FoxP3⁺CD4⁺CD25⁺ T cells with regulatory properties can be cultured from colonic mucosa of patients with Crohn's disease

Abstract

CD4⁺CD25⁺ regulatory T cells (T_regs) are involved in the maintenance of peripheral tolerance and ensure a balanced immune response competent of fighting pathogens and at the same time recognizing commensals as harmless. This feature is lost in Crohn's disease (CD). The forkhead/winged helix transcription factor FoxP3 is a master gene for T_reg function and defects in the FoxP3 gene lead to a clinical picture similar to inflammatory bowel disease (IBD). Murine colitis can be cured by adoptive transfer of T_regs and ex vivo-generated gut-specific T_regs represent an attractive option for therapy in CD. Thus, defective T_regs could contribute to the development of CD. We cultured biopsies of colonic mucosa in the presence of high concentrations of interleukin (IL)-2 and IL-4 to overcome the anergic nature of naturally occurring CD4⁺CD25⁺ T_regs in the mucosa. We investigated the expression of FoxP3 and regulatory potential of gut-derived CD4⁺CD25⁺ T cells cultured from patients with CD and healthy individuals. The FoxP3 expression was analysed by reverse transcriptase polymerase chain reaction (RT-PCR), and the suppressive effect of FoxP3⁺CD4⁺CD25⁺ T cells on proliferation and cytokine production of autologous CD4⁺ T cells was assessed by flow cytometry. Cultured gut-derived T cells with CD4⁺CD25⁺ phenotype expressed FoxP3 and were able as the freshly isolated T_regs from peripheral blood to suppress proliferation and cytokine production of autologous CD4⁺ T cells. Thus, we demonstrate that FoxP3⁺CD4⁺CD25⁺ T cells with regulatory properties can be propagated in vitro from inflamed mucosa of CD patients, which may be of interest in adoptive immunotherapy.

Introduction

Exposure of the gut mucosa to luminal antigens necessitates regulatory mechanisms to maintain a general state of hyporesponsiveness [1], and animal studies suggest that CD4⁺CD25⁺ regulatory T cells (T_regs) support the immune balance in the normal gut mucosa [2,3]. The role of T_regs in inflammatory bowel disease (IBD) is underlined by a recent study indicating that immune-modifying therapy may restore the thymic output of competent T_regs in murine colitis [4]. A breakdown of tolerance towards the resident bacterial flora plays a pivotal role in the pathogenesis of Crohn's disease (CD) [5] and defective regulatory properties of T_regs could contribute to the disturbed immune balance in CD. However, in a recent study competent T_regs could be isolated from the gut mucosa of IBD patients undergoing colon resection [6].

Hitherto, at least three different CD4⁺ regulatory T cell populations have been described as being capable of inhibiting the T cell response [7], i.e. the inducible interleukin (IL)-10-producing T cells type 1 (Tr1) and transforming growth factor (TGF)-β-producing Th3 cells arising in the periphery and the naturally occurring thymus-derived T_regs[8–10]. The mechanisms by which T_regs mediate their suppressive effect are not fully understood. The regulatory function is mediated by cell-contact-dependent signalling involving cytotoxic T lymphocyte-associated antigen 4 (CTLA-4), glucocorticoid-induced TNF receptor (GITR) and presumably lymphocyte activation gene 3 as well as by secretion of suppressive cytokines [11–13].

Several authors have addressed the difficulty in using non-specific markers for T_regs, because CD25 is also expressed diffusely on the heterogeneous population of recently activated effector T cells [14,15]. Even carefully selected CD4⁺CD25⁺ T cell clones do not represent a homogeneous population of functional suppressor cells [14]. In particular, lamina propria lymphocytes (LPL) should be defined by their suppressive function and not by the mere expression of CD25 and GITR that are also expressed on CD4⁺CD25⁻ LPLs [6].

