impaired when the Tregs and the target T cells were separated by a Transwell membrane, we conclude that CD8+ IPE Tregs use primarily a contact-dependent mechanism to suppress bystander T cell activation and that the IPE Tregs deliver active, membrane-associated TGF to the target T cells at the point of contact.
When we determined the CD4/CD8 phenotype of IPE Tregs, we observed that enriched CD8+ IPE Tregs significantly suppressed T cell activation, whereas enriched CD4+ IPE Tregs showed little capacity to suppress T cell activation. In contrast, although both CD4+ and CD8+ T cells were needed to establish RPE Tregs, it was the CD4+ RPE Treg population that was suppressive after exposure to RPE cells obtained from the posterior segment of the eye.
It was interesting to determine that up-regulation of CTLA-4 on T cells destined to become IPE Tregs was dependent on TGF signaling. We believe this to be important because CTLA-4 is the T cell surface molecule that interacts with the costimulation molecule B7 that is constitutively expressed by IPE. We speculate that the contact between CD8+ T cells and IPE, which is required for the generation of IPE Tregs, is mediated by B7-CTLA-4 interactions. Thus, IPE-derived TGF initiates the regulatory process by inducing CTLA-4 up-regulation on CD8+ T cells. Then CTLA-4:ligation provides physical stability of IPE-T cell interactions from which IPE Tregs eventually emerge. In turn, IPE Tregs up-regulate surface expression of B7 and active TGF, and a similar two-step process enables these cells to engage bystander CTLA-4+ T cells that express TGF RII and then to suppress their activation.
There are many recent reports that indicate that regulatory T cells express surface TGF and/or the secreted soluble form (15, 17, 19, 20, 21, 22, 23). Included in this list are natural regulatory CD4+CD25+ T cells (15, 17, 19, 20), CD4+CD25– T cells (21), and CD8+CD25+ T cells (22). However, this is by no means the rule because it has been reported that CD4+CD25+ T regulators suppress in the absence of TGF (27). Nakamura et al. (15) have reported evidence that cell-cell contact is important when natural CD4+CD25+ regulatory T cells inhibit bystander T cells with mTGF, but the molecular basis for "mTGF remains obscure."
In addition to its immunosuppressive and anti-inflammatory effects, TGF displays properties that have been considered to be deleterious to vision (28, 29, 30). In wound healing, TGF promotes the healing response, in part by recruiting macrophages and when excessive TGF can push wound healing toward scarring (31). Whereas scarring in many somatic tissues is not a particularly devastating outcome of wound healing, in the eye, scar formation in the corneal stroma often causes visual impairment and even blindness (32). Gliosis is a prominent feature of optic nerve head pathology in the late stages of glaucoma, and this reactive gliosis contributes to the blinding neuropathy; TGF has been suspected of promoting gliosis in this situation (30). Pertinent to the findings reported in this communication, TGF has been implicated in the changes in the outflow pathway of the eye that lead to elevated intraocular pressure in primary open angle glaucoma (29). A recent, definitive study supporting this view was reported to have produced increased resistance in the outflow path (trabecular meshwork) by infusing active TGF, especially the 1 isoform, into the anterior segment of the eye (28). We interpret these several examples to mean that soluble, active TGF can be deleterious to the eye. However, since TGF has been demonstrated to play a central role in conferring immune privilege upon the eye (1, 2, 3), a conundrum exists: how can TGF promote the dual consequences of preserving vision and yet of threatening quality vision?
Under normal circumstances, levels of total TGF2 in aqueous humor are in the nanogram per milliliter range, whereas the levels of the active isoform are in the low picogram range (14) even with TSP-1, a powerful converter of latent TGF to its active form, is present in aqueous humor (33). We suspect that in the eye IPE succeed in suppressing T cell activation in the anterior chamber, and in converting the T cells into regulators, by delivering active, mTGF to immigrating T cells that are targeted by B7-CTLA-4 interactions. Suppression of sight-threatening T cell effector mechanisms within the anterior chamber is then extended and made pervasive by B7-expressing CD8+ T cells that can deliver their own membrane-bound active TGF to targeted bystander T cells in the ocular microenvironment. In this manner immunogenic inflammation within the anterior segment of the eye is suppressed, i.e., immune privilege is present, and vision is preserved. It is pertinent to this consideration that elevated intraocular pressure and glaucoma are frequent complications of chronic inflammation of the uveal tract (iris, ciliary body, retina, and choriocapillaris) (34). In animal models where inflammation has been induced in the eye, the bulk of the TGF that is present in ocular fluids is in the active rather than latent form, and this TGF is soluble (35). This line of reasoning suggests that under normal circumstances the ocular microenvironment avoids the generation of active, soluble TGF and achieves delivery of the active form through membrane-dependent interactions between cells.
|
Acknowledgments |
|---|
. Finally, we thank Dr. Joan Stein-Streilein for her critical reading of this manuscript. |
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported by U.S. Public Health Service Grant EY 05678. 
2 This article is dedicated to Prof. J. Wayne Streilein, who died on March 15, 2004 (1935–2004).
3 Address correspondence and reprint requests to Dr. J. Wayne Streilein, c/o Dr. Joan Stein Streilein, Schepens Eye Research Institute, 20 Staniford Street, Boston, MA 02114. E-mail address: jstein{at}vision.eri.harvard.edu
4 Abbreviations used in this paper: PE, pigment epithelium; IPE, iris pigment epithelium; CBPE, ciliary body PE; RPE, retina PE; Tregs, T regulatory cells; DN TGF RII, dominant-negative TGF RII; KO, knockout; mTGF+, membrane-bound TGF+; T resp, responder T cell.
Received for publication April 6, 2005. Accepted for publication October 21, 2005.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
-producing regulatory T cells. Invest. Ophthalmol. Vis. Sci. 41: 3862-3870. on cells of the immune system. Ann. NY Acad. Sci. 685: 488-500. [Abstract] superfamily: new members, new receptors, and new genetic tests of function in different organisms. Genes Dev. 8: 133-146. suppresses host defense against bacterial infection in autoimmune MRL/lpr mice. J. Exp. Med. 180: 1693-1703. as an immunosuppressive factor in aqueous humor. Invest. Ophthalmol. Vis. Sci. 32: 2201-2211. . J. Exp. Med. 194: 629-644. (TCF-)-producing regulatory T-cells. Curr. Eye Res. 16: 900-908. [Medline]
) or not (
). These T cells were obtained from wild-type control (A) or DN TGF
). Presented are the mean cpm for triplicate cultures incubated for 72 h ± SEM. *, p < 0.05; comparing two groups. B, To evaluate whether the contact suppression of IPE Tregs is neutralized by anti-TGF