We have identified a large population of CD37+ cells in human fetal gut. Three- and four-color flow cytometry revealed a distinct surface Ag profile on this population; the majority were negative for CD4 and CD8, whereas most of the remainder expressed the CD8αα homodimer. In contrast about half of CD3+ cells expressed CD4 and half expressed CD8α. A large proportion of CD37+ cells expressed CD56, CD94, and CD161, and whereas CD3+ T cells also expressed CD161, they only rarely expressed CD56 or CD94. Further studies were conducted to determine whether the CD37+ cells have the potential to differentiate into CD3+ cells. About half of CD37+ cells contain intracellular CD3ε. Rearranged TCR γ-chains were detected in highly purified CD37+ cells as an early molecular sign of T cell commitment, and the pattern of rearrangement with V regions spliced to the most 5′ Jγ segment is reminiscent of early thymocyte differentiation. In reaggregate thymic organ cultures, CD37+ cells also gave rise to CD3+ T cells. Thus, we demonstrate that the CD37+ cells present in the human fetal gut display a distinct phenotype and are able to develop into CD3+ T cells.

In mice, numerous studies have shown the development of gut lymphocytes, particularly intraepithelial lymphocytes (IELs) 3 in the absence of a thymus, generating mainly γδ T cells (1, 2, 3, 4). Cryptopatches, tiny clusters of lymphocytes within the murine lamina propria, have also been described as source of extrathymic gut lymphocytes in euthymic and athymic mice (5, 6). Recently, mesenteric lymph nodes and Peyer’s patches have been shown to be sites of extrathymic T cell development in athymic mice but not in normal mice (7).

On the other hand, in humans, little is known about T cell development at extrathymic sites such as liver and gut. Hemopoietic progenitor cells can be detected in fetal liver and fetal bone marrow from 5 wk of gestation, at which time early T- and thymus-independent NK cell precursors start to express CD7 (8). The human thymus is then colonized by CD7+ emigrants between 7 and 8 wk of gestation, where they go through well-defined stages of maturation and develop into T cells (9, 10).

It is generally considered that T cells populate the human intestine in response to bacterial and food Ags. However in the lamina propria of human fetal intestine, even in utero the largest organ in the body, T cells can be identified by immunohistochemistry at 12–14 wk of gestation (11, 12), and their number increases rapidly with gestational age, so that by 19–22 wk of gestation, the density of T cells in some parts of the lamina propria approaches that of postnatal bowel (13). Although it is possible that these cells are being elicited by foreign Ags in amniotic fluid, we consider that alternative explanations, such as local differentiation from local T cell precursors is a possibility. A large proportion of fetal gut lamina propria T cells are dividing and express markers of activation, such as HLA-DR, αEβ7, and CD45RO, commonly associated with memory T cells that have previously undergone antigenic stimulation (13). The cells also lack the α4β7 integrin (13) by which circulating cells home into the lamina propria (14), again suggestive of a local rather than blood-borne origin.

Previous studies have detected TCR-β mRNA transcripts in fetal intestine at 14 wk of gestation (15), concurrently with pre-TCR-α (pTα) (13), which associates with the TCR β-chain at the cell surface before recombination of the TCR α-chain (16). Thus, TCR gene rearrangement seems to take place within the fetal intestine, but it is unclear whether this represents local differentiation into T cells or immature thymic emigrants that home to the gut. We have also previously shown that following transplantation of human fetal intestine into SCID mice, pTα mRNA transcripts were still present in the fetal gut specimens after an engraftment period of 30 days (13). In thymus, pTα-expressing pro-T cells have a lifespan of 3 days (17), making it unlikely that the pTα transcripts in the grafts were from residual cells. Thus, there might be a yet unidentified T cell progenitor population contributing to T cell generation in human fetal gut. A likely candidate is the population of CD37+ cells abundant in the fetal lamina propria (13), which might represent a population directly derived from fetal hemopoietic sites.

In this study, we use double immunohistochemistry, flow cytometry, analysis of TCR γ-chain rearrangement, and transfer of sorted CD37+ cells into murine reaggregate thymic organ cultures (RTOCs) to characterize the population of CD37+ cells in the human fetal gut, and provide strong evidence that these CD37+ cells are able to develop into mature CD3+ T cells.

Human fetal gut samples were obtained from the Medical Research Council Tissue Bank, Hammersmith Hospital (London), and the Princess Ann Hospital (Southampton). The study was approved by the Southampton and Southwest Hampshire Local Research Ethics committee. The fetal intestinal specimens (n = 51) were between 7 and 20 wk of gestation (conceptional age) as assessed by hand or foot length measurements. Adult normal ileum specimens (n = 5) and colon specimens (n = 16) were obtained from surgical resections for bowel malignancies. Tissue specimens were taken at >10 cm distance from the focal lesion. All samples were macroscopically and microscopically normal. Tissue was immediately used for isolation of lymphocytes as well as snap frozen in liquid nitrogen for immunohistochemistry.

Cryostat sections were cut at 6 μm and fixed in 4% paraformaldehyde for 20 min. CD37+ cells were detected by sequential double immunohistochemistry (18) using a peroxidase-conjugated secondary and diaminobenzidine with the first Ab to give a brown reaction product and an alkaline phosphatase-conjugated Ab and fast blue with the second Ab to give a blue reaction product. Abs were CD3 (UCHT-1; Dako), CD7 (DK24; Dako), HRP rabbit anti-mouse IgG (Dako), and alkaline phosphatase rabbit anti-mouse IgG (Dako).

For all immunohistochemistry experiments, negative controls included using secondary Abs alone or irrelevant isotype controls.

For quantification of CD37+ and CD3+ cell subsets, areas of lamina propria were mapped using a computer mouse, and the cell density was calculated per square millimeter by image analysis.

LPLs from fetal gut specimens and adult normal colon specimens were isolated as previously described (19). Briefly, epithelial cells and IEL were removed by incubation for 3 periods of 15 min (fetal tissue) or up to 6 periods of 30 min (adult tissue) under intense stirring with 1 mM DTT and 1 mM EDTA in HBSS (Invitrogen Life Technologies). The remaining tissue was digested using collagenase type I (Sigma-Aldrich) for 1–2 h at 37°C in RPMI 1640 medium (Invitrogen Life Technologies) supplemented with 10% FCS and antibiotics. The cell suspension was then subjected to further purification on a Percoll (Sigma-Aldrich) density gradient containing 30–40–60–100% layers. The cell fraction between 40 and 60% was collected and washed twice in HBSS. The number of cells obtained from each fetal gut specimen varied from 1 × 106 to 5 × 106 cells, depending on the age. From adult gut specimens 5 to 10 × 106 cells were recovered. More than 95% of the cells were viable as judged by trypan blue exclusion test.