The forkhead/winged helix transcription factor FoxP3 is a master gene for the function of T_regs[16]. Genetic defects in FoxP3 cause IPEX (immune dysregulation, polyendocrinopathy, enteropathy, X—linked syndrome), a rare disorder which includes severe enteropathy [17]. In humans FoxP3 is expressed mainly in T_regs, although it may not be the definitive discriminative marker for thymus-derived T_regs because FoxP3 expression can be induced in CD4⁺CD25⁻ T cells whereby they acquire regulatory properties [18,19].

CD4⁺CD25⁺ T_regs can be propagated from peripheral blood and retain their regulatory properties in vitro[20]. The rationale of propagating gut-derived T_regs is that they have a high expression of gut-homing integrin α₄β₇ and may show a higher tendency of homing to inflamed gut mucosa than T_regs from peripheral blood with a skin-specific integrin expression [21]. Such gut-derived T_regs may be able to restore the immune balance in CD and in a recent study adoptive transfer of T_regs reversed established colitis in mice [3]. However, it remains unknown whether regulatory cells can ameliorate or even reverse inflammatory disease in humans.

In the present study we use a culture principle with high amounts of IL-2 and IL-4 to expand naturally occurring T_regs from colonic mucosa. We show that mucosal T cells with FoxP3⁺CD4⁺CD25⁺ phenotype and regulatory properties can be cultured from healthy individuals and CD patients. This approach could be instrumental in the functional characterization and therapeutical use of organ-specific T_regs.

Materials and methods

Biopsy specimens

Biopsies from colonic mucosa were obtained from four CD patients (two females and two males, mean age 32 years) diagnosed by established clinical, radiological and histopathological criteria. All biopsies were obtained from inflamed mucosa. The CD patients received azathioprine 100–200 mg daily and 5-aminosalicylic acid (3–4 g daily). For comparison, biopsies were obtained from three healthy controls (one female and two males, mean age 40 years). The study was approved by the Research Ethics Committee of the County of Aarhus (J. no. 1997/3856 and J. no. 1998/4330 with amendment).

Culture of T cells from biopsies of colonic mucosa

The principle of the culture system has been described previously [22]. Biopsies of colonic mucosa were washed twice in RPMI-1640 and placed in 5 ml culture medium consisting of 90% RPMI-1640, 10% human antibody serum, penicillin G 100 U/ml, streptomycin 100 µg/ml, supplemented with 2000 U/ml IL-2 (ProLeukin, Chiron BV, Amsterdam, the Netherlands) and 500 U/ml IL-4 (Pepro Tech, Inc., Rocky Hill, New Jersey, USA) to facilitate the growth of T_regs[10]. Neither antigen nor feeder cells were added to the T cell cultures in order to expand them. Activation was dependent only on in vivo-derived antigen contained in the biopsies. After placing the colonic biopsy in the culture medium, T cells migrate from the mucosal biopsies without needing further T cell purification and the cytokine milieu promotes the preferential proliferation of CD4⁺CD45RO⁺ memory T cells [23]. Within 4 weeks, the T cell culture proliferates to 50 × 10⁶ cells and T cell function is preserved during continuous long-term cultivation [22].

Cytokine measurements

Flow cytometric measurements of interferon (IFN)-γ and tumour necrosis factor (TNF)-α levels were carried out with the human Th1/Th2 cytometric bead array cytokine kit (BD Biosciences, San Jose, CA, USA) according to the manufacturer's protocol. Supernatant (50 µl) was harvested from the same wells in 96-well plates where proliferation studies were performed, thereby obtaining paired information about T cell proliferation and cytokine production.

T cell separation

The T cell cultures were washed with MACS buffer and prefiltered with MACS filter, pore size 30 µm (Miltenyi Biotec, Bergisch Gladbach, Germany). Isolation of CD4⁺ T cells was performed with the CD4⁺ T cell isolation kit on LS⁺ columns according to the manufacturer's instructions. The CD25⁺ fraction was positively selected using CD25 microbeads and MS⁺ columns. To increase the purity of the positively selected CD25⁺ T cells, we applied only 2 µl of CD25 microbeads per 10⁷ cells. The CD4⁺CD25⁻ population underwent a second round of CD25 depletion (20 µl of CD25 microbeads per 10⁷ cells). For some experiments the CD4⁺CD25⁺ regulatory T cell isolation kit was used.

