Intestinal Ag exposure during neonatal life influences appropriate adult immune responses. To define the mechanisms shaping the T cell repertoire during this period, we examined T cell differentiation and receptor diversity in the intestine of human infants. Developmental phenotypes of intraepithelial and lamina propria intestinal T cells from infants aged 1 day to 2 years were assessed ex vivo by flow cytometry and in situ by triple-fluorescent immunohistochemistry. Gene recombination-specific enzymes were assessed by PCR. TCR β-chain V region gene diversity was determined by sequencing. Several different early lineage T cell populations were present neonatally: CD3+48 T cells were present at birth and numbers decreased during the neonatal period; CD3+4+8+ T cells were present in low numbers throughout infancy; and CD3+4+8 or CD3+48+ T cells increased with age. Very early lineage T cells, CD327+ and CD32+7+, were present neonatally, but were essentially absent at 1 year. Most lamina propria T cells differentiated rapidly after birth, but maturation of intraepithelial T cells took place over 1 year. Intestinal samples from infants less than 6 mo old contained transcripts of T early α and TdT, and 15 of 19 infant samples contained mRNA for RAG-1, some coexpressing RAG-2. TCR β-chain repertoires were polyclonal in infants. Immature T cells, pre-T cells, and genes involved in T cell recombination were found in the intestine during infancy. T cell differentiation occurs within the neonatal human intestine, and the TCR repertoire of these developing immature T cells is likely to be influenced by luminal Ags. Thus, mucosal T cell responsiveness to environmental Ag is shaped in situ during early life.

The ontogeny of human intestinal T cells is poorly understood. It has been suggested that some undergo extrathymic development, and there is some evidence that the repertoire, particularly for γδ T cells, is shaped in infancy. However, the complexity of the αβ TCR repertoire in early life has received little attention.

In mice, extrathymic T cell maturation of a subpopulation of intestinal intraepithelial lymphocytes (IEL)3 has been proposed and much evidence to substantiate this theory has accumulated in recent years. The majority of intestinal intraepithelial T lymphocytes are CD8+ and express the αβ TCR. In mice, but not in human adults, many of these CD8+ IEL express CD8 as an αα homodimer (1, 2), rather than as an αβ heterodimer. Intestinal T cells of the TCRαβ CD8αα phenotype can express TCR β-chains capable of reacting with self-superantigens (1, 2, 3), but any T cells that are likely to respond in a self-reactive manner would normally undergo clonal deletion in the thymus. As TCRαβ CD8αα cells are found exclusively in the intestinal epithelial compartment, it has been suggested, therefore, that they have developed in the intestine (3), thus evading negative selection and clonal deletion in the thymus. Further evidence for intestinal T cell development in mice is the presence of mRNA for recombination-activating proteins RAG-1 and RAG-2 (1, 4, 5, 6, 7). In contrast, a recent study indicates that recombination events in mouse intestinal mucosal T cells are confined to athymic individuals and occur in the draining mesenteric lymph nodes (8). However, even in mice, the extent to which mucosal T cells differentiate in situ during the neonatal stage is still unclear.

In humans, evidence for extrathymic maturation in the intestine is limited. Immature T cells (9, 10) and TCRαβ CD8αα cells (11) have been found in human fetal intestine, but their presence in the intestine after birth has not been reported. Expression of RAG mRNA has been shown in human adult IEL (12), but recombination-activating proteins are not exclusive to T cells (7, 13, 14). Conversely, another gene associated with TCR rearrangement, TdT, was absent from the intestine of human adults (15). T early α (TEA) is expressed in immature T cells at the point when double-negative (DN) thymocytes become immature single-positive (SP) T cells (16), and is required for TCR recombination (17). TEA protein and concordant expression of TEA mRNA was reported in the intestine of human fetuses (9), but has not been investigated postnatally.

The TCRαβ and TCRγδ repertoire of IEL in human adults is oligoclonal (18, 19, 20) and it is assumed that this reflects the need to tightly regulate the immune response in the intestine. However, Holtmeier et al. (20) have reported a polyclonal repertoire in midgestation, retained after delivery, but becoming oligoclonal during childhood. Regulation of this restricted repertoire is not understood, but if IEL undergo TCR rearrangement in the intestine, then it is likely that intestinal Ags present during rearrangement are important in shaping the repertoire. Therefore, an understanding of where and when T cells mature will help us to understand the factors driving the development of the oligoclonal repertoire evident in adulthood.

To determine whether T cell maturation occurs in the intestine of human infants, we have investigated the phenotype of T cells and their precursors, the expression of genes and proteins associated with TCR rearrangement, and the repertoire of TCRαβ T cells.

Intestinal tissue was obtained from surgical resection specimens. Only normal tissue close to the resection margins was studied, so that any disease process leading to the surgery did not influence the results (Table I). All experiments were conducted blind and patient age and sample information was retrieved from the study database at the end of the study. Fresh tissue was obtained for the work with informed parental consent. All aspects of this research were approved by the Research Ethics Committee for United Bristol Healthcare Trust.

Table I.

Patient sample summary informationa

IdSampleSample DescriptionAgeWeaned?Procedure
PT01 Duplication cyst 6 day Removal of ileal duplication cyst and associated normal ileum 
PT02 Normal ileum 6 day   
PT27 Normal ileum 2 mo Ileostomy closure 
PT09 Normal ileum 2 mo 1 wk Reversal of ileostomy (closure). Stoma was created at birth, so colon was isolated from gut contents at birth. 
PT23 Normal ileum 2 mo 3 wk Ileostomy reversal (NEC pre-term, birth at 26 wk) 
PT03 Normal ileum 3 mo 3 wk Ileostomy reversal 
PT19 Normal ileum 4 mo Staging and resection of Hirschsprung’s disease creation of ileostomy 
PT18 Normal ileum 12 mo Ileostomy reversal 
PT25 Normal colon 1 day Colostomy creation for ano-rectal malformation 
PT26 Normal colon 5 day Colostomy formation suspected Hirschsprung’s disease 
PT10 Normal colon (never seen food Ags) 2 mo 1 wk See patient E, PT09, colon was isolated from gut contents at birth 
PT20 Normal colon 4 mo See patient L, PT19 
PT07 Normal sigmoid colon 6 mo Staging and resection for Hirschsprung’s disease, creation of colostomy 
PT11 Normal colon 9 mo Colostomy reversal, closure of loop stoma 
PT13 Normal colon 9 mo 1 wk Colostomy reversal 
PT17 Normal rectum 12 mo See patient K, PT18 
PT24 Normal colon 17 mo 2 wk Colostomy reversal (anorectal malformation) 
PT14 Normal colon 18 mo Colostomy reversal 
PT22 Normal colon 2 years 2 mo Colostomy reversal 
PT06 Normal sigmoid colon 4 years 9 mo Colostomy creation 
IdSampleSample DescriptionAgeWeaned?Procedure
PT01 Duplication cyst 6 day Removal of ileal duplication cyst and associated normal ileum 
PT02 Normal ileum 6 day   
PT27 Normal ileum 2 mo Ileostomy closure 
PT09 Normal ileum 2 mo 1 wk Reversal of ileostomy (closure). Stoma was created at birth, so colon was isolated from gut contents at birth. 
PT23 Normal ileum 2 mo 3 wk Ileostomy reversal (NEC pre-term, birth at 26 wk) 
PT03 Normal ileum 3 mo 3 wk Ileostomy reversal 
PT19 Normal ileum 4 mo Staging and resection of Hirschsprung’s disease creation of ileostomy 
PT18 Normal ileum 12 mo Ileostomy reversal 
PT25 Normal colon 1 day Colostomy creation for ano-rectal malformation 
PT26 Normal colon 5 day Colostomy formation suspected Hirschsprung’s disease 
PT10 Normal colon (never seen food Ags) 2 mo 1 wk See patient E, PT09, colon was isolated from gut contents at birth 
PT20 Normal colon 4 mo See patient L, PT19 
PT07 Normal sigmoid colon 6 mo Staging and resection for Hirschsprung’s disease, creation of colostomy 
PT11 Normal colon 9 mo Colostomy reversal, closure of loop stoma 
PT13 Normal colon 9 mo 1 wk Colostomy reversal 
PT17 Normal rectum 12 mo See patient K, PT18 
PT24 Normal colon 17 mo 2 wk Colostomy reversal (anorectal malformation) 
PT14 Normal colon 18 mo Colostomy reversal 
PT22 Normal colon 2 years 2 mo Colostomy reversal 
PT06 Normal sigmoid colon 4 years 9 mo Colostomy creation 
a

Summary data for all small and large intestinal samples investigated during the course of the study, ordered by intestinal region and chronological age. In some cases both small and large intestinal samples were taken from one patient and are indicated by the same patient Id.

Fresh tissue was collected on ice in RPMI 1640 (Sigma-Aldrich, Poole, Dorset, U.K.). The mucosa was dissected free of underlying tissue and two 5 mm × 10 mm pieces were mounted in Tissue-Tek OCT compound (RA Lamb, Sussex, U.K.) and frozen in liquid nitrogen-cooled isopentane for immunohistochemistry. Three match-head-size pieces of mucosa were stored in 1 ml of TRIzol (Invitrogen Life Technologies, Paisley, U.K.), for mRNA extraction within 1 mo. The remaining mucosa (1–2 cm2) was used for the isolation of lymphocytes, as previously described (21).

Isolated IEL and lamina propria lymphocytes (LPL) were passed through 70-μm sieves and purified over discontinuous Percoll gradients (Amersham, Bucks, U.K.) (40:60%). The cell count and viability was determined by trypan blue exclusion. Aliquots of each fraction were analyzed by flow cytometry and 106 cells were resuspended in 1 ml of TRIzol for mRNA isolation.

All Abs were titrated using PBLs for use with a maximum of 107 cells in a staining volume of 50 μl. Abs were supplied by Caltag-Medsystems (Towcester, U.K.), unless stated otherwise. Cell aliquots (50–100,000) were triple-stained with combinations of Abs against CD4 (FITC, IgG1; BD Biosciences, Oxford, U.K.), CD8 (R-PE, IgG2a), CD3 (tricolor, IgG2a), CD2 (FITC, IgG1), CD5 (R-PE, IgG1), and CD7 (R-PE, IgG2a). Isotype controls were FITC-mouse IgG1 and R-PE mouse IgG2a (DAKO, Cambridgeshire, U.K.) and analyzed using a FACSCalibur Flow Cytometer (BD Biosciences). Twenty thousand CD3+ events/sample were analyzed using CellQuest Pro Software (BD Biosciences).

RNA was extracted from whole fresh mucosa and from fractionated IEL and LPL using TRIzol according to the manufacturer’s protocol. Reverse transcription was performed using Expand Reverse Transcriptase (Roche, Sussex, U.K.) according to the manufacturer’s protocol. Integrity of the cDNA was confirmed by PCR amplification of a housekeeper gene, GAPDH, and visualized on a 2% agarose gel stained with ethidium bromide. The cDNA was kept at −20°C for later use.