Freshly isolated fetal and adult LPLs were incubated for 20 min in PBS with 10% human AB serum to block Fc receptors. The cells (2 × 105) were labeled by three- and four-color staining procedure using FITC-, PE-, Cy5-, and allophycocyanin-conjugated mAbs according to standard procedures and were analyzed on a FACScan (BD Biosciences). Abs were CD3ε (UCHT1; Dako), CD4 (B-F5; Diaclone), CD7 (DK24l Dako), CD8 (DK25l Dako), CD8β (2ST8.5H7; Coulter-Immunotech), CD16 (B-E16; Diaclone), CD34III (B-K34; Diaclone), CD56 (B-A19; Diaclone), CD94 (HP-3D9; Dako), CD103 (Ber-ACT8; Dako), CD161 (DX12; BD Biosciences), TCRαβ (T10B9.1A-31; BD Biosciences), TCRγδ (B1.1; BD Biosciences), and Vα24 (C15, Coulter-Immunotech). Fluorochrome-conjugated isotype control Igs were used for all analyses.

The intracellular staining for CD3ε was performed with a commercial kit Cytofix/Cytoperm (BD Biosciences). Cells were labeled on the surface with the CD7-FITC and CD3-PE. They were then fixed and permeabilized (Cytofix/Cytoperm) for 20 min, washed with WashPerm (BD Biosciences), and labeled with an anti-CD3ε Ab conjugated to allophycocyanin or isotype-matched Ig control.

For sorting, up to 5 × 106 fetal LPLs as well as up to 20 × 106 adult LPLs were labeled in two colors using CD7-FITC and CD3-PE. Subsets of CD37+cells and CD3+7+cells (1–8 × 104) were obtained by two rounds of sorting and reached a purity of 99–99.5%. Immediately afterwards, they were used to generate RTOCs or snap frozen for PCR.

Thymic stromal cell suspensions were prepared from 2-deoxyguanosine-treated BALB/c fetal thymic rudiments as previously described (20, 21). Freshly isolated and sorted fetal CD37+ cells or CD3+7+ cells were mixed together with the murine stromal cells (1 × 106) in 1.5-ml Eppendorf tubes and pelleted by centrifugation. Following removal of the supernatant, the cell pellet was carefully transferred to the surface of a 0.8-mm Nucleopore filter (Corning Costar UK) in organ culture. Cells were harvested after 14 days and immediately stained with fluorochrome-labeled Abs and analyzed on the FACScan. As negative control, empty murine stromal cell cultures were processed in the same way.

Subsets of CD37+cells and CD3+7+cells were lysed with proteinase K (Promega) at 56°C for 1 h. Lysates were used immediately or stored at −80°C for a maximum of 4 wk.

Primers for the detection of the Vγ1 to Vγ8 gene segments, the Vγ9 as well as Vγ10 and Vγ11 gene segments were used in conjunction with either Jγ or JP primers as described by McCarthy et al. (22).

The TCR γ-chain PCR was performed in a total volume of 50 μl consisting of 0.2 μM of each primer, 200 μM of each dNTP, 1.5 μM MgCl2, 1× Taq Gold buffer (Applied Biosystems), 2 U of Taq Gold polymerase (Applied Biosystems), and 10 μl of the DNA lysate, covered with 50 μl of paraffin oil.

Thirty-five cycles of Jγ primer-containing PCR mix and 40 cycles of the JP primer-containing PCR mix were performed with an initial denaturation step (95°C for 10 min), each cycle consisting of a denaturation step (94°C for 1 min), an annealing step (55°C for 1 min), and an extension step (72°C for 1 min). PCR products were visualized on ethidium bromide-stained 10% Tris borate-EDTA polyacrylamide gels (Bio-Rad) generating bands in the size range of 70–110 bp. Water was used as a negative control, and for a positive control, DNA from a T cell lymphoma cell line (Jurkat) were run in parallel with each PCR.

PCR products of CD37+cells as well as CD3+7+cells from two fetal gut samples were ligated into the pGEM-T Easy vector using the pGEM-T Easy vector kit (Promega) according to the manufacturer’s instructions and cloned into JM 109 competent cells. Plasmid minipreps from randomly selected colonies were used for sequencing.

Double immunohistochemical staining of cryostat sections from 29 human fetal gut samples aged 7–20 wk gestation, five adult ileum samples and six adult colon samples showed numerous CD37+ cells within the lamina propria of all samples. The cell density was remarkably stable in fetal specimens of all ages as well as in adult intestine (Fig. 1 A). CD37+ cells were also occasionally seen within the epithelium.

FIGURE 1.

Distribution of CD37+ and CD3+7+ cells in human fetal and adult gut. A, CD37+ and CD3+ cell density (log scale) in the lamina propria of fetal intestine with increasing gestational age (7–8 wk, n = 4; 9–10 wk, n = 9; 11–13 wk, n = 7; 14–16 wk, n = 5; 17–20 wk, n = 4) and adult ileum (n = 5) and colon (n = 6). Bars represent the mean ± SD. B, Flow cytometric analysis of LPLs using CD7-FITC and CD3-PE mAbs on fetal gut specimens aged 7, 10, and 14.1 wk as well as an adult colon. The lymphoid population was gated based on side scatter and CD7 expression (not shown).

FIGURE 1.

Distribution of CD37+ and CD3+7+ cells in human fetal and adult gut. A, CD37+ and CD3+ cell density (log scale) in the lamina propria of fetal intestine with increasing gestational age (7–8 wk, n = 4; 9–10 wk, n = 9; 11–13 wk, n = 7; 14–16 wk, n = 5; 17–20 wk, n = 4) and adult ileum (n = 5) and colon (n = 6). Bars represent the mean ± SD. B, Flow cytometric analysis of LPLs using CD7-FITC and CD3-PE mAbs on fetal gut specimens aged 7, 10, and 14.1 wk as well as an adult colon. The lymphoid population was gated based on side scatter and CD7 expression (not shown).

Close modal

Some CD3+ cells were found in fetal gut at early gestational age, and their number increased steadily up to 20 wk gestation, where they reached approximately one-third of the number seen in postnatal specimens (Fig. 1 A) (13). Interestingly, in the two youngest fetal gut samples aged 7.0 and 7.5 wk gestation, just around the time of thymic colonization, many CD37+ cells were present, whereas no CD3+ cells were seen.

The results were confirmed by flow cytometry. LPLs isolated from 22 fetal gut samples aged 7–14.1 wk gestation and 10 adult colon specimens were stained with fluorochrome-labeled CD3 and CD7, and representative plots are shown (Fig. 1 B). The lymphoid population was gated on side scatter and CD7 expression. In fetal gut at early gestational age, the vast majority of cells were CD7 single-positive and no cells expressed CD3. With progressing gestational age, the proportion of CD3/7 double-positive cells increased. In adult colon, the pattern was reversed with the majority of CD7-positive cells expressing CD3 and only a small percentage being single-positive for CD7.