Peripheral blood mononuclear cells were isolated using standard density centrifugation over a Ficoll-Paque gradient (Pharmacia, Uppsala, Sweden) followed by the above-described immunomagnetic separation procedure. The cells were stained with fluorescein isothiocyanate (FITC)-conjugated anti-CD25 after separation to allow flow cytometric assessment of their purity.

Flow cytometry

All antibodies were titrated against recommend isotype controls, and cells were stained with optimal amounts of each antibody for 30 min at 4°C. Data (20 000 events) were acquired on a FACSCalibur® flow cytometer and analysed using Cell Quest® software (Becton Dickinson, San Diego, CA, USA). We used the following antibodies: anti-CD25 FITC (murine IgG₁ clone 2A3), anti-CD4 APC (murine IgG₁ clone SK3) together with relevant IgG subtype-specific isotype controls (Becton Dickinson), anti-GITR phycoerythrin (PE) (murine IgG₁, clone 110416), anti-Fas ligand (FasL), PE (murine IgG_2b, clone 100413) and recommended isotype control (R&D Systems, Inc., MN, USA).

Labelling of T cells with PKH67

This method allows a dynamic assessment of the proliferation in a 96-well plate of targeted populations, without bias from proliferation of the co-cultured population (Fig. 1). The proliferating T cell population is labelled with PKH67 (Sigma-Aldrich Denmark A/S, Copenhagen, Denmark), which binds irreversibly to cell membranes without significantly affecting cell growth [24]. Upon cell division the probe is partitioned between the two daughter cells and with increasing cell doublings the fluorescence intensity of the labelled population decreases (Fig. 1).

Fig. 1.

Upper panel, the freshly isolated PKH67-labelled autologous CD4⁺CD25⁻ peripheral blood lymphocytes (PBLs) proliferate upon stimulation with platebound OKT-3 and soluble anti-CD28, which is illustrated by the gradual movement of the PKH67^bright mother population to a PKH67^dim population of daughter cells (sharing the membrane PKH-labelling upon cell division). After 4 days 75% of the PBLs have undergone mitosis. The proliferation is inhibited by co-culture (ratio 1 : 1) with freshly isolated CD4⁺CD25⁺ PBLs (not labelled, appearing as the indicated PKH-negative population) and cultured gut-derived CD4⁺CD25⁺ T cells (PKH-negative), but not with the cultured CD4⁺CD25⁻ (PKH-negative) counterpart, indicating that suppressive properties are confined to the CD4⁺CD25⁺ T cell population. As indicated, 2% of the PBLs have initially escaped PKH-labelling. The dot-plots are gated on lymphocytes on forward side-scatter as surrogate marker for living cells. This experiment is representative for four Crohn's disease (CD) patients.

Freshly isolated CD4⁺CD25⁻ peripheral blood lymphocytes (PBLs) were labelled with the probe according to the manufacturer's description. Briefly, 15 × 10⁶ cells in 3 ml diluent C were mixed with 12 µl PKH67 dissolved in 3 ml diluent C and stained for 4 min at room temperature. Labelling was terminated by incubation with human antibody serum for 1 min and finally the cells were washed twice in RPMI-1640.

Regulatory T cell functional assay

Assessment of T cell regulatory function was performed with culture conditions, as described by Jonuleit and coworkers [10]. The gut-derived T cell cultures and the PBLs were separated into CD4⁺CD25⁺ and CD4⁺CD25⁻ subsets and cells were adjusted to 5 × 10⁵ cells per ml using a Coulter counter (Beckman Coulter, Inc., Fullerton, CA, USA). The subsets were co-cultured in duplicate for 4 days in 96-well plates (200 µl) in the ratio of one CD25⁺ to one CD25⁻ T cell, the latter being PKH67 labelled.