Stored cDNA was used as a template for four sets of PCR amplifications designed to identify transcripts of genes that are associated with germline recombination and rearrangement of the TCR gene locus. Primer pairs were used to investigate the expression of mRNA for TdT (15): TCAGAGTTCTGAAACCCATCCT (antisense) and AGACTCCACCAATTGCTGTACA (sense); TEA: GGCAGACAGACTTGTCACTGGAT (antisense) and GGGACACTCCATGGTGTTGTTGTTG (sense); RAG-1 (12): CTTGGCTTTCCAGAGAGTCC (antisense) and TGGATCTTTACCTGAAGATG (sense); and RAG-2 (12): CATCATCTTCATTATAGGTGTC (antisense) and TGGAAGCAACATGGGAAATG (sense). The same PCR conditions were used for all sets of PCR. A total of 1–3 μl of cDNA was amplified using appropriate primer pairs (10 pmol of each), 0.2 mM dNTPs and 0.25 U of TaqDNA polymerase (Roche). Each reaction consisted of 32 cycles: 94°C for 20 s, 55°C for 30 s, and 72°C for 60 s. The last cycle was followed by a final extension period of 7 min at 72°C and then chilled to 4°C. After each set of reactions, the PCR products were visualized on a 2% agarose gel as described previously.

TdT, TEA, and RAG-1 products were sequenced. PCR products were cloned directly into TA-cloning vector (Invitrogen Life Technologies) and competent Escherichia coli transformed and grown overnight. White colonies were picked, subcultured, and plasmid DNA extracted using the QIAprep Spin Miniprep kit (Qiagen, Crawley, U.K.). The plasmids were screened for appropriately sized inserts using EcoRI restriction enzyme (Roche). Subsequently, DNA sequencing was performed using a T7 primer, by The Sequencing Service (School of Life Sciences, University of Dundee, Dundee, Scotland; www.dnaseq.co.uk). The derived nucleotide sequence was compared with the published sequences for TdT, TEA, and RAG-1 using the nucleotide basic local alignment search tool (BLAST) search facility provided by the National Center for Biotechnology Information (NCBI) (www.ncbi.nlm.nih.gov, National Center for Biotechnology Information, Public Database of Molecular Biology Information. NCBI, U.S. National Library of Medicine, 8600 Rockville Pike, Bethesda, MD 20894).

TCR β-chain V region (TCRBV) 4, 6, and 12 genes were sequenced in 11 samples from eight children. These are reported to be the three most commonly used TCR β-chains (22). Stored cDNA was used as a template for PCR amplifications, and each specific TCRBV PCR was conducted on all patient samples simultaneously. No contaminating sequences were found between unrelated patient samples. A primer specific to a conserved region of the TCR β-chain C region (TCACCCACCAGCTCAGCT) was used with a second primer specific for the appropriate TCRBV: TCRBV4 (CCACATATGAGAGTGGATTTGTCA), TCRBV6 (CAGTGATCGGTTCTCTGCAGA, primed for; 6s3, 6s4, 6s5, 6s7, 6s11, 6s14), TCRBV12 (TACTGACAAAGGAGAAGTCTCAGA, primed for; 12s2, 12s4).

For each reaction, a 1–3 μl template was amplified using appropriate primer pairs (2 nM), 1 mM dNTPs (Roche), and 0.4 U TaqDNA polymerase (Promega, Southampton, U.K.). PCR amplification of all TCRBVs was performed at 94°C for 20 s, 55°C for 30 s, and 72°C for 60 s, with a final extension period of 7 min at 72°C and then chilled to 4°C. For amplification of TCRBV6, the reaction consisted of 40 cycles; for TCRBV4 and TCRBV12 the reaction consisted of 36 cycles. After each set of reactions, the PCR products were resolved on 2% agarose gels. Appropriately sized bands were cut from the gel using sterile scalpel blades and purified using the QIAquick Gel Extraction kit (Qiagen) before cloning in the pCR2.1 TOPO TA cloning kit (Invitrogen Life Technologies), and sequencing as described previously. Derived sequences were compared with published sequences from the NCBI database, as previously described.

Frozen sections were cut at 5 μm, air-dried for 1 h, and then fixed for 10 min in precooled acetone at 4°C. Rehydrated sections were blocked with 10% normal goat and 10% human serum in PBS. Primary Abs were applied at room temperature for 3 h, and directly conjugated secondary Abs for 1 h at room temperature, with PBS washes before and after the secondaries. For double and triple labeling, all primary Abs were applied simultaneously. Sections were mounted in Vectashield (Vector Laboratories, Peterborough, U.K.). Controls included substitution of primary Abs with isotype-matched Ag-irrelevant Abs and incubations with secondary Ab only.

For both large and small bowel immunohistochemical sections, cell counting was performed using 5–7 crypt or crypt-villus units from three to four distinct well-orientated representative fields of view, at ×200 or ×400 magnification on a Leica DMR microscope with a triple filter and mercury lamp, using the Image Pro-Plus capture and analysis software (Media Cybernetics U.K., Berkshire, U.K.). Cell counts were made from measured areas of whole intestine and numbers of IEL or LPL found within the measured area were equated to cells/mm2.

Correlations between different cell types and age were assessed by regression analysis using Pearson correlation (r) and are indicated where appropriate in the text. Statistical significance was defined as p < 0.05. Data derived from flow cytometry analysis was separated into groups on a preweaning/postweaning basis to allow us to conduct two sample unpaired t tests and thus give a reflection of changes due to weaning. We proposed that 9 mo (6 mo after the usual onset of weaning) would be an appropriate time point for us to assume good exposure to food Ag had occurred and some adaptation of the T cells would be evident. Data derived from flow cytometry analysis was subjected to two sample unpaired t tests. Significance was defined as p < 0.05 by Student’s t test.

Aliquots of isolated intestinal lymphocytes from five small bowel samples (ages 6 days-12 mo, median = 2.5 mo) and 11 colonic samples (ages 1 day-4.75 years, median = 9 mo) were analyzed by flow cytometry. The distribution of CD4 and CD8 within the CD3+ gate of IEL and LPL fractions of small and large intestine was determined (Table II).

Table II.

Phenotype of lymphocytes in the small and large intestine by flow cytometrya

All Values Mean % (SD)IELsLPLs
AgeSmall intestineLarge intestineSmall intestineLarge intestine
<9 mo>9 mo<9 mo>9 mo<9 mo>9 mo<9 mo>9 mo
Samples (n13 13 
% Lymphocytes CD3+ Mean (SD) 11.4 (10.5) 28.0 (3.1) 21.0 (15.7) 33.8 (13.6) 28.03 (18.0) 64.27 (0.5) 51.59 (27.0) 54.18 (18.1) 
Samples (n
 3+4+ 18.2 (24.1) 46.9 (NA) 20.2 (16.5) 41.4 (24.5) 52.5 (14.9) 57.7 (NA) 57.6 (11.9) 51.5 (9.9) 
 3+8+ 20.2 (15.5) 26.9 (NA) 34 (23.1) 31.3 (10.6) 19.9 (8.4) 24.3 (NA) 24.6 (6.7) 27.7 (9.3) 
 3+4+8+ 3.3 (2.3) 3 (NA) 1.9 (1.8) 3.5 (4.2) 4.7 (3.9) 1.7 (NA) 2.2 (2.1) 6.4 (5.2) 
 3+48 58.3 (36.8) 23.2 (NA) 43.9 (33.9) 23.9 (15.1) 22.9 (14.0) 16.3 (NA) 15.6 (10.0) 14.5 (3.8) 
Samples (n
 37+2 8.7 (NA) 0.5 (NA) 9.3 (7.6) 3.9 (5.2) 5.9 (NA) 0.8 (NA) 5.7 (3.6) 1.9 (2.3) 
 37+2+ 5.3 (NA) 0.5 (NA) 6.2 (3.2) 3.4 (3.4) 6.2 (NA) 1.2 (NA) 8.2 (4.4) 6.0 (6.9) 
 3+7+2 5.6 (NA) 6.5 (NA) 2.5 (0.5) 2.0 (0.7) 8.2 (NA) 8.9 (NA) 6.4 (0.8) 3.5 (1.8) 
 3+72+ 13.5 (NA) 15.7 (NA) 23.7 (14.2) 22.3 (2.7) 34.4 (NA) 53.0 (NA) 60.3 (7.3) 45.6 (22.8) 
 7 (3+/−2+/−66.9 (NA) 76.9 (NA) 58.2 (16.6) 68.3 (6.6) 45.3 (NA) 36.1 (NA) 19.4 (9.5) 43.0 (33.8) 
Samples (n
 35+2 2.2 (NA) 10.8 (NA) 3.3 (2.2) 14.6 (20.0) 3.3 (NA) 6.4 (NA) 6.1 (4.4) 17.6 (22.1) 
 35+2+ 0.2 (NA) 0.2 (NA) 1.3 (1.1) 0.5 (0.6) 0.3 (NA) 0.3 (NA) 1.0 (0.6) 1.2 (0.8) 
 3+5+2 5.8 (NA) 7.9 (NA) 2.8 (0.7) 2.4 (1.5) 9.1 (NA) 11.1 (NA) 6.5 (0.8) 2.9 (2.1) 
 3+5+2+ 12.8 (NA) 14.7 (NA) 22.7 (12.8) 24.1 (8.0) 32.7 (NA) 52.2 (NA) 60.7 (9.0) 49.6 (24.5) 
 5 (3+/−2+/−79.0 (NA) 66.5 (NA) 69.9 (14.8) 58.4 (30.1) 54.5 (NA) 30.0 (NA) 25.6 (9.0) 28.7 (0.6) 
All Values Mean % (SD)IELsLPLs
AgeSmall intestineLarge intestineSmall intestineLarge intestine
<9 mo>9 mo<9 mo>9 mo<9 mo>9 mo<9 mo>9 mo
Samples (n13 13 
% Lymphocytes CD3+ Mean (SD) 11.4 (10.5) 28.0 (3.1) 21.0 (15.7) 33.8 (13.6) 28.03 (18.0) 64.27 (0.5) 51.59 (27.0) 54.18 (18.1) 
Samples (n
 3+4+ 18.2 (24.1) 46.9 (NA) 20.2 (16.5) 41.4 (24.5) 52.5 (14.9) 57.7 (NA) 57.6 (11.9) 51.5 (9.9) 
 3+8+ 20.2 (15.5) 26.9 (NA) 34 (23.1) 31.3 (10.6) 19.9 (8.4) 24.3 (NA) 24.6 (6.7) 27.7 (9.3) 
 3+4+8+ 3.3 (2.3) 3 (NA) 1.9 (1.8) 3.5 (4.2) 4.7 (3.9) 1.7 (NA) 2.2 (2.1) 6.4 (5.2) 
 3+48 58.3 (36.8) 23.2 (NA) 43.9 (33.9) 23.9 (15.1) 22.9 (14.0) 16.3 (NA) 15.6 (10.0) 14.5 (3.8) 
Samples (n
 37+2 8.7 (NA) 0.5 (NA) 9.3 (7.6) 3.9 (5.2) 5.9 (NA) 0.8 (NA) 5.7 (3.6) 1.9 (2.3) 
 37+2+ 5.3 (NA) 0.5 (NA) 6.2 (3.2) 3.4 (3.4) 6.2 (NA) 1.2 (NA) 8.2 (4.4) 6.0 (6.9) 
 3+7+2 5.6 (NA) 6.5 (NA) 2.5 (0.5) 2.0 (0.7) 8.2 (NA) 8.9 (NA) 6.4 (0.8) 3.5 (1.8) 
 3+72+ 13.5 (NA) 15.7 (NA) 23.7 (14.2) 22.3 (2.7) 34.4 (NA) 53.0 (NA) 60.3 (7.3) 45.6 (22.8) 
 7 (3+/−2+/−66.9 (NA) 76.9 (NA) 58.2 (16.6) 68.3 (6.6) 45.3 (NA) 36.1 (NA) 19.4 (9.5) 43.0 (33.8) 
Samples (n
 35+2 2.2 (NA) 10.8 (NA) 3.3 (2.2) 14.6 (20.0) 3.3 (NA) 6.4 (NA) 6.1 (4.4) 17.6 (22.1) 
 35+2+ 0.2 (NA) 0.2 (NA) 1.3 (1.1) 0.5 (0.6) 0.3 (NA) 0.3 (NA) 1.0 (0.6) 1.2 (0.8) 
 3+5+2 5.8 (NA) 7.9 (NA) 2.8 (0.7) 2.4 (1.5) 9.1 (NA) 11.1 (NA) 6.5 (0.8) 2.9 (2.1) 
 3+5+2+ 12.8 (NA) 14.7 (NA) 22.7 (12.8) 24.1 (8.0) 32.7 (NA) 52.2 (NA) 60.7 (9.0) 49.6 (24.5) 
 5 (3+/−2+/−79.0 (NA) 66.5 (NA) 69.9 (14.8) 58.4 (30.1) 54.5 (NA) 30.0 (NA) 25.6 (9.0) 28.7 (0.6) 
a