LPLs isolated from six fetal gut samples (10–14 wk gestation) were characterized by three- and four-color flow cytometry. CD37+ and CD3+7+ cells were analyzed for CD34, CD103, CD4, CD8αα, and CD8αβ expression, their CD4/8 double-negative (DN) and CD4/8 double-positive (DP) status, as well as the TCRαβ and TCRγδ expression (Table I).

Table I.

Characterization of fetal CD37+ and CD3+7+ cellsa

CD34CD103CD4CD8DNDPTCRαβTCRγδ
CD37+ 29 ± 5.7b 41 ± 9.8 13 ± 3.7 39 ± 4.1 49 ± 3.8 3 ± 1 0.8 ± 0.7 0.51 ± 0 
CD3+7+ 24 ± 6.5 34 ± 7.2 46 ± 7.2 49 ± 7.7 20 ± 12.7 23 ± 5 77 ± 14.4 25 ± 0 
CD34CD103CD4CD8DNDPTCRαβTCRγδ
CD37+ 29 ± 5.7b 41 ± 9.8 13 ± 3.7 39 ± 4.1 49 ± 3.8 3 ± 1 0.8 ± 0.7 0.51 ± 0 
CD3+7+ 24 ± 6.5 34 ± 7.2 46 ± 7.2 49 ± 7.7 20 ± 12.7 23 ± 5 77 ± 14.4 25 ± 0 
a

Studied in six fetal gut samples of 10—14 wk gestation.

b

Values are expressed as arithmetic means ± SD of percentages.

A substantial proportion of CD37+ and CD3+7+ cells stained for the stem cell marker CD34 with moderate to bright intensity, suggesting recent arrival from fetal hemopoietic sites (Table I). However, in adult gut, only few LPLs expressed CD34 at low intensity level (data not shown). The integrin CD103, which recognizes the epithelial ligand E-cadherin, was also detected in a significant proportion of fetal CD37+ and CD3+7+ lamina propria cells (Table I).

Among the fetal CD37+ cells, only a minor fraction expressed CD4 (13 ± 3.7), whereas a large percentage of the CD3+ cells were CD4 positive (46 ± 7.2; Table I, Fig. 2,A). CD8 expression among the CD37+ cells was high (39 ± 4.1; Table I, Fig. 2,A). The vast majority of CD37+ CD8+ cells in the fetal lamina propria of all samples expressed the CD8αα homodimer (95 ± 2.3; Fig. 2 B).

FIGURE 2.

Phenotypic characterization of human fetal LPLs. Three-and four-color flow cytometric analysis of fetal LPLs. A, Subsets of CD3+7+ cells (top row) and CD37+ cells (bottom row) were analyzed for their CD4 and CD8 expression, their CD4/8 DN status as well as their TCR αβ and TCR γδ expression. The results are representative of six separate experiments. B, Subsets of CD3+7+ cells (top row) and CD37+ cells (bottom row) were analyzed for their CD8α and CD8β expression. Among the CD3+7+8+ cells, the proportion of CD8αα-expressing cells in this population declined from 52 to 10% with increasing gestational age, associated with an increment in CD8αβ-expressing cells. The vast majority of CD37+8+ cells were CD8αα homodimeric in all tested fetal gut samples. The results are representative of five separate experiments.

FIGURE 2.

Phenotypic characterization of human fetal LPLs. Three-and four-color flow cytometric analysis of fetal LPLs. A, Subsets of CD3+7+ cells (top row) and CD37+ cells (bottom row) were analyzed for their CD4 and CD8 expression, their CD4/8 DN status as well as their TCR αβ and TCR γδ expression. The results are representative of six separate experiments. B, Subsets of CD3+7+ cells (top row) and CD37+ cells (bottom row) were analyzed for their CD8α and CD8β expression. Among the CD3+7+8+ cells, the proportion of CD8αα-expressing cells in this population declined from 52 to 10% with increasing gestational age, associated with an increment in CD8αβ-expressing cells. The vast majority of CD37+8+ cells were CD8αα homodimeric in all tested fetal gut samples. The results are representative of five separate experiments.

Close modal

For CD3+ cells, 49% were CD8 positive (Table I, Fig. 2,A). With increasing gestational age, the proportion of CD8αα-expressing cells in this population declined from 52 to 10%, associated with an increment in CD8αβ-expressing cells (Fig. 2 B).

Within the CD37+ cell subset, a large percentage were DN (49 ± 3.8) and only a minor fraction was DP (3 ± 1; Table I, Fig. 2,A). In the CD3+7+ cell population, the proportion of DN cells was highly variable (20 ± 12.7), and a substantial fraction was DP (23 ± 5; Table I, Fig. 2 A).

As expected, CD37+ cells of all fetal gut samples were TCRαβ and TCRγδ negative (Table I, Fig. 2,A). In the CD3+ cells, the majority showed TCRαβ expression (77 ± 14.4), and the remaining fraction TCRγδ expression (25 ± 0; Table I, Fig. 2 A). In three additional fetal gut samples (10.6–14 wk gestation) only a small percentage of the CD3+ T cells expressed the invariant Vα24-chain (1.8 ± 1.4).

In human adult small intestine, a CD37+ population is present in the gut epithelium that expresses NK markers (18, 23, 24), and a subset displays intracellular CD3ε found in NK and T progenitor cells (25, 26). We first investigated LPLs isolated from four human fetal gut samples of 12.8–13.3 wk gestation for their NK marker expression by flow cytometric analysis.

In the CD37+ population, only a minor fraction were CD16 positive (13% ± 8.6), whereas a substantial proportion expressed CD56 (48% ± 7) with moderate to bright intensity (Fig. 3,A). Two NK receptors of the C-lectin-like family were expressed by a large fraction of CD37+ cells. CD94, a glycoprotein that forms heterodimers with molecules of the NKG2 family and recognizes the nonclassical MHC class I molecule HLA-E, was present on 44% ± 8.5 of the CD37+ cells. The NKR-P1A receptor (CD161) was expressed on 70% ± 8.8 of CD3, 7+ cells (Fig. 3,A). CD3+ T cells showed a different pattern. Only few cells expressed CD16 (8% ± 4.6), CD56 (19% ± 20.1), and CD94 (12% ± 2.6). However, CD161 was seen on 41% ± 8.7 (Fig. 3 A).

FIGURE 3.

Fetal CD37+ cells and CD3+ T cells express distinct patterns of NK receptors. A, Subsets of CD3+7+ cells (top row) and CD37+ cells (bottom row) were analyzed for their NK receptor expression. The results are representative of four separate experiments. B, A population of CD37+ cells was stained with the NK markers CD94 and CD161. CD37+94+ cells (top row) and CD37+161+ (bottom row) were analyzed for their CD4 and CD8 expression. C, CD3+7+161+ cells were analyzed for their CD4 and CD8 expression. The results are representative of five separate experiments.

