| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Divisions of
*
Infection, Inflammation and Repair,
Human Genetics, and
Cancer Sciences,
Department of Surgery, University of Southampton, Southampton;
¶
Department of Immunobiology, GKT School of Medicine, London, United Kingdom; and
||
Department of Anatomy, University of Birmingham, Birmingham, United Kingdom
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
homodimer. In contrast about half of CD3+ cells expressed CD4 and half expressed CD8. A large proportion of CD3–7+ 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 CD3–7+ cells have the potential to differentiate into CD3+ cells. About half of CD3–7+ cells contain intracellular CD3
. Rearranged TCR 
-chains were detected in highly purified CD3–7+ cells as an early molecular sign of T cell commitment, and the pattern of rearrangement with V regions spliced to the most 5' J segment is reminiscent of early thymocyte differentiation. In reaggregate thymic organ cultures, CD3–7+ cells also gave rise to CD3+ T cells. Thus, we demonstrate that the CD3–7+ cells present in the human fetal gut display a distinct phenotype and are able to develop into CD3+ T cells. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
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, E
7, and CD45RO, commonly associated with memory T cells that have previously undergone antigenic stimulation (13). The cells also lack the 47 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- mRNA transcripts in fetal intestine at 14 wk of gestation (15), concurrently with pre-TCR- (pT) (13), which associates with the TCR -chain at the cell surface before recombination of the TCR -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, pT mRNA transcripts were still present in the fetal gut specimens after an engraftment period of 30 days (13). In thymus, pT-expressing pro-T cells have a lifespan of 3 days (17), making it unlikely that the pT 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 CD3–7+ 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 -chain rearrangement, and transfer of sorted CD3–7+ cells into murine reaggregate thymic organ cultures (RTOCs) to characterize the population of CD3–7+ cells in the human fetal gut, and provide strong evidence that these CD3–7+ cells are able to develop into mature CD3+ T cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
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.
Immunohistochemistry
Cryostat sections were cut at 6 µm and fixed in 4% paraformaldehyde for 20 min. CD3–7+ 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.
Image analysis
For quantification of CD3–7+ 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.
Lamina propria lymphocyte (LPL) isolation
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 x 106 to 5 x 106 cells, depending on the age. From adult gut specimens 5 to 10 x 106 cells were recovered. More than 95% of the cells were viable as judged by trypan blue exclusion test.
Flow cytometric analysis and cell sorting
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 x 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 CD3 (UCHT1; Dako), CD4 (B-F5; Diaclone), CD7 (DK24l Dako), CD8 (DK25l Dako), CD8 (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), TCR (T10B9.1A-31; BD Biosciences), TCR (B1.1; BD Biosciences), and V24 (C15, Coulter-Immunotech). Fluorochrome-conjugated isotype control Igs were used for all analyses.
The intracellular staining for CD3 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-CD3 Ab conjugated to allophycocyanin or isotype-matched Ig control.
For sorting, up to 5 x 106 fetal LPLs as well as up to 20 x 106 adult LPLs were labeled in two colors using CD7-FITC and CD3-PE. Subsets of CD3–7+cells and CD3+7+cells (1–8 x 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.
RTOCs
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 CD3–7+ cells or CD3+7+ cells were mixed together with the murine stromal cells (1 x 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.
PCR for TCR -chain rearrangement
Subsets of CD3–7+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 V1 to V8 gene segments, the V9 as well as V10 and V11 gene segments were used in conjunction with either J or JP primers as described by McCarthy et al. (22).
The TCR -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, 1x 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 J 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 CD3–7+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.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
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 CD3–7+ 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. 1A). CD3–7+ cells were also occasionally seen within the epithelium.
|
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. 1B). 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.
Surface phenotype of fetal gut CD3–7+ and CD3+ cells
LPLs isolated from six fetal gut samples (10–14 wk gestation) were characterized by three- and four-color flow cytometry. CD3–7+ and CD3+7+ cells were analyzed for CD34, CD103, CD4, CD8, and CD8 expression, their CD4/8 double-negative (DN) and CD4/8 double-positive (DP) status, as well as the TCR and TCR expression (Table I).
|
Among the fetal CD3–7+ 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. 2A). CD8 expression among the CD3–7+ cells was high (39 ± 4.1; Table I, Fig. 2A). The vast majority of CD3–7+ CD8+ cells in the fetal lamina propria of all samples expressed the CD8 homodimer (95 ± 2.3; Fig. 2B).
|
-expressing cells in this population declined from 52 to 10%, associated with an increment in CD8-expressing cells (Fig. 2B). Within the CD3–7+ cell subset, a large percentage were DN (49 ± 3.8) and only a minor fraction was DP (3 ± 1; Table I, Fig. 2A). 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. 2A).