The phenotype of IEL and LPL fractions of cells derived from the small and the large intestine were analyzed by FACS. Cells were gated as lymphocytes based on known boundaries of forward scatter (FSC) and side scatter (SSC) defined with PBLs and LPLs, then gated as CD3 and CD3+ lymphocytes based on CD3 expression. Isotype negative controls were used to determine the CD3+/− threshold, and the position of the quadrants within the CD3+ and CD3 populations. CD3low cells were included with the CD3+ population for analysis of CD4/CD8 expression. For the purpose of CD2/CD5 and CD2/CD7 pre-T cell marker analysis, cells expressing low levels of CD3 but above the nonspecific threshold (determined with IgG isotype controls) were included as CD3 cells as these were most likely to contain immature pre-T cells. Results of CD3/CD4/CD8 expression are expressed as a percentage of the total lymphocytes satisfying both the CD3+ gate and the FSC/SSC lymphocyte gate, mean percentage values (SD). Results of CD2/CD7 and CD2/CD5 expression are expressed as a percentage of total lymphocytes within the FSC/SSC lymphocyte gate, mean percentage values (SD). The proportion of DN cells decreased with age, while that of total SP cells (CD3+4+ and CD3+8+) increased, and that of DP cells changed little. In analysis of infants aged <5 years, CD3+CD4CD8 (DN) cells were negatively correlated with age: IEL −0.86 and −0.64 (for ileum and colon, respectively) and LPL −0.41 and −0.79. Pre-T cells expressing CD7 (CD3CD7+2+/− by FACS) tended to decrease with age (NS).

Most strikingly, numerous CD3+ cells expressed neither CD4 nor CD8 (DN cells). The proportion of DN cells decreased with age, so that while DN cells accounted for 99% of CD3+ IEL of the ileum of the 6-day-old child (patient A), in the 12-mo-old child (patient K) just 23% of CD3+ IEL were DN. There was an inverse correlation between age and the proportion of DN cells in the small intestine (r = −0.52 and −0.47 for IEL and LPL, respectively), although this trend was not significant. These data indicate that DN cells are present in both epithelial and lamina propria compartments of small and large intestine in early infancy. In the epithelial compartment, they appear to become less numerous as the child matures, although this trend is less clear in the lamina propria.

Cells coexpressing CD4 and CD8 (double-positive (DP) cells) accounted for up to 10% of CD3+ cells in both IEL and LPL. The proportion of DP cells changed little with age.

The proportion of total SP cells (CD3+4+ plus CD3+8+ cells) increased with age. There was a significant negative correlation between the proportion of DN cells and total SP cells in both the small (r = −0.99, p < 0.0001, IEL; and r = −0.97, p < 0.006 LPL) and large intestine (r = −0.99, p < 0.0001, IEL; and r = −0.89, p < 0.0003 LPL). When the CD8+ fraction of SP cells was studied, there was a strong negative correlation between CD3+8+ cells and CD3+ DN cells in the IEL compartment. This was significant in the large intestine (age <5 years, n = 11, r = −0.65, p < 0.03) but not in the small intestine (ages <5 years, n = 5, r = −0.85, p = 0.067) although the study may be underpowered. This indicates that in the epithelium, DN cells decrease concordant with increased CD3+8+ cells, most probably due to recruitment of CD3+8+ cells to the IEL compartment.

The increase in CD4+ and CD8+ SP cells, and decrease in DN cells, with age, was confirmed by observations of a single patient (D) from whom two large bowel samples (PT7 and PT13) were obtained at 6 and 9 mo of age. Between these ages this patient displayed an increased proportion of CD3+4+ SP IEL (from 3.2 to 20.1%), an increased proportion of CD3+8+ SP IEL (from 1.8 to 49.0%) and LPL (from 14.0 to 32.3%), and a decreased proportion of CD3+ DN IEL (from 94.8 to 27.5%) and LPL (from 30.3 to 13.1%).

To further investigate the expression of CD4 and CD8, immunohistochemistry studies were performed. Cell counting was conducted on five ileal samples (ages 6 days-12 mo) and eight colonic samples (ages 1 day-18 mo) (Table III). These studies also showed that CD4+ cells were often as numerous as CD8+ cells, indeed CD4+ cells accounted for 20–77% of T cells in the ileal mucosa and 54–78% in the colonic mucosa.

Table III.

Phenotype of lymphocytes in the small and large intestine by immunohistochemistrya

