HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Podd, B. S. Articles by Camerini, V. Search for Related Content PubMed PubMed Citation Articles by Podd, B. S. Articles by Camerini, V.

T Cells in Cryptopatch Aggregates Share TCR Variable Region Junctional Sequences with T Cells in the Small Intestinal Epithelium of Mice¹

Bradley S. Podd^2,*,, Joseph Thoits^2,, Nicholas Whitley, Hao-Yuan Cheng, Kimberly L. Kudla^*, Hiroko Taniguchi, Joanna Halkias, Kerstin Goth and Victoria Camerini^3,

^* Department of Pediatrics, University of Virginia Health Sciences Center, Charlottesville, VA 22908; Department of Pediatrics and Department of Surgery, Childrens Hospital Los Angeles, and Saban Research Institute, University of Southern California, Los Angeles, CA 90027; and Department of Pediatrics and the Center for Immunology, University of California, Irvine, CA 92697

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
The role of cryptopatch aggregates in the development of intestinal intraepithelial lymphocytes (IEL) is a matter of controversy. Therefore, an important question is whether T cells in cryptopatch aggregates are lineally related to IEL. We hypothesized that if ⁺ IEL derive from T cells in cryptopatch aggregates, then a clonal relationship would exist between the two populations. To test this hypothesis, we compared the sequence of rearranged TCR gamma variable region 5 genes in ⁺ IEL and cryptopatch cells. We purified IEL by FACS and cryptopatch cells were isolated from frozen sections of the intestine by laser-assisted microdissection. PCR showed that TCR gamma variable region 5 was rearranged in ⁺ IEL and in CD3⁺ cryptopatch cells, but not in CD3^– cryptopatch cells. DNA sequence analysis showed that the frequency of in-frame junctions in cryptopatch aggregates was at a level consistent with positive selection in both wild-type and athymic nude mice. In addition, the predicted amino acid sequences of V-J junctions present in ⁺ IEL and cryptopatch cells were encoded by identical nucleotide sequences. By contrast, the frequency of in-frame joints was significantly reduced in cryptopatch cells isolated from TCR -deficient mice, indicating that the enrichment of in-frame joints in cryptopatch cells must normally depend on expression of surface TCR. Our results are consistent with the hypothesis that a subset of ⁺ IEL are related to T cells in cryptopatch aggregates. The precise role of cryptopatch aggregates in intestinal ⁺ T cell homeostasis still needs to be determined.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Most T cells developing in the fetal thymus express canonical TCR that give rise to dendritic epidermal T cells (DETC)⁴ in the skin (1, 2, 3, 4). By contrast, most T cells developing in the adult thymus express highly diverse TCR that are distributed widely in the body, whereas few cells express TCR. Despite the paucity of ⁺ T cells in the adult thymus and in nonmucosal sites in the periphery, ⁺ T cells are enriched in the intestinal epithelium and in nearly every mucosal/epithelial location in the adult mouse seemingly distributed by their TCR (TCR -chain V region; TCRGV) and variable (TCR -chain V region; TCRDV) gene usage (5, 6, 7, 8). For example, a large fraction of TCR ⁺ intestinal intraepithelial lymphocytes (IEL) express TCRGV gene 5 (TCRGV5), often in conjunction with TCRDV4 in C57BL/6 mice (nomenclature of Garman et al. (2)). DETC express TCRGV3, and the few ⁺ T cells in nonmucosal/epithelial sites in the periphery of mice use TCRGV1.1, TCRGV1.2, and TCRGV2 in combination with a variety of TCRDV gene products (9). The mechanisms responsible for the tissue localization of ⁺ T cells remain undefined (5, 9, 10). Studies in TCR transgenic mice demonstrate that normally nonmucosal TCRGV gene products do not restrict the development of TCR ⁺ IEL, suggesting that tissue-specific Ags may not play a major role in the pattern of ⁺ TCR gene usage at least in the intestine (11, 12, 13). Studies of DETC cells, however, suggest that the expression of "skin-seeking" receptors during development in the thymus may dictate patterns of homing that are unique to these T cells (10, 12). The anatomic location and time point at which precursors of TCR ⁺ IEL acquire "gut-seeking" signals are not known.

The extent that T cell precursors of IEL develop in the thymus remains controversial (14, 15, 16). In addition, the anatomic site where extrathymic development of T cells would occur in the intestine, and whether these events lead to the generation of TCR⁺ IEL in mice is unsettled (17, 18, 19, 20, 21, 22). This is particularly enigmatic in light of the centralized and highly regulated development of T cells known to occur in the thymus (23). The identification of cell clusters, called cryptopatch aggregates, distributed throughout the intestinal lamina propria, offered one potential site where a nonthymic pathway of T cell development would be centralized in the intestine (24, 25, 26, 27). Cryptopatch aggregates, unlike Peyer’s patches and intestinal lymphoid follicles (ILF), consist of CD117^–, lineage negative (lin^–), and CD117⁺, IL-7R⁺, lin^– cells, with few TCR ⁺ and TCR ⁺ cells, and a larger population of CD11c⁺ dendritic cells (25). Transfer of purified CD117⁺, IL-7R⁺, lin^– cells, but not the CD117^–, lin^– counterpart, gave rise to TCR ⁺ and TCR ⁺ IEL in SCID mice, supporting the fact that CD117⁺, IL-7R⁺, lin^– cells include precursors of IEL (26). Kinetic and ontogenic studies have also linked cells in cryptopatch aggregates with TCR ⁺ IEL (and TCR ⁺ IEL). Cryptopatch aggregates are established after birth in mice, but before the expansion of TCR ⁺ IEL later in postnatal life (24, 27, 28). A similar kinetic relationship was observed in adoptive transfer studies using cryptopatch- (26, 27) and bone marrow-derived cells (25, 29). In each case, restoration of cryptopatch aggregates preceded reconstitution of IEL. The absence, or the marginal level of mRNA and protein for RAG-2 and mRNA for other genes required for T cell lineage progression has cast doubt on whether cryptopatch aggregates are primary lymphoid organs (30, 31, 32). Moreover, recent studies demonstrated that mice lacking cryptopatch aggregates were replete with IEL subsets, at least at the population level, suggesting that cryptopatch aggregates are not obligate for IEL development (33, 34). Hence, the extent to which cryptopatch aggregates are primary lymphoid organs and give rise to T cell precursors of IEL remains controversial.