FIGURE 3.

Fetal CD37+ cells and CD3+ T cells express distinct patterns of NK receptors. A, Subsets of CD3+7+ cells (top row) and CD37+ cells (bottom row) were analyzed for their NK receptor expression. The results are representative of four separate experiments. B, A population of CD37+ cells was stained with the NK markers CD94 and CD161. CD37+94+ cells (top row) and CD37+161+ (bottom row) were analyzed for their CD4 and CD8 expression. C, CD3+7+161+ cells were analyzed for their CD4 and CD8 expression. The results are representative of five separate experiments.

Close modal

Having demonstrated that a substantial proportion of fetal CD37+ cells express NK markers, particularly CD94 and CD161, we further characterized the CD37+ CD94+/161+ cells for CD4 and CD8 expression using LPLs of five fetal gut samples (9.8–12.6 wk gestation). Interestingly, CD8α was expressed by the majority of CD37+ CD94+ cells (73 ± 1.5, Fig. 3,B) and a large proportion of the CD37+ CD161+ cells (40 ± 11.6, Fig. 3,B), whereas CD4 expression was low (6 ± 3.4 and 3 ± 2.4, Fig. 3,B). However, in CD3+161+ cells, the pattern was different. A large proportion expressed CD4 (56 ± 17) and CD8 (32 ± 2, Fig. 3 C).

Having demonstrated that CD37+ cells already exists very early in the fetal lamina propria around 7–8 wk of gestational age, we investigated the NK marker expression and CD8 expression on four samples aged 7.0–9.5 wk gestation. Because there were essentially no CD3+ cells in these samples, only CD37+ cells were analyzed, and results are shown in Table II. A small fraction of early fetal CD37+ cells expressed CD16, whereas a substantial proportion expressed CD56 and CD161. CD16 and CD56 expression seemed to be similar in early fetal gut samples compared with the older fetal gut samples around 13 wk of gestational age described above (Table II, Fig. 3,A). In contrast, CD161 expression was lower among the early fetal gut samples than in the older ones. CD94 expression was very low in the two youngest fetal gut specimens and increased dramatically with progressing gestational age (Table II, Fig. 3,A). The CD8 status of these samples was tested and showed a similar pattern. Low numbers of CD8-expressing cells were seen in the two youngest fetal gut samples with increasing numbers in the older fetal samples (Table II). CD34 expression was also particularly high in the two 7-wk-old specimens (Table II).

Table II.

Expression of NK receptors, CD8, and CD34 by early fetal CD37+ cellsa

CD16CD56CD94CD161CD8CD34
Fetal gut       
 7.0 wk 23b 72 1.4 42 2.5 43 
 7.5 wk 36 56 3.4 25 52 
 9.0 wk 13 39 19 56 17 26 
 9.5 wk 11 49 17 48 22 33 
CD16CD56CD94CD161CD8CD34
Fetal gut       
 7.0 wk 23b 72 1.4 42 2.5 43 
 7.5 wk 36 56 3.4 25 52 
 9.0 wk 13 39 19 56 17 26 
 9.5 wk 11 49 17 48 22 33 
a

Studied in four fetal gut samples of 7.0–9.5 wk gestation.

b

Values are expressed as percentages of CD37+ cells.

CD37+ LPLs isolated from four human fetal gut samples of 14–16 wk gestation were investigated for their expression of intracellular CD3ε by flow cytometric analysis. Intracellular CD3ε expression was seen in a substantial fraction of the CD37+ cells (42 ± 17, Fig. 4).

FIGURE 4.

CD37+ cells express intracellular CD3ε. CD37+ cells were analyzed for their intracellular CD3 (icCD3) expression by three-color flow cytometric analysis. A substantial fraction of CD37+ shows icCD3. The results are representative of four separate experiments.

FIGURE 4.

CD37+ cells express intracellular CD3ε. CD37+ cells were analyzed for their intracellular CD3 (icCD3) expression by three-color flow cytometric analysis. A substantial fraction of CD37+ shows icCD3. The results are representative of four separate experiments.

Close modal

To investigate whether CD37+ cells have the potential to develop along the T cell lineage, we looked at TCR γ-chain rearrangement as an early molecular sign of T cell lineage commitment. LPLs were isolated from six fetal gut samples of 11.9–14.1 wk gestation and five adult colon specimens. Both subsets of CD37+ and CD3+7+ cells were obtained by two rounds of sorting reaching a purity of >99.5% (Fig. 5, A and B). The PCR for the TCR γ-chain rearrangement consists of two different reactions. In the VγJP reaction, rearrangements using the J segments JP, JP1, or JP2 are detected, whereas rearrangements using the Jγ1 or Jγ2 segment are seen in the VγJ reaction. Rearranged TCR γ-chains were detected within the CD37+ cell subset of all fetal gut samples and 3/5 adult colon samples in the VγJP reaction. The rearrangement was polyclonal (Fig. 5 C). In the VγJ reaction, two of six fetal CD37+ samples and one of six adult CD37+ samples displayed a weak signal.

FIGURE 5.

CD37+ cells show TCR γ-chain rearrangement. Flow cytometric analysis of fetal LPLs using CD7-FITC and CD3-PE Abs before sorting (A) and after two rounds of sorting (B) reaching a purity of >99%. C, Rearranged TCR γ-chains are detectable within the CD37+ cell subset and the CD3+ cell subset of fetal gut samples and adult gut samples in the VγJP reaction. PCR products are generated in the size range 70–110 bp (see marker lane M) and the rearrangement appears to be polyclonal in both subsets, generating bands in the CD37+ subset (arrowheads) and a broad smear in the CD3+ subsets. Outside lane, negative control (neg).

FIGURE 5.

CD37+ cells show TCR γ-chain rearrangement. Flow cytometric analysis of fetal LPLs using CD7-FITC and CD3-PE Abs before sorting (A) and after two rounds of sorting (B) reaching a purity of >99%. C, Rearranged TCR γ-chains are detectable within the CD37+ cell subset and the CD3+ cell subset of fetal gut samples and adult gut samples in the VγJP reaction. PCR products are generated in the size range 70–110 bp (see marker lane M) and the rearrangement appears to be polyclonal in both subsets, generating bands in the CD37+ subset (arrowheads) and a broad smear in the CD3+ subsets. Outside lane, negative control (neg).

Close modal

The CD3+ cell subsets of all fetal gut and all adult colon samples showed TCR γ-chain rearrangement generating a polyclonal smear in both the VγJ and VγJP reaction (Fig. 5 C).

PCR products of the VγJP reaction from two fetal gut samples were cloned and sequenced (Table III). Twelve sequences were obtained from the 37+ subsets and 13 sequences from the 3+7+ subsets. Different V segments were used in both subsets and N/P nucleotide insertions were present. Apart from one TCR γ-chain rearrangement using the JP2 segment, all other rearrangements involved the JP1 segment. Sequences of rearranged TCR γ-chains were diverse. Within CD37+ and CD3+ cells, respectively, 9 of 12 and 11 of 13 sequences were represented only once.