IdSampleAgeCD4-CD8 Double Fluorescent ImmunohistochemistryCD3-CD19-CD16a-CD5-CD7 Triple Fluorescent Immunohistochemistry
IELsLPLs (%)IELsLPLs (%)
CD8+ IELCD4+CD8+CD4+8+ (DP)CD5+CD7CD7+CD5CD7+CD5+CD5+CD7CD7+CD5CD7+CD5+
Small intestine             
 A PT2 6 days 129 94 (20) 223 (48) 152 (32) 0 (0) 38 (100) 0 (0) 249 (43) 328 (56) 2 (<1) 
 Q PT27 2.1 mo 108 590 (66) 271 (30) 38 (4) 8 (16) 38 (78) 3 (6) 42 (13) 254 (77) 34 (10) 
 E PT9 2.3 mo 74 155 (42) 173 (47) 42 (11) 0 (0) 73 (100) 0 (0) 92 (23) 241 (60) 70 (17) 
 M PT23 2.8 mo 202 1482 (67) 578 (26) 158 (7) 12 (12) 81 (82) 6 (6) 107 (17) 390 (62) 131 (21) 
 B PT3 3.8 mo 138 (70) 49 (25) 11 (6) 0 (0) 165 (100) 0 (0) 206 (20) 725 (71) 97 (9) 
 K PT18 12 mo 200 808 (77) 124 (12) 113 (11) 0 (0) 88 (82) 20 (19) 123 (9) 716 (52) 540 (39) 
 Mean (SD)  <9 mo 104.4 (71) 491 (588.9) 258.8 (196.7) 80.2 (69.3) 4 (5.7) 79 (52) 1.8 (2.7) 139.2 (85.5) 387.6 (198) 66.8 (50.8) 
  >9 mo 200 (NA) 808 (NA) 124 (NA) 113 (NA) 0 (NA) 88 (NA) 20 (NA) 123 (NA) 716 (NA) 540 (NA) 
 Mean % (SD)  <9 mo  52.8 (21.4) 35.1 (11.2) 12.1 (11.6) 5.7 (7.9) 91.9 (11.2) 2.4 (3.3) 23.1 (11.7) 65.2 (8.3) 11.7 (7.9) 
  >9 mo  77.3 (NA) 11.9 (NA) 10.8 (NA) 0.0 (NA) 81.5 (NA) 18.5 (NA) 8.9 (NA) 51.9 (NA) 39.2 (NA) 
Large intestine             
 O PT25 1 day 18 259 (54) 90 (19) 131 (27) 0 (0) 14 (82) 3 (18) 12 (11) 77 (73) 16 (15) 
 P PT26 5 days 165 564 (55) 411 (40) 43 (4) 9 (8) 46 (41) 57 (51) 29 (11) 130 (49) 105 (40) 
 E PT10 2.3 mo 97 1502 (61) 462 (19) 502 (20) 0 (0) 23 (62) 14 (28) 93 (20) 265 (56) 113 (24) 
 L PT20 3.9 mo 25 2075 (81) 393 (15) 99 (4) 0 (0) 146 (100) 0 (0) 58 (12) 393 (79) 46 (9) 
 D PT7 6.1 mo 34 184 (56) 106 (32) 38 (12) 0 (0) 12 (100) 0 (0) 38 (6) 375 (63) 179 (30) 
 F PT11 9 mo ND ND ND ND 0 (0) 22 (100) 0 (0) 44 (12) 281 (78) 35 (10) 
 D PT13 9.3 mo 160 1050 (78) 258 (19) 36 (3) No cells No cells No cells 155 (11) 938 (66) 328 (23) 
 N PT24 17.5 mo 931 2700 (55) 1960 (40) 224 (5) 0 (0) 39 (100) 0 (0) 192 (35) 235 (42) 125 (23) 
 I PT14 18 mo 30 1131 (74) 264 (17) 138 (9) ND ND ND ND ND ND 
 Mean (SD)  <9 mo 83.2 (67.6) 939 (747.4) 286.7 (161) 141.5 (180.8) 1.3 (3.4) 37.6 (49.8) 10.6 (21.1) 61.3 (48.5) 351.3 (283.8) 117.4 (108.4) 
  >9 mo 480.5 (637.1) 1915.5 (1109.5) 1112 (1199.3) 181 (60.8) 0 (NA) 39 (NA) 0 (NA) 192 (NA) 235 (NA) 125 (NA) 
 Mean % (SD)  <9 mo  64.2 (12.1) 24.1 (9.9) 11.7 (10.2) 1.3 (3.3) 80.9 (24.6) 17.7 (22.1) 11.9 (4) 66.5 (11.2) 21.6 (11.2) 
  >9 mo  64.5 (13.1) 28.7 (16.2) 6.8 (3.1) 0.0 (NA) 100.0 (NA) 0.0 (NA) 34.8 (NA) 42.6 (NA) 22.6 (NA) 
IdSampleAgeCD4-CD8 Double Fluorescent ImmunohistochemistryCD3-CD19-CD16a-CD5-CD7 Triple Fluorescent Immunohistochemistry
IELsLPLs (%)IELsLPLs (%)
CD8+ IELCD4+CD8+CD4+8+ (DP)CD5+CD7CD7+CD5CD7+CD5+CD5+CD7CD7+CD5CD7+CD5+
Small intestine             
 A PT2 6 days 129 94 (20) 223 (48) 152 (32) 0 (0) 38 (100) 0 (0) 249 (43) 328 (56) 2 (<1) 
 Q PT27 2.1 mo 108 590 (66) 271 (30) 38 (4) 8 (16) 38 (78) 3 (6) 42 (13) 254 (77) 34 (10) 
 E PT9 2.3 mo 74 155 (42) 173 (47) 42 (11) 0 (0) 73 (100) 0 (0) 92 (23) 241 (60) 70 (17) 
 M PT23 2.8 mo 202 1482 (67) 578 (26) 158 (7) 12 (12) 81 (82) 6 (6) 107 (17) 390 (62) 131 (21) 
 B PT3 3.8 mo 138 (70) 49 (25) 11 (6) 0 (0) 165 (100) 0 (0) 206 (20) 725 (71) 97 (9) 
 K PT18 12 mo 200 808 (77) 124 (12) 113 (11) 0 (0) 88 (82) 20 (19) 123 (9) 716 (52) 540 (39) 
 Mean (SD)  <9 mo 104.4 (71) 491 (588.9) 258.8 (196.7) 80.2 (69.3) 4 (5.7) 79 (52) 1.8 (2.7) 139.2 (85.5) 387.6 (198) 66.8 (50.8) 
  >9 mo 200 (NA) 808 (NA) 124 (NA) 113 (NA) 0 (NA) 88 (NA) 20 (NA) 123 (NA) 716 (NA) 540 (NA) 
 Mean % (SD)  <9 mo  52.8 (21.4) 35.1 (11.2) 12.1 (11.6) 5.7 (7.9) 91.9 (11.2) 2.4 (3.3) 23.1 (11.7) 65.2 (8.3) 11.7 (7.9) 
  >9 mo  77.3 (NA) 11.9 (NA) 10.8 (NA) 0.0 (NA) 81.5 (NA) 18.5 (NA) 8.9 (NA) 51.9 (NA) 39.2 (NA) 
Large intestine             
 O PT25 1 day 18 259 (54) 90 (19) 131 (27) 0 (0) 14 (82) 3 (18) 12 (11) 77 (73) 16 (15) 
 P PT26 5 days 165 564 (55) 411 (40) 43 (4) 9 (8) 46 (41) 57 (51) 29 (11) 130 (49) 105 (40) 
 E PT10 2.3 mo 97 1502 (61) 462 (19) 502 (20) 0 (0) 23 (62) 14 (28) 93 (20) 265 (56) 113 (24) 
 L PT20 3.9 mo 25 2075 (81) 393 (15) 99 (4) 0 (0) 146 (100) 0 (0) 58 (12) 393 (79) 46 (9) 
 D PT7 6.1 mo 34 184 (56) 106 (32) 38 (12) 0 (0) 12 (100) 0 (0) 38 (6) 375 (63) 179 (30) 
 F PT11 9 mo ND ND ND ND 0 (0) 22 (100) 0 (0) 44 (12) 281 (78) 35 (10) 
 D PT13 9.3 mo 160 1050 (78) 258 (19) 36 (3) No cells No cells No cells 155 (11) 938 (66) 328 (23) 
 N PT24 17.5 mo 931 2700 (55) 1960 (40) 224 (5) 0 (0) 39 (100) 0 (0) 192 (35) 235 (42) 125 (23) 
 I PT14 18 mo 30 1131 (74) 264 (17) 138 (9) ND ND ND ND ND ND 
 Mean (SD)  <9 mo 83.2 (67.6) 939 (747.4) 286.7 (161) 141.5 (180.8) 1.3 (3.4) 37.6 (49.8) 10.6 (21.1) 61.3 (48.5) 351.3 (283.8) 117.4 (108.4) 
  >9 mo 480.5 (637.1) 1915.5 (1109.5) 1112 (1199.3) 181 (60.8) 0 (NA) 39 (NA) 0 (NA) 192 (NA) 235 (NA) 125 (NA) 
 Mean % (SD)  <9 mo  64.2 (12.1) 24.1 (9.9) 11.7 (10.2) 1.3 (3.3) 80.9 (24.6) 17.7 (22.1) 11.9 (4) 66.5 (11.2) 21.6 (11.2) 
  >9 mo  64.5 (13.1) 28.7 (16.2) 6.8 (3.1) 0.0 (NA) 100.0 (NA) 0.0 (NA) 34.8 (NA) 42.6 (NA) 22.6 (NA) 
a

The phenotype of IEL and LPL fractions of cells derived from the small and large intestine were analyzed by double (CD4/CD8) and triple (CD3/CD19/CD16a/CD5/CD7) fluorescent immunohistochemistry. For CD5/CD7 triple immunohistochemistry, FITC was used to exclude any cells expressing CD3, CD16a, or CD19 from the count. Individual subject data are expressed as cell counts per 1 mm2 of whole intestinal tissue, with relative percentage expression in parentheses. All data for individual subjects are included, and the mean and SD of absolute cell counts and relative percentage expression are summarized for subjects aged <9 mo and for subjects aged >9 mo. CD4CD8 (DN) cells could not be counted. There was a surprisingly large number of CD4+ cells in both the small and large intestine. Immature T cells (CD4+8+ and CD316a197+5+/−) were observed in the small and large intestine. The proportion of CD4+CD8+ (DP) cells decreased in subjects aged >9 mo (NS), which may reflect the trend for increased numbers of mature SP T cells. In the lamina propria, the proportion of pre-T cells (CD316a197+5) was decreased in subjects aged >9 mo (NS), again this may reflect an increase in mature T cells rather than an absolute decrease in pre-T cells.

The results suggest that ontogeny of LPL preceded that of IEL, i.e., in 9 of 11 samples aged 9 mo or less, IEL contained a smaller proportion of CD3+ SP cells and a larger proportion of immature cells than LPL. Soon after birth, both the small and the large intestine are populated by numerous CD3+ LPL that have already differentiated into SP cells and, with increasing age, the remaining immature (DN and DP) cells (except for a residual 10% or so which remain immature) differentiate into, or are replaced by, SP cells. This was particularly apparent in the small intestine, where shortly after birth 99% of CD3+ IEL were DN and 0.4% were CD3+ SP, whereas in the LPL fraction 60% of CD3+ cells were already SP and only 40% were DN.

Flow cytometry demonstrated a large proportion of DN cells and that most were lymphocytes. Those DN cells that were CD3+ have been described above. However, there appeared to be other cells of similar light scatter characteristics that did not stain for CD3, suggesting that these may be CD3 lymphocyte precursors. To clarify this, expression of CD2, CD5, and CD7 were assessed in the IEL and LPL from seven subjects (age 1 day-27 mo, Table II). In both compartments of the small and large intestine, four quadrants were defined within the CD3+/− gated lymphocytes based on their expression of CD2 and CD5, or CD2 and CD7.

CD37+2+ and CD37+2 populations, which are both most likely early T cell precursors, were at highest levels in tissues from subjects aged 9 mo or less, dropping to almost undetectable levels in subjects aged >9 mo (Table II). CD3CD7+CD2 cells (the phenotype of bone marrow-derived T cells or NK precursors) were found in IEL and LPL in small and large intestine. CD5+CD2+ cells were also present in both immunologic compartments in the small and large intestine. Most of these cells also stained positively for CD3, but up to 11% of the CD5+CD2+ lymphocytes were CD3, and these were likely to be immature T cells and unlikely to be precursors of either NK or B cells. For each individual sample, the proportion of CD5+CD2+ cells was greater in the LPL than the IEL fraction. The proportion of CD5+CD2+CD3 lymphocytes tended to decrease with age, with the maximum expression observed in the 1-day-old child (O), accounting for 6.5% of CD3 lymphocytes in the LPL fraction (1.6% of total lymphocytes), and 2.9% of CD3 lymphocytes in the IEL fraction (2.2% of total lymphocytes). All intestinal regions showed an increase in proportion of the CD35+2 subpopulation, possibly B cells, from <9 mo to >9 mo. The presence of immature cells was also investigated by triple immunofluorescence histology (Table III, Fig. 1). CD5 is expressed by both T and B cells, CD7 by T, NK, and their precursors. Thus, only T cell precursors will express both CD5 and CD7, whereas the presence of CD3, CD16 and CD19 would indicate mature T, NK, and B cells, respectively. In the small intestine (n = 6), SP CD7+ cells were found in both the epithelium (38–165/mm2) and the lamina propria (241–725/mm2), while SP CD5+ cells were less numerous (42–249/mm2) in the lamina propria and essentially absent from the epithelium. In the colon (n = 9), SP CD7+ cells were found in both the epithelium (0–146/mm2) and the lamina propria (77–938/mm2), while SP CD5+ cells were less numerous (12–192/mm2) in the lamina propria and, again, essentially absent from the epithelium. CD5+7+ DP cells were present in every lamina propria sample (2–540/mm2 in the ileum, 16–328/mm2 in the colon), although very few CD5+7+ DP cells were found in the epithelium (0–20/mm2 in the ileum, 0–57/mm2 in the colon).

FIGURE 1.

CD3 lymphocytes in the human infant intestine express surface membrane Ags of immature T cells. Ileal tissues from patients aged 6 days (a) and 3.9 mo (b) underwent immunohistochemistry with FITC-conjugated CD3, CD16a, and CD19, Texas Red-conjugated CD5 (CD35+, red), and 7-amino-4-methylcoumarin-3-acetic acid (AMCA)-conjugated CD7 (CD37+, blue). Colonic tissues from patients aged 4 mo (c) and 9 mo (d) underwent immunohistochemistry with AMCA-conjugated CD3, CD16a and CD19, FITC-conjugated CD5 (CD35+, green), and Texas Red-conjugated CD7 (CD37+, red). Scale bars indicate 100 μm. CD37+ pre-T cells were observed in the lamina propria, and occasionally in the epithelium (arrowed). Occasionally pre-T cells were observed coexpressing CD5 and CD7 but without expression of CD3, CD16, or CD19. These cells are the phenotype of bone marrow-derived precursors. For the orientation of villi and crypts, Lu indicates the luminal surface.