In this study, we explored the possibility that ⁺ T cells in cryptopatch aggregates are lineally related to TCR ⁺ IEL. TCRGV5 was preferentially rearranged in cryptopatch aggregates, similar to TCR ⁺ IEL. Furthermore, TCRGV5 rearrangements were only detected in CD3⁺ cryptopatch cells but not in CD3^– cryptopatch cells. Moreover, the frequency of in-frame TCRGV5 segments in cryptopatch cells was greater than would be expected to occur by chance alone and was reduced in mice unable to express a surface TCR. Despite evidence that cryptopatch aggregates function as primary lymphoid organs, we did not detect mRNA for RAG-1 gene expression in cryptopatch aggregates. Finally, the same nucleotide sequence accounted for the majority of in-frame TCRGV5 exons in cryptopatch cells and was shared with a subset of TCR ⁺ IEL. Our data support the hypothesis that T cells in cryptopatch aggregates are clonally related to TCR ⁺ IEL. Perhaps cryptopatch aggregates are reservoirs for Ag-selected ⁺ T cells in adult mice. This may, in turn, promote the restricted TCR ⁺ repertoire characteristic of IEL.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Mice

Mice between 6 and 12 wk of age were obtained from Taconic Farms and The Jackson Laboratory and housed in a laminar flow barrier facility under specific pathogen-free conditions. Male and female C57BL/6 mice (C57BL/6NTac), TCR -chain-deficient mice (TCR ^–/–) (B6.129P2-Tcrdtm1Mom) (35), and C57BL/6 nude (C57BL/6NTac-Foxn1nu) mice were used for these studies. The Institutional Animal Care and Use Committee approved all animal protocols and procedures.

Preparation of lymphocyte populations

Intestinal lymphocytes were prepared from the small intestine of mice using our previously published procedure (36). Briefly, the small intestine was dissected from its mesentery and washed in RPMI 1640 (Invitrogen Life Technologies). The intestine was opened longitudinally, and the contents were removed. The intestine was cut into 0.5-cm pieces, and mononuclear cells were released from the epithelium by shaking in calcium and magnesium-free HBSS (Invitrogen Life Technologies) supplemented with 1 mM DTT (Sigma-Aldrich). Mononuclear cells collected from the epithelial layer were filtered through stainless steel mesh, and IEL were enriched on a discontinuous 20%/40%/70% Percoll (Amersham, Biosciences) gradient at 900 x g for 20 min. IEL were collected at the 40%/70% interface, washed with RPMI 1640 with 10% FCS, and kept on ice until use, as noted below. Epithelial cells, enriched with mononuclear cells at the 20%/40% interface, were washed as described above and stored on ice until use, as noted below. Mononuclear cells were released from the thymus and spleen by mechanical disruption of the capsule between frosted glass slides. Hypotonic lysis was used to deplete RBC from splenocytes. Mononuclear cell populations were collected by centrifugation and suspended in RPMI 1640 with 10% FCS until use, as noted below.

Ab staining and cell sorting

Lymphocyte populations were prepared for cell sorting by suspension in mAb staining buffer (PBS with 5% FCS (Invitrogen Life Technologies), 0.02% NaN₃ (Sigma-Aldrich)) supplemented with 5.0 µg/ml mAb CD16/CD32 (2.4G2) (BD Biosciences) at a concentration of 1 x 10⁶ cells/ml. Pretitered mAb directly conjugated to FITC, PE, biotin, and allophycocyanin were added to cell suspensions and incubated at 4°C for 30 min. Cells were then washed in PBS and suspended in staining buffer with an optimal concentration of streptavidin PE-Cy7 or streptavidin-PerCP (Caltag Laboratories) to detect biotin-conjugated primary mAb. Cells were incubated at 4°C for 30 min and then washed three times with PBS before cell sorting. The following mAbs were used for cell surface staining before cell sorting (see below): TCR (H57 597), TCR (GL3), CD3 (145-2C11), and pan CD45 (30-F11) (all purchased from BD Biosciences or Biolegend).

After mAb staining, intestinal mononuclear cells were suspended in RPMI 1640 with 10% FCS on ice. Lymphocytes were defined by forward and side scatter characteristics, and CD45⁺, CD3⁺, TCR ⁺, or TCR ⁺ IEL were distinguished from each other and from CD45⁺, CD3^–, and CD45⁺, TCR ^– or TCR ^– IEL subsets by gating with a half-log window between cell populations of interest on a FACSVantage Cell Sorter (BD Immunocytometry Systems) or a MoFlo high-performance cell sorter (DakoCytomation). TCR ⁺ cells were isolated from suspensions of splenocytes using the same protocol. Sorted cell populations were 98–100% pure upon reanalysis. Sorted cells were collected by centrifugation at 1200 rpm for 5 min, and pellets were suspended in DNA or RNA lysis buffer and stored at –20°C until use. DNA was purified using the DNeasy tissue kit, and RNA was purified using the RNeasy kit following the included directions (Qiagen).