The wells were precoated with anti-CD3 (Okt-3) 1 µg/ml overnight at 4°C and the cells were cultured in cytokine-free culture media with soluble anti-CD28 1 µg/ml. We tested the competency of the regulatory T cell assay by performing parallel experiments with cultured gut-derived CD4⁺CD25⁺ T cells and freshly isolated autologous CD4⁺CD25⁺ PBLs, the latter serving as positive control for the regulatory T cell assay. Control neutralizing antibodies against IL-10 (5 µg/ml, JES3–19F1, BD Pharmingen) and FasL (4 µg/ml, Nok-1, BD Pharmingen) were added to some of the wells.

FoxP3 expression determined by reverse transcription-polymerase chain reaction (RT-PCR)

RNA was extracted from 10⁶ T cells using an RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The amount and purity of the obtained RNA was determined by measurements of optical density at 260 and 280 nm. The samples were treated with DNase I, amplification grade (Invitrogen A/S, Taastrup, Denmark) before reverse transcription to eliminate contaminating genomic DNA. RT-PCR was performed in a two-step procedure according to the protocol for GeneAmp Gold RNA PCR kit (PE Biosystems, Foster City, CA, USA).

The CDNA synthesis was carried out with at least 3 µg RNA in a final volume of 60 µl using multiscribe reverse transcriptase and random hexamers. The mixture was heated to 25°C for 10 min followed by heating to 42°C for 12 min. Control experiments were carried out without reverse transcriptase and the human embryonic kidney cell lines T293 and K562 served as negative controls. The second-step PCR was performed in a 50 µl reaction volume containing 10 µl CDNA, RNase-free water, 5 µl 10× PCR buffer, 1·75 m M magnesium chloride, 0·8 mM dNTP, 2·5 units Amplitaq Gold DNA polymerase and 25 pmol of each primer. Primer sequences for FoxP3 were: FoxP3 forward: 5′-TCATCCGCTGGGCCATCCTG-3′ and FoxP3 reverse: 5′-GTGGAAACCTCACTTCTTGGTC-3′ (kindly provided by Tobias Bopp, Institute of Immunology, Johannes Gutenberg-University, Mainz, Germany), in addition to previously published FoxP3 primers [19]. After 10 min ‘hot start’ at 95°C, 30 cycles of amplification were followed by 10 min extension at 72°C. Each cycle included denaturation at 94°C for 30 s, annealing at 58°C for 30 s and extension at 72°C for 1 min. The PCR products were run on 1·5% agarose gel stained with ethidium bromide and visualized using an Eagle Eye analyser (Stratagene, La Jolla, CA, USA). As control, mRNA content for GAPDH was analysed using the following primers: GAPDH forward 5′-GTGAAGGTCGGAGTCAACG-3′ and GADPH reverse 5′-CAATGCCAGCCCCAGCG-3′.

Results

Cultured gut-derived CD4⁺CD25⁺ T cells suppress T cell proliferation

The present culture principle selects for CD4⁺CD45RO⁺ memory T cells, which is also the predominant phenotype of LPLs in vivo[6,25]. Others have shown that culture with high IL-2 and IL-4 levels can overcome the anergy of CD4⁺CD25⁺ T_regs from peripheral blood [10] and we investigated the regulatory potential of CD25 expressing T cells in our mucosal T cell cultures. In four CD patients and three healthy individuals cultured gut-derived CD4⁺CD25⁺ T cells and, as expected, freshly isolated CD4⁺CD25⁺ T cells from peripheral blood (being the gold standard for human regulatory T cell function) strongly suppressed the proliferation of anti-CD3/CD28-stimulated autologous CD4⁺ PBLs when co-cultured at a ratio of 1 : 1 (Fig. 1). In contrast, gut-derived CD4⁺CD25⁻ T cells did not suppress proliferation, demonstrating that the regulatory function was contained within the subset of CD4⁺CD25⁺ T cells.