As expected, CD3–7+ cells of all fetal gut samples were TCR and TCR negative (Table I, Fig. 2A). In the CD3+ cells, the majority showed TCR expression (77 ± 14.4), and the remaining fraction TCR expression (25 ± 0; Table I, Fig. 2A). In three additional fetal gut samples (10.6–14 wk gestation) only a small percentage of the CD3+ T cells expressed the invariant V24-chain (1.8 ± 1.4).
Human fetal CD3–7+LPLs express a distinct pattern of NK receptors
In human adult small intestine, a CD3–7+ population is present in the gut epithelium that expresses NK markers (18, 23, 24), and a subset displays intracellular CD3 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 CD3–7+ 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. 3A). Two NK receptors of the C-lectin-like family were expressed by a large fraction of CD3–7+ 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 CD3–7+ cells. The NKR-P1A receptor (CD161) was expressed on 70% ± 8.8 of CD3–, 7+ cells (Fig. 3A). 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. 3A).
|
was expressed by the majority of CD3–7+ CD94+ cells (73 ± 1.5, Fig. 3B) and a large proportion of the CD3–7+ CD161+ cells (40 ± 11.6, Fig. 3B), whereas CD4 expression was low (6 ± 3.4 and 3 ± 2.4, Fig. 3B). However, in CD3+161+ cells, the pattern was different. A large proportion expressed CD4 (56 ± 17) and CD8 (32 ± 2, Fig. 3C). Early fetal CD3–7+ cells
Having demonstrated that CD3–7+ 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 CD3–7+ cells were analyzed, and results are shown in Table II. A small fraction of early fetal CD3–7+ 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. 3A). 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. 3A). 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).
|
and TCR -chain rearrangement
CD3–7+ LPLs isolated from four human fetal gut samples of 14–16 wk gestation were investigated for their expression of intracellular CD3 by flow cytometric analysis. Intracellular CD3 expression was seen in a substantial fraction of the CD3–7+ cells (42 ± 17, Fig. 4).
|
-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 CD3–7+ 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 -chain rearrangement consists of two different reactions. In the VJP reaction, rearrangements using the J segments JP, JP1, or JP2 are detected, whereas rearrangements using the J1 or J2 segment are seen in the VJ reaction. Rearranged TCR -chains were detected within the CD3–7+ cell subset of all fetal gut samples and 3/5 adult colon samples in the VJP reaction. The rearrangement was polyclonal (Fig. 5C). In the VJ reaction, two of six fetal CD3–7+ samples and one of six adult CD3–7+ samples displayed a weak signal.
|
-chain rearrangement generating a polyclonal smear in both the VJ and VJP reaction (Fig. 5C).
PCR products of the VJP reaction from two fetal gut samples were cloned and sequenced (Table III). Twelve sequences were obtained from the 3–7+ 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 -chain rearrangement using the JP2 segment, all other rearrangements involved the JP1 segment. Sequences of rearranged TCR -chains were diverse. Within CD3–7+ and CD3+ cells, respectively, 9 of 12 and 11 of 13 sequences were represented only once.
|
To demonstrate the ability of CD3–7+ cells to differentiate into CD3+ T cells, we examined the behavior of CD3–7+ and CD3+7+ cells in a long-term culture system using the RTOC system.
Freshly isolated and sorted fetal CD3–7+ cells and CD3+7+ cells (2–4 x 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 x 104) and immediately stained for flow cytometric analysis. We found that 
30% of the cells isolated from the RTOCs reconstituted with CD3–7+ 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 CD3–7+ cells added to the RTOC were at least 99.5% CD3–, indicating that there were only 100 contaminating CD3+ cells in 20,000 CD3–7+ 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.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Further characterization of the fetal LPLs revealed a distinct profile of surface Ags for CD3–7+ 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 CD3–7+ cells, the majority of cells were either DN or CD8 single-positive. Among the CD3–7+ CD8+ cells, 95% were CD8 homodimeric, whereas in the mature CD3+ T cells, the proportion of CD8 declined in favor of CD8-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 CD8 positive, which included CD3–7+ cells, but mainly mature CD3+ T cells (27).