Table III.

DNA sequence analysis of the TCR γ-chain rearrangement from fetal gut LPLsa

Seq. nobV SegmentV SequenceN/PJ SequenceJ Segment
37+ CACCTGGGACGGG GTAGTGATTGGATCAAGACGTTTGCAAAAG JP2 
 1/3/5/7 CACCTGGGACA TG CCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/3/5/7 CACCTGGGACAGG  CCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 CACCTGGGATGGG  CCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 CACCTGGGATA TA TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/3/5/7 CACCTGGGACAGG TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/3/5/7 CACCTGGGACA TATACCNCGTGT GTTGGTTCAAGATATTTGCTGAAG JP1 
 CACCTGGGACG CTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 9–11 CACCTGGGATAGG CCA TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 12 1/3/5/7 CACCTGGGACAGG CT TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
3+7+ 1/3/5/7 CACCTGGA  TGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/2/3/4/5/7 C?GCCTGGG CC GGTTGGTTCAAGATATTTGCTGAAG JP1 
 3, 4 CACCTGGGACGGG CCTTA TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 CACCTGGGATA  TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/2/3/4/5/7 CACCTGGGA  TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 CACCTGGGATAG TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/3/5/7 CACCTGGGACAG TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/3/5/7 CACCTGGGACA  TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 10  CACC  TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 11  CACC CCA TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 12  CACCTG GA TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 13 CACCTGGGATAGG  CCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
Seq. nobV SegmentV SequenceN/PJ SequenceJ Segment
37+ CACCTGGGACGGG GTAGTGATTGGATCAAGACGTTTGCAAAAG JP2 
 1/3/5/7 CACCTGGGACA TG CCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/3/5/7 CACCTGGGACAGG  CCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 CACCTGGGATGGG  CCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 CACCTGGGATA TA TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/3/5/7 CACCTGGGACAGG TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/3/5/7 CACCTGGGACA TATACCNCGTGT GTTGGTTCAAGATATTTGCTGAAG JP1 
 CACCTGGGACG CTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 9–11 CACCTGGGATAGG CCA TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 12 1/3/5/7 CACCTGGGACAGG CT TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
3+7+ 1/3/5/7 CACCTGGA  TGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/2/3/4/5/7 C?GCCTGGG CC GGTTGGTTCAAGATATTTGCTGAAG JP1 
 3, 4 CACCTGGGACGGG CCTTA TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 CACCTGGGATA  TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/2/3/4/5/7 CACCTGGGA  TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 CACCTGGGATAG TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/3/5/7 CACCTGGGACAG TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 1/3/5/7 CACCTGGGACA  TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 10  CACC  TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 11  CACC CCA TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 12  CACCTG GA TACCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
 13 CACCTGGGATAGG  CCACTGGTTGGTTCAAGATATTTGCTGAAG JP1 
a

DNA sequence analysis was performed on the CD37+ subset and CD3+7+ subset of two fetal gut samples a and b.

b

Of the 37+ subsets, sequence nos. 1–5 belong to fetal gut sample a, nos. 6–12 to fetal gut sample b. Of the 3+7+ subsets, sequence nos. 1–8 belong to fetal gut sample a, nos. 9–13 to fetal gut sample b.

To demonstrate the ability of CD37+ cells to differentiate into CD3+ T cells, we examined the behavior of CD37+ and CD3+7+ cells in a long-term culture system using the RTOC system.

Freshly isolated and sorted fetal CD37+ cells and CD3+7+ cells (2–4 × 104) from two fetal gut samples (11.3 and 11.5 wk gestation) were added into murine RTOCs. After 14 days in RTOCs, cells were harvested (1–7 × 104) and immediately stained for flow cytometric analysis. We found that ∼30% of the cells isolated from the RTOCs reconstituted with CD37+ cells expressed CD3 with moderate to high signal intensity after 2 wk in the culture system (Fig. 6). Sorted CD3+ cells in the RTOCs still continued to express CD3 (Fig. 6). The CD37+ cells added to the RTOC were at least 99.5% CD3, indicating that there were only 100 contaminating CD3+ cells in 20,000 CD37+ used to reconstitute the organ culture. Empty murine stromal cell cultures used as a negative control did not show any CD7 or CD3 signal in the subsequent flow cytometric analysis.

FIGURE 6.

CD37+ cells give rise to CD3+ T cells in RTOCs. Freshly isolated and sorted fetal CD37+ cells as well as CD3+7+ cells were added into murine RTOCs. After 14 days in culture, ∼30% of the CD37+ cells expressed CD3 and sorted CD3+ cells still continued to express CD3. The data were obtained in two separate experiments (Exp. 1, fetal gut 11.3 wk gestation; Exp. 2, 11.5 wk gestation). The harvested cell population was gated based on side scatter and their CD7 expression.

FIGURE 6.

CD37+ cells give rise to CD3+ T cells in RTOCs. Freshly isolated and sorted fetal CD37+ cells as well as CD3+7+ cells were added into murine RTOCs. After 14 days in culture, ∼30% of the CD37+ cells expressed CD3 and sorted CD3+ cells still continued to express CD3. The data were obtained in two separate experiments (Exp. 1, fetal gut 11.3 wk gestation; Exp. 2, 11.5 wk gestation). The harvested cell population was gated based on side scatter and their CD7 expression.

Close modal

In the present study, we show that the CD37+ population of cells abundant in human fetal gut appears to be developing into T cells in vivo and can give rise to T cells when placed in thymic organ cultures. CD37+ cells are abundant in the early fetal gut and also continue to persist in the adult intestine. Interestingly, CD37+ cells were already seen in fetal gut of 7 wk gestation, shortly after hemopoiesis begins in fetal liver (around 5–7.5 wk) and around the time the fetal thymus is colonized by hemopoietic progenitor cells. At this time, no CD3+ T cells were found in the fetal gut. CD3+ T cells are first seen within the fetal lamina propria around 8–9 wk in very low numbers, and their number increases dramatically reaching approximately one-third around 20 wk of that seen in adult intestine (13).

Further characterization of the fetal LPLs revealed a distinct profile of surface Ags for CD37+ cells and CD3+7+ cells. Many fetal LPLs expressed the stem cell marker CD34 suggesting recent arrival from fetal hemopoietic sites such as fetal liver and bone marrow. Among the CD37+ cells, the majority of cells were either DN or CD8 single-positive. Among the CD37+ CD8+ cells, 95% were CD8αα homodimeric, whereas in the mature CD3+ T cells, the proportion of CD8αα declined in favor of CD8αβ-positive cells with increasing gestational age. These findings agree with our previous immunohistochemical data that almost half the CD8+ cells in the lamina propria of fetal gut samples of 16–24 wk gestational age were CD8αα positive, which included CD37+ cells, but mainly mature CD3+ T cells (27).