We used autologous CD4⁺ PBLs in our co-culture experiments because the gut-derived CD4⁺CD25⁻ T cells did not proliferate in response to CD3/CD28 stimulation alone, but required the presence of IL-2. Anergy is normally considered a hallmark of T_regs; however, our observation is in line with the hyporesponsive behaviour of freshly isolated CD4⁺CD25⁻ LPLs [6].

As reported by other investigators, neutralizing antibody against IL-10 had no effect on the suppressive capacity of the cultured T_regs[26]. IL-2 driven expansion of CD4⁺ T cells may lead to CD4⁺CD25⁺ T cells with Fas—Fas ligand (FasL)-dependent cytotoxic properties [27]. We found no effect of neutralizing FasL-antibody but high amounts of IL-2 (100 IU/ml) partly reversed the suppressive effect of the CD4⁺CD25⁺ T cells (Fig. 2).

Fig. 2.

CD4⁺CD25⁻ peripheral blood lymphocytes (PBLs) from a patient with Crohn's disease (CD) were labelled with PKH67 (a) and proliferated for 4 days upon anti-CD3/CD28 stimulation (b). Proliferation of the responder population was not inhibited by co-cultured autologous gut-derived CD4⁺CD25⁻ T cells (ratio 1 : 1) (c). Proliferation was inhibited by co-cultured CD4⁺CD25⁺ T cells (d). The suppressive function of the CD4⁺CD25⁺ T cells was not abrogated by neutralizing antibody against Fas ligand (FasL) (e). However, proliferation of the responder population could be partly restored by exogenous interleukin (IL)-2, indicated by the increase in daughter cells (f). Dot-plots gated on living cells.

Gut-derived CD4⁺CD25⁺ T cells suppress cytokine production

CD is considered to be a disease characterized by cytokine imbalance rather than a lymphoproliferative disease, and regulation of cell proliferation may thus not be of primary importance in CD. We therefore addressed whether cultured gut-derived CD4⁺CD25⁺ T cells from CD patients are able to suppress the production of the proinflammatory cytokines TNF-α and IFN-γ. Supernatants were harvested from the same 96-well plates used for proliferation measurements.

During 4 days of co-culture, the gut-derived CD4⁺CD25⁺ T cells and the CD4⁺CD25⁺ PBLs displayed a strong suppressive capacity (Fig. 3). Inhibition of cytokine production was obvious in all experiments where inhibition of proliferation was also observed. However, in two additional experiments we observed significant inhibition of cytokine production without concomitant suppression of proliferation (data not shown).

Fig. 3.

When co-cultured (ratio 1 : 1), gut-derived CD4⁺CD25⁺ T cells suppress the interferon (IFN)-γ and tumour necrosis factor (TNF)-α production of CD3/CD28-stimulated CD4⁺CD25⁻ peripheral blood lymphocytes (PBLs). Within the first 4 days of culture, the suppressive effect is comparable to that of freshly isolated CD4⁺CD25⁺ T cells from peripheral blood. The cytokine levels are means of duplicate wells determined by cytometric bead array (CBA, Human TH1/TH2 kit, Becton Dickinson). Error bars indicate standard error of mean.

Fluctuation of CD25 in mucosal T cell cultures

We observed that CD25 expression of the gut-derived T cell cultures fluctuated and often mirrored the recent activation of effector T cells without regulatory potential. Up-regulation of CD25 occurred simultaneously with a high metabolic activity in T cell cultures. Therefore, the CD25 separation must be supplemented with functional testing of regulatory capacity in order to evaluate T cells with the CD4⁺CD25⁺ phenotype. Contamination by these recently activated effector T cells may obscure the functional testing of putative T_regs and we specifically selected the low-proliferating T cell cultures consisting of 1–5% CD25⁺ T cells for experiments, a strategy similar to authors who isolated CD4⁺CD25⁺ T_regs from peripheral blood [28].