Preparation of the small intestine for embedding, tissue sectioning for immunocytochemical studies, and laser microdissection

The small intestine was excised en bloc, opened longitudinally, cleaned, and placed in PBS (pH 5.5; 4°C). Five- to 6-cm lengths were immersed immediately in optimal cutting temperature compound (Miles) and frozen on dry ice. Frozen tissue blocks were cut on a cryostat microtome, and 6- to 7-µm sections were placed on coated glass slides or on Leica membrane slides (Leica Microsystems) pretreated with bonding agent (VECTABOND; Vector Laboratories) for microdissection. Tissue sections were fixed in acetone and stored at –80°C until use. Before staining, tissues sections were returned to room temperature and incubated for 20 min in PBS supplemented with 20% normal serum (Vector Laboratories) followed by incubation with biotinylated or fluorescence-conjugated mAb specific for CD45, TCR (GL-4 and GL-3), TCR (H57-597), CD3 (2C11), CD69 (H1.2F3), or CD117 (2B8) (all purchased from BD Biosciences or Biolegend) in PBS supplemented with 5% FCS. If the primary mAb was not directly conjugated to a fluorescence label, tissue sections were incubated with biotinylated goat anti-hamster IgG or biotinylated rabbit anti-rat IgG (as required) followed by incubation with avidin-HRP (ABC; avidin-biotin complex) (all obtained from Vector Laboratories). Slides examined by fluorescence microscopy were counterstained with 4',6'-diamidino-2-phenylindole (DAPI) (Molecular Probes) and mounted in Prolong Gold Anti-fade Mounting Medium (Molecular Probes). Slides examined by light microscopy were incubated with 3' 3'diaminobenzidine (DAB) or 3-amino-9-ethylcarbazole (AEC) (Vector Laboratories), and sections were counterstained by hematoxylin blue (Biomeda). Control sections were prepared following incubation with the appropriate isotype control mAb or following addition of biotinylated mAb conjugate alone. Slides were examined using a Leica DMLA microscope (Leica Microsystems) equipped with the appropriate filter sets, and images were captured using an attached SPOT camera. Positively staining cells were identified under light or fluorescence microscopy.

In vivo labeling of lymphocytes subsets by BrdU

Mice were injected i.p. daily for three consecutive days with 100 µl of a 10 mg/ml solution of BrdU (Sigma-Aldrich) in PBS. On the fourth day, the small intestine was harvested in the usual fashion for histological analysis (see above). BrdU FITC-conjugated mAb (3D4; Caltag Laboratories or BD Biosciences) was used for detection of incorporated BrdU. For immunofluorescent histology, 5-µm serial sections of intestine and thymus were cut and pretreated according to a modified protocol (Vector Laboratories). Briefly, intestinal sections were rinsed in tap water for 5 min, preheated in distilled water at 37°C for 5 min, and finally incubated in 2 N HCl at 37°C. In cases of dual labeling for surface CD3 expression, sections were incubated in 2 N HCl for 5 min, or when analyzing BrdU incorporation alone, sections were incubated for 30 min. Slides were rinsed in tap water and neutralized in 50 mM TBS (pH 7.4) for 10 min before standard immunofluorescent staining as noted above. Fluorescent images were obtained on a Leica fluorescence microscope (Leica Microsystems) using OpenLab imaging software (Improvision). BrdU⁺ cells were counted per individual cryptopatch aggregate. To normalize differences in the size of individual cryptopatch aggregates, the number of BrdU⁺ cells was corrected for the area of the aggregate. The long and short axes of each cryptopatch aggregate were measured and the area calculated (A = [xy/4]), where x and y are the short and long axes of an ellipse, respectively. The density of BrdU⁺ cells per unit area of cryptopatch was then determined. Adobe Photoshop (Adobe) was used to merge single-color images.

Laser-assisted microdissection (LMD)

Frozen sections cut at 7-µm thickness on a cryostat microtome were placed on uncoated glass slides (Fisher Scientific) for harvest by the PixCell II (Arcturus Engineering) laser capture system or on Leica Membrane slides for tissue harvest with the Leica AS LMD Microdissection system (Leica Microsystems). In either case, intestinal sections were air dried, fixed, and stored at –80°C until use. Slides were allowed to return to room temperature before further processing by immunostaining, as noted above, or were counterstained with either hematoxylin (Fisher Biochemicals) or Toludine Blue O (Electron Microscopy Sciences). Slides were prepared for LMD using two methods. Tissue sections prepared for LMD using the Leica system were air dried before dissection, whereas tissue sections for harvest using the Arcturus system were washed in distilled water followed by successive washes in 70, 95, and 100% ethanol, and a final dehydration step in xylene (Fisher Biochemicals) before air drying. Cryptopatch aggregates, the epithelium, or distinct cells within the cryptopatch aggregates were identified under light or fluorescence conditions as noted in the figure legends. The laser aperture, attenuation, and cutting speed were adjusted as per the manufacturer’s recommendation and based on tissue requirements. Imaging software captured sequential images before and after laser cutting as well as an automated image capture of the Capsure-HS LMD cap or PCR tube for inspection and documentation of capture. Cells, or groups of cells on Capsure-HS caps or membranes were harvested and placed directly into DNA or RNA lysis buffer. DNA was extracted using the PicoPure DNA kit (Arcturus Engineering), and RNA was extracted using the QIAgen Micro RNeasy kit (Qiagen) as per the manufacturer’s instructions. After DNase treatment, total cellular RNA was used to prepare cDNA using the First-Strand cDNA Synthesis kit (Amersham-Pharmacia) following the manufacturer’s instructions.