FIGURE 1.

CD3 lymphocytes in the human infant intestine express surface membrane Ags of immature T cells. Ileal tissues from patients aged 6 days (a) and 3.9 mo (b) underwent immunohistochemistry with FITC-conjugated CD3, CD16a, and CD19, Texas Red-conjugated CD5 (CD35+, red), and 7-amino-4-methylcoumarin-3-acetic acid (AMCA)-conjugated CD7 (CD37+, blue). Colonic tissues from patients aged 4 mo (c) and 9 mo (d) underwent immunohistochemistry with AMCA-conjugated CD3, CD16a and CD19, FITC-conjugated CD5 (CD35+, green), and Texas Red-conjugated CD7 (CD37+, red). Scale bars indicate 100 μm. CD37+ pre-T cells were observed in the lamina propria, and occasionally in the epithelium (arrowed). Occasionally pre-T cells were observed coexpressing CD5 and CD7 but without expression of CD3, CD16, or CD19. These cells are the phenotype of bone marrow-derived precursors. For the orientation of villi and crypts, Lu indicates the luminal surface.

Close modal

The data show that T cell precursors are present at significant levels at birth and that subsequent differentiation to SP mature T cells takes place in the mucosal population over the next 12–18 mo. To define whether this maturation takes place in situ, involving gene rearrangement and generation of a diverse TCR repertoire, rather than at an extra-mucosal site with subsequent mucosal seeding of precursors, we examined the evidence for these processes in infant mucosal T cells.

PCR products of the appropriate size for TEA were obtained from 5 of 11 cDNA samples (age 3 days-18 years) (Table IV). All samples from infants aged 6 mo or less contained TEA (n = 5). The identity of the cDNA was confirmed by cloning and sequencing.

Table IV.

Expression of TdT, TEA, and RAG genes in the intestine during childhooda

IdAgeIntestineTEATdTRAG-1RAG-2
AA 3 days SI + + + + 
A-Cyst 6 days SI cyst ND ND Neg Neg 
A-Ileum 6 days SI ND ND Neg 
AB 2 mo SI + + + + 
AC 2 mo Small + + + Neg 
3.8 mo SI ND ND Neg 
3.9 mo SI ND ND Neg Neg 
AD 5 mo SI + + + + 
12 mo SI ND ND Neg Neg 
AE 18 mo SI Neg ND ND 
AF 15 years SI Neg Neg ND ND 
AG 16 years SI Neg Neg ND ND 
2.3 mo LI ND ND Neg 
3.9 mo LI ND ND Neg 
AH 6 mo LI + + + + 
6.1 mo LI ND ND Neg 
6.1 mo Anus ND ND Neg 
9 mo LI ND ND Neg 
AI 9 mo LI Neg Neg ND ND 
9.3 mo LI ND ND Neg 
12 mo Rectum ND ND Neg 
AJ 16 mo LI Neg + + Neg 
18 mo LI ND ND Neg 
AK 18 years LI Neg Neg ND ND 
IdAgeIntestineTEATdTRAG-1RAG-2
AA 3 days SI + + + + 
A-Cyst 6 days SI cyst ND ND Neg Neg 
A-Ileum 6 days SI ND ND Neg 
AB 2 mo SI + + + + 
AC 2 mo Small + + + Neg 
3.8 mo SI ND ND Neg 
3.9 mo SI ND ND Neg Neg 
AD 5 mo SI + + + + 
12 mo SI ND ND Neg Neg 
AE 18 mo SI Neg ND ND 
AF 15 years SI Neg Neg ND ND 
AG 16 years SI Neg Neg ND ND 
2.3 mo LI ND ND Neg 
3.9 mo LI ND ND Neg 
AH 6 mo LI + + + + 
6.1 mo LI ND ND Neg 
6.1 mo Anus ND ND Neg 
9 mo LI ND ND Neg 
AI 9 mo LI Neg Neg ND ND 
9.3 mo LI ND ND Neg 
12 mo Rectum ND ND Neg 
AJ 16 mo LI Neg + + Neg 
18 mo LI ND ND Neg 
AK 18 years LI Neg Neg ND ND 
a

TdT, TEA, and RAG gene expression was assayed by PCR. TEA and TdT were expressed in samples from infants aged up to 18 mo (5 of 8 and 7 of 8, respectively). RAG-1 was found in 15 of 19 samples studied, while RAG-2 was only found in 5 of 19 samples. There was concordance between the expression of TdT and RAG-1 in 6 of 6 samples and between TEA and RAG-1 in 5 of 6 samples (indicated by bold/underlining). SI, small intestine; LI, large intestine.

Seven of 11 cDNA samples yielded PCR products of the appropriate size for TdT (Table IV). The identity was confirmed by cloning and sequencing. As for TEA, all the positive samples came from infants. All samples from children aged 6 mo or less contained TdT transcripts (n = 5).

These data suggest that TEA and TdT genes are transcribed in the intestine during the first 18 mo of life, but are unlikely to be transcribed later. All samples from infants aged 6 mo or less contained transcripts for TEA and TdT. Seven samples were obtained from the small intestine, five contained the transcripts of interest. Too few large bowel samples were studied to conclude that there were fewer transcripts than in the small bowel.

Six samples used in the TdT and TEA studies were also analyzed for RAG transcript expression (Table IV). All contained transcripts for RAG-1 or RAG-1 and RAG-2. In total, 19 cDNA samples were analyzed and RAG-1 was found in 15, while RAG-2 was only found in 5. Concurrent expression of RAG-1 and RAG-2 was found in 4 samples.

Although TEA expression is thought to be specific for T cells, TdT and RAG may be expressed by non-T cells. In additional experiments, intestinal expression of TdT protein was investigated by immunohistochemistry in seven infants aged 18 mo or less (n = 7). Thymus tissue was a positive control and an isotype-matched Ab the negative control. TdT protein was detected in four of four small intestinal samples from three different infants (A-cyst, A-Ileum, B and L) and two of four large intestinal samples from four different infants (E and F positive, L and I negative). TdT protein was expressed mostly by cells in the lamina propria (Fig. 2, c and d), at least some of which were CD3+ (Fig. 2 e), and by the occasional cell in the epithelium.

FIGURE 2.

TdT protein is expressed in the human infant intestine. TdT protein expression was investigated by immunohistochemistry in small intestinal samples from patients aged 6 days to 4 mo and large intestinal samples from patients aged 2–18 mo. An isotype control Ab (a) was used as a negative control in all tissues. Thymus was used as positive control tissue (b). TdT protein was expressed by cells in the infant small intestine (four of four positive) and large intestine (two of four positive). Examples of TdT+ immunohistochemistry are shown for the small intestine (aged 6 days (c), and 3.9 mo (d), TdT+-DAB, brown) and large intestine (e, TdT-FITC, CD3-Texas Red). TdT protein expression was observed by cells in the lamina propria and occasionally in the epithelium. Although some cells displayed a clear nuclear distribution of TdT (filled arrow), many expressed TdT outside the nucleus. This is consistent with the translation of TdT protein and protein folding in the Golgi and endoplasmic reticulum before translocation to the nucleus where it is required for TCR recombination. Occasionally CD3+ cells were TdT+ (open arrow), but the majority of TdT+ cells were CD3. Scale bars indicate 100 μm.

FIGURE 2.

TdT protein is expressed in the human infant intestine. TdT protein expression was investigated by immunohistochemistry in small intestinal samples from patients aged 6 days to 4 mo and large intestinal samples from patients aged 2–18 mo. An isotype control Ab (a) was used as a negative control in all tissues. Thymus was used as positive control tissue (b). TdT protein was expressed by cells in the infant small intestine (four of four positive) and large intestine (two of four positive). Examples of TdT+ immunohistochemistry are shown for the small intestine (aged 6 days (c), and 3.9 mo (d), TdT+-DAB, brown) and large intestine (e, TdT-FITC, CD3-Texas Red). TdT protein expression was observed by cells in the lamina propria and occasionally in the epithelium. Although some cells displayed a clear nuclear distribution of TdT (filled arrow), many expressed TdT outside the nucleus. This is consistent with the translation of TdT protein and protein folding in the Golgi and endoplasmic reticulum before translocation to the nucleus where it is required for TCR recombination. Occasionally CD3+ cells were TdT+ (open arrow), but the majority of TdT+ cells were CD3. Scale bars indicate 100 μm.

Close modal

TCR β-chain junctional sequences were obtained from three different families of TCRBV: TCRBV4, TCRBV6, and TCRBV12 (Table V). Initially, cDNA derived from small or large bowel mucosa whole tissue was used as the template. Subsequently, for two small intestinal samples and three large intestinal samples, cDNA from separated LPL and/or IEL was used. Two samples were obtained from a 6-day-old child (patient A) with an ileal duplication cyst. This allowed the effect of luminal contents on the small intestinal TCR repertoire to be determined as the cyst lay parallel to, but not in continuity with, the ileum. Small and large intestinal samples were obtained from a 2- and 1/4-mo-old infant (patient E) undergoing an ileostomy reversal. As this was created at birth, the colon had never been exposed to food Ags. Two colon samples were obtained from one child (patient D) separated by a 3-mo interval; samples were taken at 6 and 9 mo of age. Over 800 junctional regions were sequenced. The data presented are derived from samples where at least 20 junctional regions were sequenced, unless the first 10 sequences for a particular sample were identical.

Table V.