A considerable time ago we described a population of CD3–7+ 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 CD8 homodimeric (28). Likewise about the same percentage expressed intracellular CD3 (28).
We show here that fetal lamina propria CD3–7+ 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 CD3–7+ 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 CD3–7+ cells were present in the tissues, but CD3+ cells were undetectable. In comparison with CD3–7+ 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 CD3–7+ 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 -genes precedes that of - and -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 -chains in the CD3–7+ cells in human fetal and adult gut indicating that a proportion of CD3–7+ cells already shows T cell lineage commitment on a molecular level. The -chain rearrangement was polyclonal. In the VJP 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 locus (37, 38, 39) and rearrangements of the most 3' V genes to JP1 (the most 5' J segment) take place preferentially at an early stage of development. Additionally, a subset of fetal CD3–7+ expressed intracellular CD3, which is found in NK and T cell progenitors (25, 26).
To further confirm these findings, sorted fetal CD3–7+ cells were cultured in the RTOC system. After 2 wk in culture, 30% of CD3–7+ cells expressed the CD3 surface Ag. Thus, a proportion of fetal CD3–7+ 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 CD3–7+ 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 -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 CD3–7+ cells that express CD8 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 CD3–7+ 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 CD3–7+ 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 CD3–7+ cells and 30% of the CD3+ cells expressed this integrin. It is generally thought that CD103 is induced in gut lymphocytes by TGF (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 CD3–7+ cells are present in the fetal and adult gut. Fetal CD3–7+ 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.
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 U.G., J.H., and A.K. were supported by the Wellcome Trust. 
2 Address correspondence and reprint requests to Dr. Thomas T. MacDonald, Division of IIR, Mailpoint 813, School of Medicine, University of Southampton, Tremona Road, Southampton, SO16 6YD, U.K. E-mail address: T.T.MacDonald{at}soton.ac.uk
3 Abbreviations used in this paper: IEL, intraepithelial lymphocyte; pT, pre-TCR-; RTOC, reaggregate thymic organ culture; LPL, lamina propria lymphocyte; DN, double negative; DP, double positive.
Received for publication March 18, 2004. Accepted for publication February 28, 2005.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
. Proc. Natl. Acad. Sci. USA 88:43.4 integrins mediate lymphocyte attachment and rolling under physiologic flow. Cell 80:413.[Medline]
-chain transcript. J. Immunol. 159:1775.[Abstract]
chain gene. Diagn. Mol. Pathol. 1:173.[Medline]
and ) and CD3– lymphocytes isolated from normal human gut epithelium display phenotypic features different from their counterparts in peripheral blood. Eur. J. Immunol. 20:1097.[Medline]
, proteins. J. Exp. Med. 175:1055. or CD3 
. J. Immunol. 162:88. homodimer-bearing T cells in human fetal intestine. Eur. J. Immunol. 24:1703.[Medline]
genes during thymic development. J. Exp. Med. 164:1./ lymphocytes in the early fetal thymus: evidence for controlled gene rearrangement and cytokine production. J. Exp. Med. 172:847. cDNA in human fetal liver and thymus: variable regions of chains are restricted to V1 or V9, due to the absence of splicing of the V10 and V11 leader intron. Eur. J. Immunol. 24:571.[Medline]
E7 (CD103). Immunology 103:146.[Medline]
This article has been cited by other articles:
|
|
 |
|
  O. Cohavy and S. R. Targan CD56 Marks an Effector T Cell Subset in the Human Intestine J. Immunol., May 1, 2007; 178(9): 5524 - 5532. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Vijayanand, G. Seumois, C. Pickard, R. M. Powell, G. Angco, D. Sammut, S. D. Gadola, P. S. Friedmann, and R. Djukanovic Invariant Natural Killer T Cells in Asthma and Chronic Obstructive Pulmonary Disease N. Engl. J. Med., April 5, 2007; 356(14): 1410 - 1422. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