Nucleic acid preparation, PCR, cloning, and sequencing

Purified DNA (50–100 ng of template or as noted) was added to a 25- to 50-µl PCR mixture containing 100 ng of each primer in GeneChoice Taq Reaction Buffer (PGC Scientifics) supplemented with 1.5 mM MgCl₂, 2.5 mM dNTP (Amersham Biotech), and 0.5 U of thermostable DNA polymerase (Taq) (PGC Scientifics). DNA was amplified for 30 cycles in a MJ Research Thermal Cycler using the following parameters: 94°C for 1 min, 61.5°C for 2 min, and 72°C for 2 min with a final step at 72°C for 5 min. Genomic DNA was amplified using previously published primer pairs (16): VG1.2, ACATTGGTACCGGCAAAAAAC; VG2, GGGGGGAATTCCCCTCACCCATATTTTCTT; VG3, GCACTGGATCCAACTGAAAGAAG; VG4, CCAAAGAATTCTGTGTAGTTC; VG5, TCCACTGGTACCGATTCCAG; JG_pan, GGGAGCTTACCAGAGGGAATTACTATGAG; p53, 5'- TCACTGCATGGACGATCTGTTGC; and p53, 3'GATGATGGTAAGGATAGGTCGGCG (37). Linear-range amplification conditions were confirmed by multiple PCR of TCRVG5 with varying numbers of cycles, followed by densitometry of products. PCR amplification of cDNA was performed using 1–8 µl of each cDNA reaction mix derived from equivalent amounts of input RNA, in a final volume of 25–50 µl containing the manufacturer’s buffer as noted above and each primer in specified pairs. Amplifications were performed in a MTC 100 or DNA Engine Thermocycler (MJ Research). Amplification conditions were 1 cycle of 2 min at 94°C, 1 min at 61.5°C, 2 min at 72°C followed by a total of 30 cycles, and a remaining extension for 5 min at 72°C. cDNA was amplified in separate reactions using the following: RAG1, 5'-CCAAGCTGCAGACATTCTAGCACTC; RAG1, 3'-CAACATCTGCCTTCACGTCGATCC; villin, 5'-GCTTGAAGTAGCTCCGGAAA; villin, 3'-TCCTGGCTATCCACAAGACC; 5' -actin, 5'ATGGATGACGATATCGCT; and 3' -actin, 5'ATGAGGTAGTCTGTCAGGT.

PCR products were resolved on a 1.8% agarose gel and visualized by UV light following ethidium bromide staining. PCR-amplified TCRGV5 gene segments were purified using the QIAquick PCR Purification Kit as per the manufacturer’s instructions (Qiagen). Amplified gene segments were cloned into the TA cloning kit vector pCR2.1 as per the manufacturer’s instructions (Invitrogen Life Technologies). The sequence of cloned TCRGV genes was determined with the ABI Prism Automated DNA Sequencer using universal primers specific for the plasmid backbone.

	Results

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
Cryptopatch aggregates can be isolated by LMD of thin frozen intestinal sections

Cryptopatch aggregates have been previously isolated for molecular and cell lineage studies from the small intestine using stereomicroscopy and manual dissection under high-power magnification (26, 27, 34). However, this technique requires specialized equipment and training. Purification of cryptopatch cells expressing the stem cell factor receptor (c-kit ligand or CD117), but lacking multilineage-specific markers from lamina propria-derived mononuclear cell suspensions is an alternative method (32), although the anatomic location of cells collected in this way cannot be assured. To circumvent both limitations, we used LMD (38) to isolate cryptopatch aggregates and distinct cell populations within these aggregates from frozen thin sections of the intestine. Cryptopatch aggregates were distinguished from ILF by mAb staining as described previously (24, 34). A large proportion of cells within cryptopatch aggregates expressed CD117 (Fig. 1A), whereas fewer cells expressed CD3 (Fig. 1B). For LMD, cryptopatch aggregates were captured from intestinal tissue sections on glass slides onto Capsure-HS caps (Fig. 2, A–C) or were dissected from slides by membrane-based laser dissection (Fig. 2, D–F) as described in Materials and Methods. Microscopic re-examination of the intestinal sections and visualization of isolated cryptopatch cells after each harvest confirmed the fidelity of the isolation. To confirm that cells in cryptopatch aggregates were isolated free of the intestinal epithelium, a possible source of rearranged TCRGV5 gene segments in resident TCR ⁺ IEL, cryptopatch aggregates and the epithelial layer were harvested from adjacent regions in the intestine and separately examined for the expression of villin, a protein abundantly expressed by epithelial cells throughout the intestine (39) (Fig. 2, D–F). mRNA for villin was detected in RNA prepared from intestinal epithelial cells, but not detected in cryptopatch aggregates despite abundant mRNA for -actin (Fig. 2G). Hence, cryptopatch aggregates can be isolated by LMD free of contamination by the overlying epithelium and its resident T cells.

View larger version (81K): [in this window] [in a new window]   FIGURE 1. Cryptopatch aggregates include CD3+ cells. Immunohistochemical identification of cryptopatch aggregates in the small intestine of mice showing clusters of CD117+ (A) and individual dispersed CD3+ (B) cells in cryptopatch aggregates in the small intestine of mice. Frozen sections of the small intestine were incubated with CD117 (A) and CD3 mAb followed by biotinylated rabbit anti-rat IgG or biotinylated goat anti-hamster IgG, respectively. CD117+ (A) and CD3+ (B) cells were identified after incubation with ABC and visualized with DAB (black). Cryptopatch aggregates consist of a large population of CD117+ cells but contain CD3+ cells. CD3+ cells were also clustered around the perimeter of the aggregate and in the intestinal epithelial layer. Slides were counterstained with hematoxylin and are shown at x500 original magnification. Representative sections are shown following the examination of >10 mice. Control staining using isotype mAb was negative (data not shown).