Summary of small and large intestinal T cell repertoirea

Patient IdAge (mo)Intestinal SourceTCRVB% UniqueNo. Clones StudiedNo. Different TCRVB SequencesNo. Expanded Clones-(Copies Per Clone)Mean Junctional Length
2.3 SI (M) 14 1 (14) 14.0 
12 SI (M) 15 1 (15) 10.0 
A (Cyst) 0.2 SI (M) 96 24 23 1 (2) 12.9 
A (Ileum) 0.2 SI (M) 86 21 18 3 (2, 2, 2) 12.4 
12 SI (M) 100 21 21 14.3 
  SI (LPL) 90 19 17 2 (2, 2) 13.3 
A (Cyst) 0.2 SI (M) 12 100 20 20 11.7 
A (Ileum) 0.2 SI (M) 12 70 27 19 5 (2, 2, 2, 3, 4) 12.5 
3.8 SI (LPL) 12 90 20 18 2 (2, 2) 12.1 
12 SI (M) 12 91 21 19 1 (3) 13.4 
  SI (LPL) 12 96 22 21 1 (2) 12.8 
2.3 LI (M) 81 21 17 3 (2, 2, 3) 12.4 
6.1 LI (M) 61 28 17 6 (2, 2, 2, 3, 3, 5) 13.4 
  LI (LPL) 81 26 21 4 (2, 2, 2, 3) 12.3 
9.3 LI (M) 50 28 14 6 (2, 2, 2, 4, 4, 6) 13.4 
9.0 LI (M) 74 23 17 4 (2, 2, 4) 12.7 
  LI (LPL) 28 29 6 (2, 2, 3, 6, 7, 7) 13.0 
12.0 LI (M) 38 21 3 (3, 6, 7) 11.9 
18.0 LI (M) 74 19 14 4 (2, 2, 2, 3) 14.4 
6.1 LI (M) 57 23 13 6 (2, 2, 2, 3, 3, 4) 14.2 
  LI (LPL) 95 22 21 1 (2) 13.9 
9.0 LI (M) 87 23 20 3 (2, 2, 2) 13.0 
  LI (LPL) 70 23 16 5 (2, 2, 2, 2, 4) 13.7 
12.0 LI (M) 87 23 20 3 (2, 2, 2) 14.4 
18.0 LI (M) 75 24 18 3 (2, 3, 4) 13.6 
  LI (IEL) 100 21 21 14.1 
  LI (LPL) 100 23 23 13.8 
56.0 LI (IELs) 29 24 5 (3, 3, 5, 5, 6) 13.9 
6.1 LI (M) 12 43 21 6 (2, 2, 2, 2, 4, 6) 13.4 
9.0 LI (M) 12 70 23 16 5 (2, 2, 2, 3, 3) 13.9 
12.0 LI (M) 12 67 27 18 2 (3, 8) 13.1 
18.0 LI (LPL) 12 100 23 23 12.8 
Patient IdAge (mo)Intestinal SourceTCRVB% UniqueNo. Clones StudiedNo. Different TCRVB SequencesNo. Expanded Clones-(Copies Per Clone)Mean Junctional Length
2.3 SI (M) 14 1 (14) 14.0 
12 SI (M) 15 1 (15) 10.0 
A (Cyst) 0.2 SI (M) 96 24 23 1 (2) 12.9 
A (Ileum) 0.2 SI (M) 86 21 18 3 (2, 2, 2) 12.4 
12 SI (M) 100 21 21 14.3 
  SI (LPL) 90 19 17 2 (2, 2) 13.3 
A (Cyst) 0.2 SI (M) 12 100 20 20 11.7 
A (Ileum) 0.2 SI (M) 12 70 27 19 5 (2, 2, 2, 3, 4) 12.5 
3.8 SI (LPL) 12 90 20 18 2 (2, 2) 12.1 
12 SI (M) 12 91 21 19 1 (3) 13.4 
  SI (LPL) 12 96 22 21 1 (2) 12.8 
2.3 LI (M) 81 21 17 3 (2, 2, 3) 12.4 
6.1 LI (M) 61 28 17 6 (2, 2, 2, 3, 3, 5) 13.4 
  LI (LPL) 81 26 21 4 (2, 2, 2, 3) 12.3 
9.3 LI (M) 50 28 14 6 (2, 2, 2, 4, 4, 6) 13.4 
9.0 LI (M) 74 23 17 4 (2, 2, 4) 12.7 
  LI (LPL) 28 29 6 (2, 2, 3, 6, 7, 7) 13.0 
12.0 LI (M) 38 21 3 (3, 6, 7) 11.9 
18.0 LI (M) 74 19 14 4 (2, 2, 2, 3) 14.4 
6.1 LI (M) 57 23 13 6 (2, 2, 2, 3, 3, 4) 14.2 
  LI (LPL) 95 22 21 1 (2) 13.9 
9.0 LI (M) 87 23 20 3 (2, 2, 2) 13.0 
  LI (LPL) 70 23 16 5 (2, 2, 2, 2, 4) 13.7 
12.0 LI (M) 87 23 20 3 (2, 2, 2) 14.4 
18.0 LI (M) 75 24 18 3 (2, 3, 4) 13.6 
  LI (IEL) 100 21 21 14.1 
  LI (LPL) 100 23 23 13.8 
56.0 LI (IELs) 29 24 5 (3, 3, 5, 5, 6) 13.9 
6.1 LI (M) 12 43 21 6 (2, 2, 2, 2, 4, 6) 13.4 
9.0 LI (M) 12 70 23 16 5 (2, 2, 2, 3, 3) 13.9 
12.0 LI (M) 12 67 27 18 2 (3, 8) 13.1 
18.0 LI (LPL) 12 100 23 23 12.8 
a

TCR β-chain gene junctional sequences were obtained from three different families of β-chains: TCRBV4, TCRBV6, and TCRBV12. cDNA was derived from whole mucosa tissue (M), separated LPLs (LPL), or separated IELs (IEL) of the small and large bowel. The data is presented as the percentage of repertoire attributed to unique sequences (m, monoclonal), the total number of clones studied, the number of different TCRBV sequences identified, and the number of expanded clones found with numbers of clones expressing the same sequence in parentheses, within each sample. In the small intestine (SI), the repertoire of TCRBV6 and 12 were polyclonal whereas TCRBV 4 was monoclonal. In the large intestine (LI), in infants aged 18 mo or less the repertoire of TCRBV4, 6, and 12 were polyclonal. In direct contrast to the polyclonal TCRBV6 expression in the IELs of the 18-mo-old child, the child aged 4 years 9 mo showed a TCRBV6 IEL repertoire which was oligoclonal containing only seven unique clones.

In the small intestine, the TCRBV6 and TCRBV12 repertoire were polyclonal in all subjects studied (Table V). Few clones were identified more than once in whole tissue and LPL samples. In the normal ileum (2T6) and cyst (1T6) samples from the 6-day-old child (A), none of the 41 unique TCRBV6 clones identified were shared between the two sites. Furthermore, none of the 39 unique TCRBV12 clones were shared between the two sites. In a sample from a 12-mo-old (K), 38 of 43 TCRBV12 clones were unique and only one TCRBV12 sequence (TCRBV12s2-CAIGGQTKNIQY-TCRBJ2.4) from this patient was found to be common between the mucosa and the LPL fraction and was used by only two clones. In contrast, the TCRBV4 repertoire in the small intestine of both the 2-mo-old and the 1-year-old child were monoclonal. The finding of polyclonality and the lack of shared clones suggests the lymphocytes had not homed to the small intestine, normal ileum, or cyst under any specific antigenic influence, but arrived by chance, i.e., that luminal Ags do not influence homing at this age.

In the large intestine, few clonal expansions of TCRBV6 and TCRBV12 were evident in the mucosal samples (Table V). Of the 21–27 sequences obtained per sample, most were different and six or fewer clones were found more than once. Occasionally a mucosal tissue sample displayed a more restricted TCRBV6 or TCRBV12 repertoire. However, when a corresponding LPL or IEL sample was studied, very few common clones were found and overall the repertoire of the patient was polyclonal. The TCR repertoire did not appear to be more restricted with increasing age, at least in the infants studied (aged 0–18 mo). In the oldest infant (18-mo-old, I) no TCRBV12 clonal expansions were found in LPL, and no TCRBV6 clonal expansions were found in IEL and LPL. Only one TCRBV6 sequence was found to be common, accounting for 4 of 24 clones from the mucosa and 1 of 21 clones from the IEL fraction. For comparison, TCRBV6 was studied in fractionated IEL from a child aged almost 5 years and at this age the TCRBV6 repertoire of IEL was more restricted, with 7 different sequences used by 24 clones (29% unique). Interestingly, the sample from the colon that had not been exposed to the fecal stream (E) was indistinguishable from the others; 17 of 21 sequences were different and just three clones appeared more than once (80% unique). Thus, although this child displayed a monoclonal TCRBV4 repertoire in its small intestine mucosa (E, Table V), the large intestinal mucosa displayed a polyclonal TCRBV4 repertoire, consistent with other large intestinal samples, despite being isolated from the normal flow of luminal Ags at birth.

Using TCRBV4, similar results were obtained: 19–29 sequences were obtained per sample and most samples were polyclonal with six or fewer clones found more than once. Some evidence of clonal restriction was found in the TCRBV4 repertoire of two samples. One mucosal sample (K, 12 mo, 38% unique) and one fractionated LPL sample (F, 9 mo, 27% unique) displayed a more restricted TCR repertoire. Of the two patients from which LPL fractions had been isolated, patient F had no sequences shared between the mucosa and the LPL fraction, and patient D had only one sequence common to the mucosal and LPL fractions which was used by only three clones. When the fractionated LPL samples were compared, the LPL from the older child (F, 9 mo) displayed a more restricted TCRBV4 repertoire, 27% of clones were unique, compared with the LPL from patient D (6 mo) in whom 80% were unique. When the sequences obtained from patient D at 6 and 9 mo were compared, only one sequence from the first LPL fraction, used by 1 of 26 clones, was found to be retained in the mucosa at 9 mo, and was slightly expanded accounting for 4 of 28 clones. In fact, the mucosal samples from patient D displayed a slightly more restricted TCRBV4 repertoire at 9 mo than at 6 mo, 50 and 60% unique, respectively.

The frequency with which each joining gene was used was determined from the 764 TCR sequences derived from the samples described above. The most commonly used TCRBJ segment was J2s3, followed by J2s7, J2s1, and J1s1 (Fig. 3). This is similar to that described for Jβ usage in adult PBLs (22).

FIGURE 3.

TCRBJ usage in human infant intestine is similar in T cells using TCRBV4, 6, and 12. The frequency with which each joining gene was used was determined from 764 TCR sequences. The most commonly used TCRBJ segment was J2s3, followed by J2s7, J2s1, and J1s1. This is similar to that reported for human adult PBLs.

FIGURE 3.

TCRBJ usage in human infant intestine is similar in T cells using TCRBV4, 6, and 12. The frequency with which each joining gene was used was determined from 764 TCR sequences. The most commonly used TCRBJ segment was J2s3, followed by J2s7, J2s1, and J1s1. This is similar to that reported for human adult PBLs.

Close modal

The junctional length was determined by counting the amino acids between the first (5′) serine residue (3′ end of the variable region) and the second (3′) phenylalanine residue of the joining segment. The mean junctional length was 13 aa (Table V). There was no trend toward longer junctional regions with increasing age.

This study provides the first evidence that, in the human infant, differentiation of epithelial and mucosal T cells takes place in the intestine over the first 12–18 mo after birth, giving rise to a polyclonal population of T cells.