A considerable time ago we described a population of CD37+ cells in the small bowel epithelium of healthy individuals, particularly prominent at the villus tip (18). More recent studies have characterized these cells (28, 29, 30), and some are remarkably similar to the cells we find in abundance in fetal gut, although we need to emphasize that the cells we investigated were from lamina propria and not epithelium. In adult small intestine, CD3 IELs displayed a distinct NK marker pattern (28, 29, 30). Approximately 10% of CD3 IELs expressed CD8 of which the majority were also CD8αα homodimeric (28). Likewise about the same percentage expressed intracellular CD3ε (28).

We show here that fetal lamina propria CD37+ cells around 13-wk gestational age also express a distinct NK marker profile with low levels of CD16, a substantial expression of CD56, CD94, and high CD161 expression. Functionally, it has been demonstrated that NK-like CD3-IELs in adult small intestine possess a cytotoxic lymphokine-activated killer activity superior to that of CD3+ IELs (28). Thus, NK receptor-bearing fetal CD37+ cells might be involved in the innate immune response to pathogens immediately after birth, when the gut is exposed to external Ags. Functional studies still need to be conducted to test this hypothesis.

We were also able to study two fetal gut samples aged 7 and 7.5 wk, a time when CD37+ cells were present in the tissues, but CD3+ cells were undetectable. In comparison with CD37+ cells from specimens older than 9 wk, these very young samples showed three striking differences: namely, very low numbers of CD8+ and CD94+ cells and reduced numbers of CD161+ cells. Although one should be cautious in interpreting data from only two specimens, the CD8 and CD94 results are particularly profound. This raises the possibility that in the gut, NK receptor expression might be part of a genetic program of T cell differentiation (31), as well as markers associated with activation of mature T cells (32, 33). In mice, CD94/NKG2 expression has been detected on very early fetal NK cells and fetal T cells. It has been attributed to inhibitory NK functions and seems to play an important role in maintaining self tolerance (34).

A novel part of this study is that we also demonstrated the ability of CD37+ cells to differentiate into CD3+ T cells. On a molecular level, TCR genes undergo sequentially ordered rearrangement during T cell differentiation. The rearrangement of the TCR γ-genes precedes that of β- and α-chain genes and is one of the earliest signs of T cell lineage commitment common to all T cells (35, 36). We were able to detect rearranged TCR γ-chains in the CD37+ cells in human fetal and adult gut indicating that a proportion of CD37+ cells already shows T cell lineage commitment on a molecular level. The γ-chain rearrangement was polyclonal. In the VγJP reaction different V segments were rearranged to the JP1 segment in 24 of 25 rearrangements (and not to JP or JP2 segments). In fetal T cells, sequential rearrangement seems to occur in the TCR γ locus (37, 38, 39) and rearrangements of the most 3′ Vγ genes to JP1 (the most 5′ Jγ segment) take place preferentially at an early stage of development. Additionally, a subset of fetal CD37+ expressed intracellular CD3ε, which is found in NK and T cell progenitors (25, 26).

To further confirm these findings, sorted fetal CD37+ cells were cultured in the RTOC system. After 2 wk in culture, ∼30% of CD37+ cells expressed the CD3 surface Ag. Thus, a proportion of fetal CD37+ cells seem to be precursors for CD3+ T cells. Similar events have been shown in fetal and embryonic liver of 6–8 wk gestational age (before thymic colonization) where CD3 cells gave rise to mature CD3+ T cells when cultured suggesting that these cells are rather not thymic emigrants and that T cell maturation can occur in an extrathymic environment (40).

Thus, fetal CD37+ cells might be a heterogeneous population with a subset being NK cells or NK cell precursors on the basis of their NK receptor expression. Other subsets seem to be T cell precursors that express intracellular CD3 (13), show TCR γ-chain rearrangement, and give rise to CD3+ T cells in the RTOC system. Although it would be tempting to suggest that the majority population of CD37+ cells that express CD8αα and CD161 are the precursors of the CD3+ cells with the same surface markers, we feel that such a conclusion would be premature. This could be tested by repopulating RTOCs with CD37+ cells, further subdivided according to NK marker status. However, these experiments are demanding and may not be technically possible given the relatively small numbers of cells isolated from individual fetal specimens and the rigorous nature of the sorting.

The fetal CD37+ subpopulation might also be seen as the physiological counterpart of the immunophenotypically aberrant IEL population seen in refractory sprue, which resembles a celiac disease-like enteropathy, but is resistant to strict gluten-free diet. In this disease, a massive expansion of IELs is observed with an unusual phenotype presenting with intracellular CD3, but lacking surface CD3-TCR complexes, often lacking CD8 and CD4, and a variable expression of CD94 (41, 42, 43, 44). Such IELs are considered neoplastic and have been suggested to represent a cryptic variant of enteropathy-associated T cell lymphoma (45).

Another interesting feature of the neoplastic cells in enteropathy-associated T cell lymphoma, as well as normal mucosal T cells is the expression of the gut-specific integrin CD103 (46). We were somewhat surprised to find that ∼40% of the CD37+ cells and 30% of the CD3+ cells expressed this integrin. It is generally thought that CD103 is induced in gut lymphocytes by TGFβ (47). On the other hand, there is now some evidence that CD103 is expressed on human thymocytes and that interaction with the CD103 ligand, E-cadherin, on thymic stromal cells enhances thymocyte proliferation (48). In the gut, E-cadherin is expressed on epithelial cells, so it is unlikely that CD103 lamina propria cells will have the opportunity to interact with the ligand, unless the cells traffic in and out of the epithelium.

Taken together, we could demonstrate that CD37+ cells are present in the fetal and adult gut. Fetal CD37+ cells display a distinct phenotype and a proportion is able to develop into CD3+ T cells in a possibly thymus-independent way. Our present data are consistent with the concept that fetal or embryonic intestine contains precursors that are either uncommitted pluripotent stem cells or committed to the T cell lineage and appropriate maturational signals would trigger their differentiation into CD3+ T cells.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

U.G., J.H., and A.K. were supported by the Wellcome Trust.

3

Abbreviations used in this paper: IEL, intraepithelial lymphocyte; pTα, pre-TCR-α; RTOC, reaggregate thymic organ culture; LPL, lamina propria lymphocyte; DN, double negative; DP, double positive.