View larger version (103K): [in this window] [in a new window]   FIGURE 2. Cryptopatch aggregates can be isolated from the small intestine using two laser dissection systems. Segments of the small intestine were fixed in acetone before staining with H&E (A–C) or with hematoxylin (D–F). Representative cryptopatch aggregates were visualized (A), and the cellular aggregate was captured free from surrounding intestinal cell types (B) onto the Capsure-HS cap film support (C) using standard laser settings on a PixCell II instrument (Arcturus Engineering). Alternatively, cryptopatch aggregates were visualized after hematoxylin staining (D) and the aggregate was dissected from the surrounding intestinal tissue (E) using the AS LMD Microdissection system (Leica Microsystems). The ability of LMD to harvest cryptopatch cells with a high degree of purity from the adjacent epithelial cells (F) was confirmed by RT-PCR analysis of villin expression of cDNA derived separately from cryptopatch aggregates and from the epithelium (G). RT-PCR demonstrated no detectable mRNA for villin in cryptopatch aggregates (lane 1), whereas -actin mRNA was abundant (lane 2). By contrast, villin was readily detected in the epithelial layer harvested by LMD (lane 3). Lanes 5 and 6 show villin and -actin expression from mRNA derived from small intestinal epithelial cells prepared freshly from the intestine, as described for the preparation of mucosal lymphocytes.

We used PCR to determine whether TCRGV to VJ gene rearrangements had occurred in cryptopatch cells and how TCRGV gene usage compared with IEL. Although rearranged TCRGV1.2, TCRGV2, and TCRGV3 were detected in TCR ⁺ and TCR ⁺ IEL, rearranged TCRGV5 was enriched in TCR ⁺ IEL relative to TCR ⁺ IEL (Fig. 3, A and B). Cryptopatch aggregates had consistent rearrangement of TCRGV1.2 and TCRGV5 gene segments, although the level of TCRGV1.2 was lower than for TCRGV5 (Fig. 3D). Rearrangement of TCRGV2 and TCRGV3 were inconsistently detected in DNA derived from cryptopatch aggregates, whereas TCRGV4 was never detected (data not shown). By contrast, rearrangement of TCRGV5 was not as abundant in DNA derived from unfractionated thymocytes when compared with other TCRGV rearrangements (Fig. 3C).

View larger version (41K): [in this window] [in a new window]   FIGURE 3. TCRGV5 rearrangements are enriched in TCR + IEL and cells in cryptopatch aggregates. Genomic DNA was extracted from IEL purified by FACS (A–B), from unfractionated thymocytes (C), and from cryptopatch aggregates isolated by LMD (D). V to J rearrangement of the TCRGV gene segment was determined using the TCR VG-JG primer pairs noted. Following amplification, PCR products were resolved on a 1.8% agarose gel stained with ethidium bromide and visualized by exposure to UV light. The results are representative of a minimum of four independent experiments for each cell subset.

View larger version (70K): [in this window] [in a new window]   FIGURE 4. Rearrangement of TCRGV5 is linked to CD3+ cells in cryptopatch aggregates, although no evidence of RAG-1 gene expression was found. Cells expressing CD3 were identified in cryptopatch aggregates in thin frozen sections of the small intestine following incubation with ABC visualized with AEC (red). Tissue sections were counterstained with hematoxylin before LMD and are shown at x500 original magnification. CD3+ and CD3– cryptopatch cell populations were identified in multiple sections (A) and harvested separately by LMD as described in Materials and Methods. Postdissection images documented selective harvest of respective cell populations (B). Genomic DNA was extracted from respective CD3+ and CD3– cryptopatch cell subsets shown in B above and amplified using primers for TCRGV5 and p53 (C). Results are representative of experiments performed on intestines isolated from two individual mice. RT-PCR analysis of cDNA derived from cryptopatch aggregates or thymus was serially diluted and amplified for expression of RAG-1 and -actin (D). RNA samples were reverse transcribed with random oligo priming, and the resultant reverse transcription products were serially diluted and amplified by PCR with appropriate primer pairs as noted in the Materials and Methods. The + is the positive control using thymus cDNA amplified in parallel with cryptopatch cDNA, and – is the negative control amplification with no added template cDNA. Data are representative of four independent experiments.

View larger version (36K): [in this window] [in a new window]   FIGURE 5. Cryptopatch aggregates contain few T cells, a subset of which expresses CD69. T cells in cryptopatch aggregates from C57BL/6 mice (A–F) were identified in frozen sections of the small intestine after staining with mAb CD3 FITC (A) and TCR PE (B). An overlay of single-color images is shown in (C). CD3+ cells are scattered throughout the cryptopatch aggregate, whereas few TCR + cells are present. A CD3+, TCR + cell residing within the border of the cryptopatch is indicated by an arrow. Staining with CD3 FITC (D) and CD69 PE (E) are shown, with an overlay of single-color images in F. A CD3+, CD69+ cell is indicated by an arrow. Frozen sections from a TCR –/– (G–L) mouse were stained with Abs to TCR (G), CD3 (H). An overlay of single-color images is shown with DAPI counterstain in (I). Close examination revealed that each TCR + cell costained with CD3, although CD3 intensity varied. Original magnification, x320. Frozen sections were stained with Abs to CD3 (J) and CD69 (K), and an overlay of single-color images is shown with DAPI (L). Original magnification x320, except for (G–I), enlarged from x200. Data shown are representative of multiple cryptopatch aggregates examined from at least three individual mice.