We have demonstrated significant numbers of T cell precursors in the infant intestinal mucosa. Soon after birth, CD48 (DN) cells accounted for up to 99% of IELs and up to 42% of LPLs. Most of the DN cells expressed CD3 at a moderate or low level, although a CD3 population was also identified which are likely to be even more immature. CD3+ DN cells are likely to be either immature T cells or CD3+ NK cells, while DP cells could represent a slightly later stage of immature T cell development (23). However, some DP cells, proposed to be activated T cells, are normally present in adults and are increased in adults with various autoimmune disorders (24, 25). We have observed that a proportion of the DP cells were CD25 or CD25low (data not shown), in contrast to in vitro studies which have demonstrated that activated DP cells display an unusually high level of CD25 (26). Thus, it is more likely that the cells we have observed are immature DP cells than activated circulating DP cells. To confirm the likely T cell lineage of CD3low/− cells, expression of CD2, CD5, and CD7 was studied. CD2 is expressed on a large proportion of developing T cells, before CD3 expression, and by NK cells. CD5 is expressed by T and B cells and their bone marrow-derived precursors. CD7 is a marker of early lineage T cells, expressed before CD3 on nonmyeloid and nonerythroid bone marrow emigrant cells. CD7 may be coexpressed with CD5 before CD5 down-regulation occurs, and can also be expressed by NK cells. Flow cytometry studies revealed that CD7+23 cells accounted for up to 18% of all IEL and 10% of LPL. In contrast, CD2+ CD5+ cells represented up to 6% of the CD3 LPL fraction. This combination of cell surface molecules strongly suggests that the cells are immature T cells, supported by immunohistochemistry showing coexpression of CD5 and CD7, a proportion of which were negative for NK, B, and mature T cell markers.

Howie et al. (10) reported CD7+ cells in the human fetal intestinal mucosa. Many were CD2, especially in the lamina propria (14%, ±7%). They also found a significant proportion (30%) of fetal LPL were CD3+CD4CD8. As in our study, they found that, in the lamina propria, CD4+ T cells were more common than CD8+ T cells. They did not, however, report CD4+ IEL, although we found CD4+ cells in the epithelium by flow cytometric and immunohistolochemical analyses. Together, these data suggest that DN T cell precursors accumulate in the mucosa during fetal life and do not differentiate further until after birth, whereupon they gradually mature in situ to SP cells and are joined in the mucosa by thymic emigrants. Evidence of pre-T cells in other tissues of human neonates is limited. Significant numbers of CD34+ hemopoietic stem cells are present in both cord and peripheral blood at birth, but decline rapidly with age (27, 28, 29). Hemopoietic cells are present in neonatal liver (30), but evidence for hemopoiesis in the adult human liver is controversial (31, 32).

Immature T cells are likely to be undergoing TCR rearrangement. The RAG protein binds to a pair of recombination signal sequences that flank the V, D, and J gene segments of the germline and initiates a double-strand break that facilitates recombination (33). In mice, intestinal RAG expression has been recognized for a decade (1, 4), although it is usually ascribed to CD8αα cells. RAG expression has been reported in human intestine (12, 34, 35), specifically by immature CD3CD2+TCR IEL and CD3+TCR IEL (34) in the jejunum, and by immature CD3CD2+CD7+ IEL and LPL in the small intestines of adults and children (35). It was recently reported that T cells in the human small intestine and bone marrow express two specific splice variants of the RAG-1 gene (1A/2 and 1A/1B/2) not found in thymocytes (35). It was proposed that T cells using the RAG-1 1A/2 and 1A/1B/2 variants migrate directly from the bone marrow to the small intestine, where they use the 1A/2- and 1A/1B/2-specific RAG-1 variants during TCR gene recombination in situ (35). As our RAG-1 primers are located in exon 2, we would detect all isoforms of RAG-1, and so our finding that 15 of 19 intestinal samples express RAG-1 could correspond to the expression of a gut-specific RAG isoform and certainly merits investigation. After expression of a pre-TCR, consisting of an inframe TCR β-chain and pre-Tα, RAG expression is down-regulated and this is accompanied by a proliferative phase at the end of which CD4 and CD8 are both expressed (DP). After this proliferative phase, RAG is re-expressed and the TCR-α gene locus is rearranged. Thus, RAG may be present in both DN and DP T cells. Furthermore, we have demonstrated CD3+ DN and DP cells in the human postnatal intestine.

The TEA promoter is located immediately upstream of the TCR-Jα cluster within the TCRαδ locus (36, 37). It is expressed in immature T cells at the point when DN thymocytes become immature SPs (38), and has been reported in human fetal intestine, but has not previously been investigated postnatally (9, 10). TdT is also present in immature lymphocytes, although the precise stage of T cell maturation has not been defined. Expression of TdT was absent from the intestine of human adults (15) but has not been previously investigated in neonates. Together RAG, TdT, and TEA are associated with ongoing TCR rearrangement and junctional modification. Our finding of these three genes in human intestinal mucosa during infancy strongly supports the hypothesis that T cells are undergoing maturation in the gut mucosa after birth.

We have shown that in early childhood, the TCRαβ T cell repertoire in the small and large intestine is essentially polyclonal. Mucosal, LPL, and IEL cDNA was used as the template for PCR. When TCRBV6 and TCRB12 were studied, very few clones were identified more than once. The population was so polyclonal that, within each individual, few clones were shared between the mucosal sample and either the IEL or LPL fractions. This confirms findings reported by Holtmeier et al. (20), that humans displayed a polyclonal γδ repertoire in midgestation which was retained after delivery, but became oligoclonal during childhood. This is in stark contrast to the findings in adults (18, 19, 39), which have consistently been that the adult intestinal TCR repertoire, at least of IEL, is oligoclonal. In our study only a few samples differed from a polyclonal TCR repertoire. One sample displayed an oligoclonal TCRBV6 repertoire at almost 5 years old, thus being more like the adult repertoire than the other infant samples. By comparing mucosal samples from one patient at 6 and 9.3 mo old, and comparing fractionated LPL in two patients aged 6 and 9 mo old, some restriction of the TCRBV4 repertoire with increasing age seemed to be evident, although in the older mucosal samples from patients aged 12 mo (38% unique) and 18 mo (73% unique) this pattern of increasingly restricted TCRBV4 repertoire was variable.

Together with the T cell phenotype data, then, our findings suggest that, at birth, the intestine is seeded with a polyclonal population of immature T cells. Although relatively few cells were found in the intestine, most of the sequenced TCRB transcripts were different, which implies that they have not undergone clonal expansion. Our data hold true for mucosal samples as well as LPL and IEL fractions, indicating that our findings are not due to experimental artifact. In fact, in experiments not reported here, we have found that mucosal samples of adult intestine show an oligoclonal pattern, which has been widely reported (18, 19, 20). It is possible that either the infant polyclonality is diluted by clonal deletion in situ or migration of clones to other tissues or, more likely, that as the commensal flora stabilizes in later infancy, a few flora-associated clones are favored and expand to become dominant. Changes in chemokine receptor expression during infancy may also contribute to tissue-specific redistribution of clones. In mice, CCL25 is up-regulated upon Ag activation and demonstrates a very restricted pattern of expression essentially limited to the intestine and thymus (40, 41, 42, 43, 44).

TCR junctional length has been the topic of some debate. Moss et al. (45) reported that the length of the junctional region was no longer in adult peripheral blood than in fetal cord blood. However, Hall et al. (46) reported that the proportion of junctional regions expressing 10 or more amino acids was significantly lower in cord blood than in adults. It has been suggested that with greater exposure to Ags, TdT is used to add more nucleotides to increase diversity. Our data suggest that the mean junctional length does not change significantly between birth and 1 year of age. Although TdT transcripts are present and protein is expressed, there is no evidence to support the hypothesis that later rearranged junctional segments have a greater length in response to Ag exposure over time.

The phenotype and repertoire of T cells in intestinal mucosa differs between infancy and adulthood. We have provided evidence for a population of T cells in the infant gut which is immature and undergoing receptor gene recombination amid a polyclonal T cell population. This is taking place at a time of great quantitative and qualitative change in Ag exposure, food and microbial, and it seems likely that this environmental influence must, in some way, shape the repertoire. We were surprised, therefore, that in the two individuals tested in which the intestine had been isolated from environmental influence, there was no difference in clonality from tissues exposed in the normal way. It is possible, then, that these early waves of TCR rearrangement are influenced by absorbed Ags recirculating from other areas of the intestine as well as by locally absorbed luminal Ag. The critical events which cause clonal deletion or selective expansion in late infancy are probably more influenced by microflora and diet, and it is probably these influences which shape the oligoclonal repertoire which is crucial to preventing hypersensitivity and inflammatory responses in the adult gut.

We thank Eleri Cusick and Lucinda Huskisson of the Department of Pediatric Surgery for their valued assistance in collecting suitable surgical specimens; and Daniel Carroll and Daniela Vieten for their assistance in obtaining informed parental consent.

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

This work was funded by the Biotechnology and Biological Sciences Research Council (BBSRC) U.K., Grant No. 7/D13513.

3

Abbreviations used in this paper: IEL, intraepithelial lymphocyte; TEA, T early α; DN, double negative; LPL, lamina propria lymphocyte; TCRBV, TCR β-chain V region; DP, double positive. SP, single positive.