Gut-derived CD4⁺CD25⁺ T cells co-express GITR

The expression of CD25 is not sufficient to define a regulatory T cell population. Studies on human peripheral blood and mice indicate that co-expression of glucocorticoid-induced TNF receptor (GITR) is a hallmark of T_regs[12,14]. In the present study the gut-derived CD4⁺CD25⁺ T cells expressed higher levels of GITR than the CD4⁺CD25⁻ T cell subset (Fig. 4). Surprisingly, the CD4⁺CD25⁻ T cell subset expressed measurable amounts of GITR, which is also a characteristic of freshly isolated CD4⁺CD25⁻ LPLs [6]. We therefore conclude that GITR does not clearly demarcate a population of CD25⁺ T_regs.

Fig. 4.

Flow cytometric evaluation of the immunomagnetic separation of cultured gut-derived T cells from a patient with Crohn's disease (CD). The CD4⁺CD25⁺ T cells expressed higher levels of glucocorticoid-induced tumour necrosis factor receptor (GITR) than the CD4⁺CD25⁻ subset, but equivalent levels of Fas ligand (FasL). Dot plots of 20 000 events gated on living cells.

IL-2-driven expansion of murine CD4⁺ T cells may lead to CD4⁺CD25⁺ T cells with Fas—Fas ligand (FasL)-dependent cytotoxic properties [27]. Therefore, we investigated the level of FasL expression on the IL-2/IL-4-expanded gut-derived T cells. The expression of FasL on the cultured CD4⁺CD25⁺ T cells and the CD4⁺CD25⁻ subset (Fig. 4) were comparable to the level on freshly isolated CD4⁺ PBLs from the same individual.

Cultured gut-derived CD4⁺CD25⁺ T cells express FoxP3

Both murine and human studies suggest that regulatory T cell function depends critically upon the forkhead/winged helix transcription factor FoxP3. IPEX is caused by defects in the FoxP3 gene and the clinical picture may mimic IBD. Thus, following magnetic separation of the gut-derived T cell cultures we evaluated the FoxP3 expression by RT-PCR. FoxP3 mRNA was expressed preferentially in the CD4⁺CD25⁺ T cell subset, in both gut-derived as well as in freshly isolated PBLs from CD patients and healthy individuals. Despite CD25-depletion, we also observed a minor expression of FoxP3 in CD4⁺CD25⁻ T cells, although less than in the CD4⁺CD25⁺ subset (Fig. 5).

Fig. 5.

Nuclear transcription factor FoxP3 is expressed preferentially in the CD4⁺CD25⁺ T cell subset. FoxP3 expression was evaluated by reverse transcription-polymerase chain receptor (RT-PCR) in T cell subsets separated by MACS beads in cultured gut-derived T cells and peripheral blood from a Crohn's disease (CD) patient. The human 293 embryonic kidney cells (T293) and K562 served as negative controls.

Discussion

We have demonstrated that human CD4⁺CD25⁺ T cells with regulatory properties can be propagated from colonic mucosal biopsies from CD patients and healthy individuals. The cultured organ-specific CD4⁺CD25⁺ T cells displayed properties qualitatively comparable to freshly isolated T_regs from peripheral blood, i.e. suppression of T cell proliferation and cytokine production of anti-CD3/CD28-stimulated autologous CD4⁺ PBLs.

The problems of using non-specific markers for T_regs such as CD25 and GITR seem to be even more pronounced in LPLs than in PBLs [6]. We observed that cultured gut-derived CD4⁺CD25⁺ T cells had higher GITR expression than CD4⁺CD25⁻ T cells which, on the other hand, were not devoid of GITR. While the transcription factor FoxP3 seems to be expressed selectively by CD4⁺CD25⁺ TR cells in mice [16], FoxP3 expression in humans is inducible and does not appear to be restricted to the CD25⁺ subset [18, 19, 28]. In our cultured gut-derived T cells we observed that FoxP3 was expressed preferentially, but not exclusively in the CD25⁺ T cells.