Selection of in-frame TCRGV5 exons in cryptopatch aggregates is independent of the thymus but requires a functional TCR

The frequency of in-frame TCRGV to J joints has been used to predict lineage commitment and to estimate the probability of cellular selection mediated by surface TCR expression in thymocytes (40). We used a similar assay to compare the frequency of in-frame TCRGV5 genes isolated from cryptopatch cells to TCR ⁺ IEL. We found on average that 78% of V to J joints in TCRGV5 exons derived from TCR ⁺ IEL were in-frame (Table I). By contrast, only 33% of V to J joints in TCRGV1.2 genes isolated from TCR ⁺ IEL were in-frame, indicating that, although this gene rearrangement was detected by PCR, few TCR ⁺ IEL were selected to express this TCRGV chain. To maximize the probability of detecting overlap between V region rearrangements in IEL and cells in cryptopatch aggregates, we focused our analysis on TCRGV5-J exons. We found on average that 53% of joints in TCRGV5 gene segments isolated from cryptopatch cells were in-frame. By contrast, the frequency of in-frame joints in exons cloned from TCR ⁺ IEL was 43%, TCR ⁺ cells isolated from the spleen was 36%, whereas IEL lacking surface CD3 expression had a frequency of in-frame rearrangements of 34%. To determine whether the frequency of in-frame joints varied between different cryptopatch aggregates, we harvested individual aggregates from the proximal, middle, and distal region of the small intestine. The frequency of in-frame rearrangements varied from 20 to 86% in any given region, with an average frequency of 60% (Table II) similar to the frequency of in-frame rearrangements obtained from pooled cryptopatch aggregates (Table I).

View larger version (61K): [in this window] [in a new window]   FIGURE 6. The number of proliferating cells in cryptopatch aggregates does not vary along the length of the intestine. The small intestine was harvested from mice injected with BrdU. The degree of BrdU incorporation in cryptopatch aggregates (A) in the proximal (top panel), middle (middle panel), and distal (lower panel) portion of the small intestine was determined from an overlay of single-color fluorescence images (right panel) after staining with mAb BrdU FITC (left panel) and counterstaining with DAPI (middle panel). Few BrdU+ cells are present in cryptopatch aggregates as compared with extensive BrdU+ staining of intestinal epithelial cells. B, Quantitative analysis of proliferating cells in cryptopatch aggregates. The number of BrdU+ cells per individual cryptopatch aggregate corrected for the surface area of each aggregate was determined as described in Materials and Methods. The graphs show the number of proliferating cells per 1000 µm2 of cryptopatch aggregate in the proximal, middle, and distal third of the small intestine. *, p = NS when the proximal, middle, and distal regions were compared with each other using the Student’s two-tailed t test. The data are representative of images obtained from the analysis of multiple cryptopatches from intestines derived from 3 to 5 individual mice. Original magnification, x320.

View larger version (26K): [in this window] [in a new window]   FIGURE 7. Few T cells are proliferating in cryptopatch aggregates. Proliferating T cells in cryptopatch aggregates (A) were identified after staining frozen sections prepared from mice injected with BrdU after staining with mAb BrdU FITC (a) and CD3 PE (b). An overlay of single-color images counterstained with DAPI is shown in c. Examples of CD3+ T cells in cryptopatch aggregates that costained with BrdU are indicated by arrows. B, Quantitative analysis of proliferating T cells in cryptopatch aggregates. The number of BrdU+, CD3+ cells expressed as a proportion of the total number of CD3+ cells per individual cryptopatch aggregate are shown for the proximal, middle, and distal small intestine. The data are representative of images obtained from the analysis of multiple cryptopatch aggregates from three individual mice. Original magnification, x320.

TCRGV5 CDR3 DNA sequences are shared between TCR ⁺ IEL and cryptopatch cells

If ⁺ T cells in cryptopatch aggregates are lineally related to TCR ⁺ IEL, then CDR3 sequences of rearranged TCRGV5 gene segments would be shared between these populations. Three of 14 different amino acid sequences (21% of the repertoire) predicted from DNA sequences across the CDR3 of TCRGV5 exons isolated from cryptopatch cells were identical with those isolated from TCR ⁺ IEL. In fact, these isolates accounted for 71% (90 of 126) of all in-frame TCRGV5 exons isolated from cryptopatch cells and 66% (19 of 29) of sequences isolated from TCR ⁺ IEL (Table III). DNA sequences across the CDR3 were identified between cryptopatch cells and TCR ⁺ IEL, supporting a clonal relationship (Fig. 8). Isolates from TCR ⁺ IEL also included CDR3 DNA sequences that were not present in cryptopatch aggregates. TCRGV5 exons isolated from IEL and cryptopatch cells had N region additions, although V region deletions were uncommon (Table III, Fig. 8, and data not shown). J-segment nucleotide removal was, however, more common in cryptopatch cells (8 of 14 exons), but not in exons isolated from TCR ⁺ IEL. Taken together, these data support that T cells in cryptopatch aggregates are clonally related to a subset of TCR ⁺ IEL. Furthermore, sequence identity at the protein level with variability in the DNA sequences may indicate that multiple clones sharing common antigenic pressure contribute to the TCR repertoire of IEL.