1
Guy-Grand, D., N. Cerf-Bensussan, B. Malissen, M. Malassis-Seris, C. Briottet, P. Vassalli.
1991
. Two gut intraepithelial CD8+ lymphocyte populations with different T-cell receptors–a role for the gut epithelium in T-cell differentiation.
J. Exp. Med.
173
:
471
.
2
Poussier, P., P. Edouard, C. Lee, M. Binnie, M. Julius.
1992
. Thymus-independent development and negative selection of T-cells expressing T-cell receptor-α/β in the intestinal epithelium: evidence for distinct circulation patterns of gut-derived and thymus-derived T-lymphocytes.
J. Exp. Med.
176
:
187
.
3
Rocha, B., P. Vassalli, D. Guy-Grand.
1991
. The V-β repertoire of mouse gut homodimeric-α CD8+ intraepithelial T-cell receptor α/β+ lymphocytes reveals a major extrathymic pathway of T-cell differentiation.
J. Exp. Med.
173
:
483
.
4
Guy-Grand, D., C. V. Broecke, C. Briottet, M. Malassisseris, F. Selz, P. Vassalli.
1992
. Different expression of the recombination activity gene RAG-1 in various populations of thymocytes, peripheral T-cells and gut thymus-independent intraepithelial lymphocytes suggests two pathways of T-cell receptor rearrangement.
Eur. J. Immunol.
22
:
505
.
5
Lin, T., G. Matsuzaki, H. Yoshida, N. Kobayashi, H. Kenai, K. Omoto, K. Nomoto.
1994
. CD3CD8+ intestinal intraepithelial lymphocytes (IEL) and the extrathymic development of IEL.
Eur. J. Immunol.
24
:
1080
.
6
Boll, G., A. Rudolphi, S. Spiess, J. Reimann.
1995
. Regional specialization of intraepithelial T-cells in the murine small and large-intestine.
Scand. J. Immunol.
41
:
103
.
7
Murray, A. M., B. Simm, K. W. Beagley.
1998
. Cytokine gene expression in murine fetal intestine: potential for extrathymic T cell development.
Cytokine
10
:
337
.
8
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.
197
:
333
.
9
Koningsberger, J. C., A. Chott, T. Logtenberg, L. Wiegman, R. S. Blumberg, G. P. V. 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
.
10
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
.
11
Latthe, M., L. Terry, T. T. MacDonald.
1994
. High frequency of CD8-α-α homodimer-bearing T-cells in human fetal intestine.
Eur. J. Immunol.
24
:
1703
.
12
Lynch, S., D. Kelleher, R. McManus, C. O’Farrelly.
1995
. RAG1 and RAG2 expression in human intestinal epithelium: evidence of extrathymic T-cell differentiation.
Eur. J. Immunol.
25
:
1143
.
13
Hayakawa, S., S. Saito, N. Nemoto, F. Chishima, K. Akiyama, H. Shiraishi, J. Hayakawa, M. Karasakisuzuki, K. T. Fujii, M. Ichijo, et al
1994
. Expression of recombinase-activating genes (RAG-1 and RAG-2) in human decidual mononuclear-cells.
J. Immunol.
153
:
4934
.
14
Chun, J. J. M., D. G. Schatz, M. A. Oettinger, R. Jaenisch, D. Baltimore.
1991
. The recombination activating gene-1 (RAG-1) transcript is present in the murine central nervous system.
Cell
64
:
189
.
15
Taplin, M. E., M. E. Frantz, C. Canning, J. Ritz, R. S. Blumberg, S. P. Balk.
1996
. Evidence against T-cell development in the adult human intestinal mucosa based upon lack of terminal deoxynucleotidyltransferase expression.
Immunology
87
:
402
.
16
DeVillartay, J. P., D. Lewis, R. Hockett, T. A. Waldmann, S. J. Korsmeyer, D. I. Cohen.
1987
. Deletional rearrangement in the human T-cell receptor α-chain locus.
Proc. Natl. Acad. Sci. USA
84
:
8608
.
17
Mauvieux, L., I. Villey, J. P. DeVillartay.
2001
. T early α (TEA) regulates initial TCRVAJA rearrangements and leads to TCRJA coincidence.
Eur. J. Immunol.
31
:
2080
.
18
Balk, S. P., E. C. Ebert, R. L. Blumenthal, F. V. McDermott, K. W. Wucherpfennig, S. B. Landau, R. S. Blumberg.
1991
. Oligoclonal expansion and CD1 recognition by human intestinal intraepithelial lymphocytes.
Science
253
:
1411
.
19
Blumberg, R. S., C. E. Yockey, G. G. Gross, E. C. Ebert, S. P. Balk.
1993
. Human intestinal intraepithelial lymphocytes are derived from a limited number of T-cell clones that utilize multiple V-β T-cell receptor genes.
J. Immunol.
150
:
5144
.
20
Holtmeier, W., T. Witthoft, A. Hennemann, H. S. Winter, M. F. Kagnoff.
1997
. The TCR-δ repertoire in human intestine undergoes characteristic changes during fetal to adult development.
J. Immunol.
158
:
5632
.
21
Chott, A., C. S. Probert, G. G. Gross, R. S. Blumberg, S. P. Balk.
1996
. A common TCR β-chain expressed by CD8+ intestinal-mucosa T-cells in ulcerative colitis.
J. Immunol.
156
:
3024
.
22
Rosenberg, W. M. C., P. A. H. Moss, J. I. Bell.
1992
. Variation in human T-cell receptor V-β and J-β repertoire: analysis using anchor polymerase chain-reaction.
Eur. J. Immunol.
22
:
541
.
23
Carleton, M., M. C. Haks, S. A. A. Smeele, A. Jones, S. M. Belkowski, M. A. Berger, P. Linsley, A. M. Kruisbeek, D. L. Wiest.
2002
. Early growth response transcription factors are required for development of CD4CD8 thymocytes to the CD4+CD8+ stage.
J. Immunol.
168
:
1649
.
24
Weiss, L., A. Roux, S. Garcia, C. Demouchy, N. Haeffner-Cavaillon, M. D. Kazatchkine, M. L. Gougeon.
1998
. Persistent expansion, in a human immunodeficiency virus-infected person, of Vβ-restricted CD4+CD8+ T lymphocytes that express cytotoxicity-associated molecules and are committed to produce interferon-γ and tumor necrosis factor-α.
J. Infect. Dis.
178
:
1158
.
25
Parel, Y., C. Chizzolini.
2004
. CD4+CD8+ double positive (DP) T cells in health and disease.
Autoimmun. Rev.
3
:
215
.
26
Sullivan, Y. B., A. L. Landay, J. A. Zack, S. G. Kitchen, L. Al-Harthi.
2001
. Upregulation of CD4 on CD8+ T cells: CD4dimCD8bright T cells constitute an activated phenotype of CD8+ T cells.
Immunology
103
:
270
.
27
Li, K., J. Liu, T. F. Fok, F. W. Yau, A. Wong, C. K. Li, M. Yang, K. W. So, K. W. Chik, K. S. Tsang, et al
1999
. Human neonatal blood: stem cell content, kinetics of CD34+ cell decline and ex vivo expansion capacity.
Br. J. Haematol.
104
:
178
.
28
Li, K., F. W. Yau, T. F. Fok, K. W. So, C. K. Li, P. M. P. Yuen.
2001
. Haematopoietic stem and progenitor cells in human term and preterm neonatal blood.
Vox Sang.
80
:
162
.
29
De Vries, E., S. De Bruin-Versteeg, W. M. Comans-Bitter, R. De Groot, W. C. J. Hop, G. J. M. Boerma, F. K. Lotgering, J. J. M. van Dongen.
2000
. Longitudinal survey of lymphocyte subpopulations in the first year of life.
Pediatr. Res.
47
:
528
.
30
Brauner, P., M. Nibbelink, P. Flachs, I. Vitkova, P. Kopecky, I. Mertelikova, L. Janderova, L. Penicaud, L. Casteilla, R. Plavka, J. Kopecky.
2001
. Fast decline of hematopoiesis and uncoupling protein 2 content in human liver after birth: location of the protein in Kupffer cells.
Pediatr. Res.
49
:
440
.
31
Golden-Mason, L., A. M. Kelly, O. Traynor, G. McEntee, J. Kelly, J. E. Hegarty, C. O’Farrelly.
2001
. Expression of interleukin 7 (IL-7) mRNA and protein in the normal adult human liver: implications for extrathymic T cell development.
Cytokine
14
:
143
.
32
Golden-Mason, L., D. C. Douek, R. A. Koup, J. Kelly, J. E. Hegarty, C. O’Farrelly.
2004
. Adult human liver contains CD8pos T cells with naive phenotype, but is not a site for conventional αβ T cell development.
J. Immunol.
172
:
5980
.
33
Khor, B., B. P. Sleckman.
2002
. Allelic exclusion at the TCR β locus.
Curr. Opin. Immunol.
14
:
230
.
34
Lundqvist, C., V. Baranov, S. Hammarstrom, L. Athlin, M. L. Hammarstrom.
1995
. Intraepithelial lymphocytes: evidence for regional specialization and extrathymic T-cell maturation in the human gut epithelium.
Int. Immunol.
7
:
1473
.
35
Bas, A., S. G. Hammarstrom, M. L. K. C. Hammarstrom.
2003
. Extrathymic TCR gene rearrangement in human small intestine: identification of new splice forms of recombination activating gene-1 mRNA with selective tissue expression.
J. Immunol.
171
:
3359
.
36
DeVillartay, J. P., R. D. Hockett, D. Coran, S. J. Korsmeyer, D. I. Cohen.
1988
. Deletion of the human T-cell receptor δ-gene by a site-specific recombination.
Nature
335
:
170
.
37
DeChasseval, R., J. P. DeVillartay.
1993
. Functional characterization of the promoter for the human germ line T-cell receptor J-α (TEA) transcript.
Eur. J. Immunol.
23
:
1294
.
38
Wilson, A., J. P. DeVillartay, H. R. MacDonald.
1996
. T cell receptor δ gene rearrangement and T early α (TEA) expression in immature αβ lineage thymocytes: implications for αβ/γ δ lineage commitment.
Immunity
4
:
37
.
39
Gross, G. G., V. L. Schwartz, C. Stevens, E. C. Ebert, R. S. Blumberg, S. P. Balk.
1994
. Distribution of dominant T-cell receptor-β chains in human intestinal-mucosa.
J. Exp. Med.
180
:
1337
.
40
Kunkel, E. J., J. J. Campbell, G. Haraldsen, J. L. Pan, J. Boisvert, A. I. Roberts, E. C. Ebert, M. A. Vierra, S. B. Goodman, M. C. Genovese, et al
2000
. Lymphocyte CC chemokine receptor 9 and epithelial thymus-expressed chemokine (TECK) expression distinguish the small intestinal immune compartment: epithelial expression of tissue-specific chemokines as an organizing principle in regional immunity.
J. Exp. Med.
192
:
761
.
41
Onai, N., M. Kitabatake, Y. Y. Zhang, H. Ishikawa, S. Ishikawa, K. Matsushima.
2002
. Pivotal role of CCL25 (TECK)-CCR9 in the formation of gut cryptopatches and consequent appearance of intestinal intraepithelial T lymphocytes.
Int. Immunol.
14
:
687
.
42
Wurbel, M. A., J. M. Philippe, C. Nguyen, G. Victorero, T. Freeman, P. Wooding, A. Miazek, M. G. Mattei, M. Malissen, B. R. Jordan, et al
2000
. The chemokine TECK is expressed by thymic and intestinal epithelial cells and attracts double- and single-positive thymocytes expressing the TECK receptor CCR9.
Eur. J. Immunol.
30
:
262
.
43
Papadakis, K. A., J. Prehn, V. Nelson, L. Cheng, S. W. Binder, P. D. Ponath, D. P. Andrew, S. R. Targan.
2000
. The role of thymus-expressed chemokine and its receptor CCR9 on lymphocytes in the regional specialization of the mucosal immune system.
J. Immunol.
165
:
5069
.
44
Zabel, B. A., W. W. Agace, J. J. Campbell, H. M. Heath, D. Parent, A. I. Roberts, E. C. Ebert, N. Kassam, S. X. Qin, M. Zovko, et al
1999
. Human G protein-coupled receptor GPR-9-G/CC chemokine receptor 9 is selectively expressed on intestinal homing T lymphocytes, mucosal lymphocytes, and thymocytes and is required for thymus-expressed chemokine-mediated chemotaxis.
J. Exp. Med.
190
:
1241
.
45
Moss, P. A. H., J. I. Bell.
1995
. Sequence analysis of the human α-β-T-cell receptor CDR3 region.
Immunogenetics
42
:
10
.
46
Hall, M. A., J. L. Reid, J. S. Lanchbury.
1998
. The distribution of human TCR junctional region lengths shifts with age in both CD4 and CD8 T cells.
Int. Immunol.
10
:
1407
.