1
Lefrancois, L., L. Puddington.
1995
. Extrathymic intestinal T-cell development: virtual reality?.
Immunol. Today
16
:
16
.
2
Bandeira, A., S. Itohara, M. Bonneville, O. Burlen-De Franoux, T. MotoSantos, A. Coutinko, S. Tonegawa.
1991
. Extrathymic origin of intestinal intraepithelial lymphocytes bearing T cell antigen receptor γδ.
Proc. Natl. Acad. Sci. USA
88
:
43
.
3
Lefrancois, L., S. Olson.
1994
. A novel pathway of thymus directed T lymphocyte maturation.
J. Immunol.
153
:
987
.
4
Rocha, B., P. Vassalli, D. Guy-Grand.
1994
. Thymic and extrathymic origin of gut intraepithelial lymphocyte populations in mice.
J. Exp. Med.
180
:
681
.
5
Saito, H., Y. Kanamori, T. Takemori, H. Nariuchi, E. Kubota, H. Takahashi-Iwanaga, T. Iwanaga, H. Ishikawa.
1998
. Generation of intestinal T cells from progenitors residing in gut cryptopatches.
Science
280
:
275
.
6
Suzuki, K., T. Oida, H. Hamada, O. Hitotsumatsu, M. Watanabe, T. Hibi, H. Yamamoto, E. Kubota, S. Kaminogawa, H. Ishikawa.
2000
. Gut cryptopatches: direct evidence of extrathymic anatomical sites for intestinal lymphopoiesis.
Immunity
13
:
691
.
7
Guy-Grand, D., O. Azogui, S. Celli, S. Darche, M. C. Nussenzweig, P. Kourilsky, P. Vassalli.
2003
. Extrathymic T cell lymphopoiesis: ontogeny and contribution to gut intraepithelial lymphocytes in athymic and euthymic mice.
J. Exp. Med.
3
:
333
.
8
Sanchez, M. J., J. C. Gutierrez-Ramos, E. Fernandez, E. Leonardo, J. Lozano, C. Martinez, M. L. Toribio.
1993
. Putative prethymic T cell precursors within the early human embryonic liver: a molecular and functional analysis.
J. Exp. Med.
177
:
19
.
9
Haynes, B. F., M. E. Martin, H. H. Kay, J. Kurtzberg.
1988
. Early events in human T cell ontogeny: phenotypic characterization and immunohistologic localization of T cell precursors in early human fetal tissues.
J. Exp. Med.
168
:
1061
.
10
Haynes, B. F., S. M. Denning, K. H. Singer, J. Kurtzberg.
1989
. Ontogeny of T-cell precursors: a model for the initial stages of human T-cell development.
Immunol. Today
10
:
87
.
11
Spencer, J., T. T. MacDonald, T. Finn, P. G. Isaacson.
1986
. The development of gut associated lymphoid tissue in the terminal ileum of fetal human intestine.
Clin. Exp. Immunol.
64
:
536
.
12
Spencer, J., P. G. Isaacson, J. A. Walker-Smith, T. T. MacDonald.
1989
. Heterogeneity in intraepithelial lymphocyte subpopulations in fetal and postnatal human small intestine.
J. Pediatr. Gastroenterol. Nutr.
9
:
173
.
13
Howie, D., J. Spencer, D. DeLord, C. Pitzalis, N. C. Wathen, A. Dogan, A. Akbar, T. T. MacDonald.
1998
. Extrathymic T cell differentiation in the human intestine early in life.
J. Immunol.
161
:
5862
.
14
Berlin, C., R. F. Bargatze, J. J. Campbell, U. H. von Andrian, M. C. Szabo, S. R. Hasslen, R. D. Nelson, E. L. Berg, S. L. Erlandsen, E. C. Butcher.
1995
. α4 integrins mediate lymphocyte attachment and rolling under physiologic flow.
Cell
80
:
413
.
15
Koningsberger, J. C., A. Chott, T. Logtenberg, L. J. J. M. Wiegman, R. S. Blumberg, G. P. van Berge Henegouwen, S. P. Balk.
1997
. TCR expression in human fetal intestine and identification of an early T cell receptor β-chain transcript.
J. Immunol.
159
:
1775
.
16
Saint-ruf, C., K. Ungewiss, M. Groettrup, L. Bruno, H. J. Fehling, H. von Boehmer.
1994
. Analysis and expression of a cloned pre-T cell receptor gene.
Science
266
:
1208
.
17
Penit, C., B. Lucas, F. Vasseur.
1995
. Cell expansion and growth arrest phases during the transition from precursor (CD48) to immature (CD4+8+) thymocytes in normal and genetically modified mice.
J. Immunol.
154
:
5103
.
18
Spencer, J., T. T. MacDonald, T. C. Diss, J. A. Walker-Smith, P. J. Ciclitira, P. G. Isaacson.
1989
. Changes in intraepithelial lymphocyte subpopulations in coeliac disease and enteropathy associated T cell lymphoma (malignant histiocytosis of the intestine).
Gut
30
:
339
.
19
Marker-Hermann, E., R. Duchmann.
1998
. Isolation of T cells and establishment of T-cell lines and clones.
Methods Microbiol.
25
:
539
.
20
Anderson, G., J. J. T. Owen, N. C. Moore, E. J. Jenkinson.
1994
. Thymic epithelial cells provide unique signals for positive selection of CD4+CD8+ thymocytes in vitro.
J. Exp. Med.
179
:
2027
.
21
Anderson, G., E. J. Jenkinson, N. C. Moore, J. J. T. Owen.
1993
. Both MHC class II-positive thymic epithelial and mesenchyme cells are required for T-cell development in the thymus.
Nature
362
:
70
.
22
McCarthy, K. P., J. P. Sloane, J. H. S. Kabarowski, E. Matutes, L. M. Wiedemann.
1992
. A simplified method of detection of clonal rearrangements of the T-cell receptor γ chain gene.
Diagn. Mol. Pathol.
1
:
173
.
23
Lundquist, C., V. Baranov, S. Hammarström, L. Athlin, M. L. Hammarström.
1995
. Intraepithelial lymphocytes: evidence for regional specialization and extrathymic T cell maturation in the human gut epithelium.
Int. Immunol.
24
:
1703
.
24
Jarry, A., N. Cerf-Bensussan, N. Brousse, F. Selz, G. Guy-Grand.
1990
. Subsets of CD3+ (T cell receptor αβ and γδ) and CD3 lymphocytes isolated from normal human gut epithelium display phenotypic features different from their counterparts in peripheral blood.
Eur. J. Immunol.
20
:
1097
.
25
Phillips, J. H., T. Hori, A. Nagler, N. Bhat, H. Spits, L. L. Lanier.
1992
. Ontogeny of human natural killer (NK) cells: fetal NK cells mediate cytolytic function and express cytoplasmatic CD3 ε, δ proteins.
J. Exp. Med.
175
:
1055
.
26
Wang, B., N. Wang, C. E. Whitehurst, J. She, J. Chen, C. Terhost.