View larger version (34K): [in this window] [in a new window]   FIGURE 8. DNA sequences across V-J joints are overlapping in TCR + IEL and cryptopatch cells. CDR3 DNA sequences encoding the same predicted amino acid sequences isolated from TCR + IEL and cryptopatch cells were aligned across the TCR V and J region. N region nucleotides were determined by comparison with the genomic sequence, and the predicted amino acid sequences across the CDR3 are noted. Differences in DNA nucleotide usage are highlighted by red shading of the base pair. The title of previously published sequences is shown in italics. The number of individual isolates for each predicted amino acid sequence is noted. Data are derived from a subset of that shown in Table III.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

We propose that cryptopatch aggregates are storage sites for T cell progenitors of TCR ⁺ IEL. Consistent with this, TCRGV5 gene rearrangements segregated with CD3⁺ cryptopatch cells, but not with cryptopatch cells lacking CD3 expression. Although CD69 was expressed by a subset of cryptopatch T cells, the majority of T cells in cryptopatch aggregates were not dividing. We could not determine whether T cells move between cryptopatch aggregates and the epithelium or whether this movement is directional. The exchange of T cells between cryptopatch aggregates in the intestine of a host mouse implanted with an intestinal graft was found to be low (34), consistent with low levels of T cell exchange into cryptopatch aggregates under steady-state conditions. T cell proliferation in cryptopatch aggregates may be restricted in the absence of high turnover in the IEL compartment. Perhaps proliferation of cryptopatch T cells coincides with the burst of IEL colonization evident during postnatal maturation (41) or in response to Salmonella sp. infection where TCR ⁺ IEL expand (43).

Cryptopatch T cells expressing a surface TCR likely account for the predominance of in-frame TCRGV5 rearrangements detected in cryptopatch aggregates. Whether TCR ⁺ IEL expressing CD8 contribute to the small increase in the frequency of in-frame TCRGV5 rearrangements in TCR ⁺ IEL (compared with TCR ⁺ splenocytes) is under investigation. Other cell subsets that have rearranged, have been selected, but have down-regulated surface TCR expression could also account for the accumulation of in-frame TCRGV5 exons. Late-stage, triple-negative thymocytes are one source of cells that colonize cryptopatch aggregates, may serve as long-lived progenitors of IEL, and included cells that gave rise to TCRGV5⁺ IEL (44). The accumulation of TCRGV5 rearrangements in thymocytes just before expression of CD4 and CD8 is consistent with this possibility (45). The density of late-stage thymocyte precursors may, in addition, explain the variable detection of RAG gene expression documented in prior studies (24, 29, 30). It is possible that a bona fide primary lymphoid function for cryptopatch aggregates is present in early ontogeny and would establish long-lived precursors of IEL. However, the absence of RAG gene expression in adult mice indicates that this activity, if initially present, is extinguished by adult life.

The majority of TCRGV5 exons in cryptopatch aggregates of TCR ^–/– mice were not in-frame (abortive), suggesting that expression of a surface TCR accounted for positive selection or expansion of T cells in cryptopatch aggregates. It is likely that the TCR ⁺ cells present in cryptopatch aggregates of TCR ^–/– mice account for cells with out-of-frame (abortive) TCRGV5 gene rearrangements. The TCR ⁺ subset may additionally contribute to CD3⁺ cryptopatch cells with out-of-frame rearrangements in wild-type mice. In-frame rearrangements were high in TCR ⁺ IEL, but not 100%, presumably because ⁺ T cells using other TCRVG genes were present in IEL. TCRGV5 exons in cryptopatch aggregates with the same predicted amino acid sequence were more likely to be encoded by the same DNA sequence. By contrast, TCR ⁺ IEL with the same predicted amino acid sequence were encoded by a greater variety of DNA sequences. Of note, the most commonly isolated TCRGV5 exons in our study were encoded by identical DNA sequences to a subset of TCRGV5 exons previously published (8, 16). The diversity of TCRGV5 rearrangements isolated from cryptopatch aggregates was more restricted than TCR ⁺ IEL, which encompassed a greater TCR V5 repertoire. Because we did not sample every cryptopatch aggregate in the intestine, we may have examined only a limited portion of the potential cryptopatch T cell repertoire when compared with IEL pooled from the entire small intestine. More likely, the increased diversity in the IEL TCR repertoire suggests ⁺ IEL derive from multiple sources. By contrast, pauci-clonal selection or expansion of T cells may explain the comparatively limited TCR diversity among cryptopatch aggregates. The robust population of CD11c⁺ dendritic cells in cryptopatch aggregates could support Ag-driven selection or expansion of these few clones (25, 29). The overlap in predicted amino acid sequences between TCRGV5 exons isolated from cryptopatch aggregates and TCR ⁺ IEL supports the notion of a shared antigenic pressure. The apparent independence of IEL on the formation of cryptopatch aggregates would also support that IEL derive from noncryptopatch pathways (34).

The enrichment of in-frame TCRGV5 exons is similar between wild-type mice and athymic nude mice, demonstrating that the development of cryptopatch T cells does not a priori depend on the thymus. The reduced number of TCR ⁺ IEL in athymic nude mice suggests that extrathymic pathways of IEL development, including a cryptopatch pathway, are alone inefficient in directing T cells to the epithelium, in generating T cell precursors of IEL, or that the expansion of TCR ⁺ cells requires conventional TCR ⁺ cells in trans (31, 46, 47). Whether cryptopatch aggregates in athymic nude mice have or retain a primary lymphoid function when compared with euthymic mice is currently under investigation.