Expansion of murine CD4⁺ T cells with IL-2 may generate CD4⁺CD25⁺ T cells with Fas/FasL-mediated cytotoxic properties [27] and human T_regs display perforin-dependent cytotoxicity [26]. Our cultured CD4⁺CD25⁺ T cells express low levels of FasL, comparable to the level on freshly isolated CD4⁺ PBLs, and in antibody-blocking experiments we found no evidence of Fas/FasL-mediated cytotoxicity. On the contrary, proliferation was partly restored by exogenous IL-2.

Our cultured mucosal CD4⁺CD25⁺ T cells meet the criteria to be considered offspring of genuine thymus-derived T_regs; however, we cannot know whether they have been induced from CD4⁺CD25⁻ T cell precursors during culture [29]. On the other hand, this discussion is primarily of academic interest if the goal is to achieve T cells with regulatory properties for therapeutic use.

Reduced function of T_regs may be involved in the breakdown of immunological self-tolerance in patients with multiple sclerosis [30] and in active rheumatoid arthritis [31], thus we investigated whether defective properties of T_regs play a role in CD. We observed no difference between the regulatory capacity of T_regs from CD patients and healthy individuals, neither in freshly isolated T_regs from peripheral blood nor in cultured gut-derived T_regs. In a recent study on 24 patients with CD intact T_regs were recovered from colon resections with inflamed mucosa, which together with our findings suggests that competent T_regs do exist in inflamed gut mucosa as well as in peripheral blood of CD patients. The freshly isolated T_regs from peripheral blood generally exerted a stronger suppression on T cell proliferation and cytokine production than the cultured CD4⁺CD25⁺ T cells from gut mucosa. Importantly, this was true for both CD patients and healthy individuals and could reflect diminished regulatory capacity during cultivation or more probably gut-specific properties, as freshly isolated T_regs from lamina propria have been observed to be less suppressive than their equivalent in peripheral blood [6].

Special culture conditions may be required to unravel regulatory T cell defects [30] and we cannot conclude whether our observations reflect all aspects of regulatory T cell function, or whether our findings can be extended to in vivo conditions. Secondly, we present data from mixed polyclonal T cell populations and not from clonally selected T cells. Others have shown that it may be difficult to observe pure regulatory T cell characteristics unless experiments are performed with T_reg clones [14].

The existence of intact T_regs in CD patients is a prerequisite for ex vivo propagation and therapeutic use. The idea of using T cells from colonic biopsies is to take advantage of the presumed intrinsic gut-homing potential of mucosal T cells, which we have shown to be stronger than that of PBLs, as judged by their adhesion to mucosal addressin adhesion molecule 1 (MAdCAM-1) in vitro[32]. While CD4⁺CD25⁺ T_regs from peripheral blood mainly carry skin-specific homing markers [21], our cultured gut-derived T cells including the CD4⁺CD25⁺ subset have a high expression of the gut-homing integrin α₄β₇. Although no gut-specific homing of such T cells was observed in healthy individuals [32], the gut-specific integrin expression would presumably make them prone to target the inflamed gut [33]. Of particular interest, the expression of integrin α₄β₇ has been reported to define a population of CD4⁺CD25⁺ T_regs which specifically induce IL-10-producing Tr1-like cells, whereas expression of integrin α₄β₁ induce Th3-like cells [28].

In conclusion, our T cell culture system provides the potential to cultivate CD4⁺CD25⁺ T cells with regulatory properties. However, until now we have not been able to expand CD4⁺CD25⁺ T cells with regulatory properties beyond 5% (10⁷ cells) of the entire mucosal T cell population without excessive contamination by recently activated CD25⁺ effector T cells. Thus, so far, we cannot generate sufficient T_regs by this method, as the number of T_regs required for adoptive immunotherapy has been estimated to be as high as 10⁹[34].

We have shown that cultured autologous gut-derived T cells can be infused in humans without side effects [32], but it remains unresolved whether established autoimmune/inflammatory disease in humans can be limited or even reversed by T_regs. Adoptive cellular therapy with CD4⁺CD25⁺ T_regs will require rapid and large-scale expansion in vitro and the future challenge will be to drive the cultured T cells in a regulatory direction.