1999
. T lymphocyte development in the absence of CD3 ε or CD3 γδεζ.
J. Immunol.
162
:
88
.
27
Latthe, M., L. Terry, T. T. MacDonald.
1994
. High frequency of CD8αα homodimer-bearing T cells in human fetal intestine.
Eur. J. Immunol.
24
:
1703
.
28
Leon, F., E. Roldan, L. Sanchez, C. Camarero, A. Bootello, G. Roy.
2003
. Human small-intestinal epithelium contains functional natural killer lymphocytes.
Gastroenterology
125
:
345
.
29
Eiras, P., F. Leon, C. Camarero, M. Lombardia, E. Roldan, A. Bootello, G. Roy.
2000
. Intestinal intraepithelial lymphocytes contain a CD3CD7+ subset expressing natural killer markers and a singular pattern of adhesion molecules.
Scand. J. Immunol.
52
:
1
.
30
Jabri, B., N. P. de Serre, C. Cellier, K. Evans, C. Gache, C. Carvalho, J. F. Mougenot, M. Allez, R. Jian, P. Desreumaux, et al
2000
. Selective expansion of intraepithelial lymphocytes expressing the HLA-E-specific natural killer receptor CD94 in celiac disease.
Gastroenterology
118
:
867
.
31
Bendelac, A., M. N. Rivera, S. H. Park, J. H. Roark.
1997
. Mouse CD1-specific NK1 T cells: development, specificity, and function.
Annu. Rev. Immunol.
15
:
535
.
32
Mingari, M. C., M. Ponte, S. Bertone, F. Schiavetti, C. Vitale, R. Bellomo, A. Moretta, L. Moretta.
1998
. HLA class I-specific inhibitory receptors in human T lymphocytes: interleukin 15-induced expression of CD94/NKG2A in superantigen- or alloantigen-activated CD8+ T cells.
Proc. Natl. Acad. Sci. USA
95
:
1172
.
33
Lanier, L. L., J. H. Phillips.
1996
. Inhibitory MHC class I receptors on NK cells and T cells.
Immunol. Today
17
:
86
.
34
Van Beneden, K., F. Stevenaert, A. De Creus, V. Debacker, J. De Boever, J. Plum, G. Leclercq.
2001
. Expression of Ly49E and CD94/NKG2 on fetal and adult NK cells.
J. Immunol.
166
:
4302
.
35
Haars, R., M. Kronenberg, W. M. Gallatin, I. L. Weissmann, I. L. Owen, L. Hood.
1986
. Rearrangement and expression of T cell antigen receptor and γ genes during thymic development.
J. Exp. Med.
164
:
1
.
36
Blom, B., M. C. Verschuren, M. H. Heemskerk, A. Q. Bakker, E. J. van Gastel-Mol, I. L. Wolvers-Tettero, J. J. van Dongen, H. Spits.
1999
. TCR gene rearrangement and expression of the pre-T cell receptor complex during human T-cell differentiation.
Blood
93
:
3033
.
37
Krangel, M. S., H. Yssel, C. Brocklehurst, H. Spits.
1990
. A distinct wave of human T cell receptor γ/δ lymphocytes in the early fetal thymus: evidence for controlled gene rearrangement and cytokine production.
J. Exp. Med.
172
:
847
.
38
Alexandre, D., P. Chuchana, M. G. Roncarolo, H. Yssel, H. Spits, G. Lefranc, M. P. Lefranc.
1991
. Reciprocal hybrid joints demonstrate successive V-J rearrangements on the same chromosome in the human TCR gamma locus.
Int. Immunol.
3
:
973
.
39
Zhang, X. M., C. Tonnelle, M. P. Lefranc, S. Huck.
1994
. T cell receptor γ cDNA in human fetal liver and thymus: variable regions of γ chains are restricted to Vγ1 or V9, due to the absence of splicing of the V10 and V11 leader intron.
Eur. J. Immunol.
24
:
571
.
40
Poggi, A., M. Sargiacomo, R. Biassoni, N. Pella, S. Sivori, V. Revello, P. Costa, M. Valtieri, G. Russo, M. C. Mingari.
1993
. Extrathymic differentiation of T lymphocytes and natural killer cells from human embryonic liver precursors.
Proc. Natl. Acad. Sci. USA
90
:
4465
.
41
Cellier, C., N. Patey, L. Mauvieux, B. Jabri, E. Delabesse, J. P. Cervoni, M. L. Burtin, D. Guy-Grand, Y. Bouhnik, R. Modigliani, et al
1998
. Abnormal intestinal intraepithelial lymphocytes in refractory sprue.
Gastroenterology
114
:
471
.
42
Daum, S., D. Weiss, M. Hummel, R. Ullrich, W. Heise, H. Stein, E. O. Riecken, H. D. Foss.
2001
. Intestinal Lymphoma Study Group: frequency of clonal intraepithelial T lymphocyte proliferations in enteropathy-type intestinal T cell lymphoma, coeliac disease, and refractory sprue.
Gut
49
:
804
.
43
Farstad, I. N., F. E. Johansen, L. Vlatkovic, J. Jahnsen, H. Scott, O. Fausa, A. Bjorneklett, P. Brandtzaeg, T. S. Halstensen.
2002
. Heterogeneity of intraepithelial lymphocytes in refractory sprue: potential implications of CD30 expression.
Gut
51
:
372
.
44
Mention, J. J., M. Ben Ahmed, B. Begue, U. Barbe, V. Verkarre, V. Asnafi, J. F. Colombel, P. H. Cugnenc, F. M. Ruemmele, E. Macintyre, et al
2003
. Interleukin 15: a key to disrupted intraepithelial lymphocyte homeostasis and lymphomagenesis in celiac disease.
Gastroenterology
125
:
730
.
45
Cellier, C., E. Delabesse, C. Helmer, N. Patey, C. Matuchansky, B. Jabri, E. Macintyre, N. Cerf-Bensussan, N. Brousse.
2000
. Refractory sprue, coeliac disease, and enteropathy-associated T-cell lymphoma. French Coeliac Disease Study Group.
Lancet
356
:
203
.
46
Spencer, J., N. Cerf-Bensussan, A. Jarry, N. Brousse, D. Guy-Grand, A. S. Krajewski, P. G Isaacson.
1988
. Enteropathy-associated T cell lymphoma (malignant histiocytosis of the intestine) is recognized by a monoclonal antibody (HML-1) that defines a membrane molecule on human mucosal lymphocytes.
Am. J. Pathol.
132
:
1
.
47
Robinson, P. W., S. J. Green, C. Carter, J. Coadwell, P. J. Kilshaw.
2001
. Studies on the transcriptional regulation of the mucosal T-cell integrin αEβ7 (CD103).
Immunology
103
:
146
.
48
Kutlesa, S., J. T. Wessels, A. Speiser, I. Steiert, C. A. Muller, G. Klein.
2002
. E-cadherin-mediated interactions of thymic epithelial cells with CD103+ thymocytes lead to enhanced thymocyte proliferation.
J. Cell Sci.
115
:
4505
.