Cryptopatch T cells may uniquely impact the repertoire and function of IEL. If subsets of atypical IEL arise from cryptopatch T cell precursors, or aspects of their differentiation are dependent on cryptopatch aggregates, certain functional perturbations may be evident in mice lacking cryptopatch aggregates. The dramatic increase in the number of IEL in mice deficient in lymphotoxin signaling (which lack cryptopatch aggregates) when compared with IEL in wild-type mice (34) may be one indication. By contrast, in a model of ileitis (48), defects in the formation of cryptopatch aggregates were associated with a reduction in TCR ⁺ IEL and the subsequent development of inflammatory bowel disease (49). Insight into the role of cryptopatch T cells will derive from identifying cryptopatch-associated defects in IEL homeostasis. The relationship or interdependence between cryptopatch aggregates and ILF, and their relative contribution to immune responses in the intestine needs to be explored (33). The contribution of cryptopatch T cells to the repertoire of TCR ⁺ IEL is a question central to immune homeostasis in the intestine.

	Acknowledgments

 
We thank Drs. Mitchell Kronenberg and David Camerini for critical reading of the manuscript, Dr. Sheree Kuo and Samar Yageneh (University of Virginia, Charlottesville, VA) for technical assistance with LMD, and Hyojin Han (University of Virginia) and Ida Theodor (FACS Core Facility, Center for Immunology, University of California, Irvine, CA) for technical assistance.

	Disclosures

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 
The authors have no financial conflict of interest.

	Footnotes

 
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 supported by a grant from the National Institutes of Health (AI440941).

² B.S.P. and J.T. contributed equally to this work.

³ Address correspondence and reprint requests to Dr. Victoria Camerini, Associate Professor of Pediatrics, Childrens Hospital Los Angeles, Saban Research Institute, University of Southern California and the Keck School of Medicine, 4650 Sunset Boulevard, Mailstop 31, Los Angeles, CA 90027. E-mail address: vcamerini{at}chla.usc.edu

⁴ Abbreviations used in this paper: DETC, dendritic epidermal T cell; TCRGV; TCR -chain V region; TCRDV, TCR -chain V region; IEL, intestinal intraepithelial lymphocyte; ILF, intestinal lymphoid follicles; lin^–, lineage negative; DAPI, 4',6'-diamidino-2-phenylindole; DAB, 3' 3'diaminobenzidine; AEC, 3-amino-9-ethylcarbazole; LMD, laser-assisted microdissection.

Received for publication June 3, 2005. Accepted for publication March 22, 2006.

	References

Top Abstract Introduction Materials and Methods Results Discussion Disclosures References

 

1. Havran, W. L., J. P. Allison. 1988. Developmentally ordered appearance of thymocytes expressing different T-cell antigen receptors. Nature 335: 443-445. [Medline]
2. Garman, R. D., P. J. Doherty, D. H. Raulet. 1986. Diversity, rearrangement, and expression of murine T cell genes. Cell 45: 733-742. [Medline]
3. Bonneville, M., C. A. Janeway, Jr, K. Ito, W. Haser, I. Ishida, N. Nakanishi, S. Tonegawa. 1988. Intestinal intraepithelial lymphocytes are a distinct set of T cells. Nature 336: 479-481. [Medline]
4. Farr, A., S. Hosier, A. Nelson, S. Itohara, S. Tonegawa. 1990. Distribution of thymocytes expressing receptors in the murine thymus during development. J. Immunol. 144: 492-498. [Abstract]
5. Itohara, S., A. G. Farr, J. J. Lafaille, M. Bonneville, Y. Takagaki, W. Haas, S. Tonegawa. 1990. Homing of a thymocyte subset with homogeneous T-cell receptors to mucosal epithelia. Nature 343: 754-757. [Medline]
6. Goodman, T., L. Lefrancois. 1988. Expression of the T-cell receptor on intestinal CD8⁺ intraepithelial lymphocytes. Nature 333: 855-858. [Medline]
7. Goodman, T., L. Lefrancois. 1989. Intraepithelial lymphocytes: anatomical site, not T cell receptor form, dictates phenotype and function. J. Exp. Med. 170: 1569-1581. [Abstract/Free Full Text]
8. Asarnow, D. M., T. Goodman, L. LeFrancois, J. P. Allison. 1989. Distinct antigen receptor repertoires of two classes of murine epithelium-associated T cells. Nature 341: 60-62. [Medline]
9. Allison, J. P., D. H. Raulet. 1990. The immunobiology of ⁺ T cells. Semin. Immunol. 2: 59-65. [Medline]
10. Xiong, N., J. E. Baker, C. Kang, D. H. Raulet. 2004. The genomic arrangement of T cell receptor variable genes is a determinant of the developmental rearrangement pattern. Proc. Natl. Acad. Sci. USA 101: 260-265. [Abstract/Free Full Text]
11. Bonneville, M., S. Itohara, E. G. Krecko, P. Mombaerts, I. Ishida, M. Katsuki, A. Berns, A. G. Farr, C. A. Janeway, Jr, S. Tonegawa. 1990. Transgenic mice demonstrate that epithelial homing of T cells is determined by cell lineages independent of T cell receptor specificity. J. Exp. Med. 171: 1015-1026. [Abstract/Free Full Text]
12. Xiong, N., C. Kang, D. H. Raulet. 2004. Positive selection of dendritic epidermal T cell precursors in the fetal thymus determines expression of skin-homing receptors. Immunity 21: 121-131. [Medline]
13. Lin, T., H. Yoshida, G. Matsuzaki, S. R. Guehler, K. Nomoto, T. A. Barrett, D. R. Green. 1999. Autospecific thymocytes that escape negative selection find sanctuary in the intestine. J. Clin. Invest. 104: 1297-1305. [Medline]
14. Bandeira, A., S. Itohara, M. Bonneville, O. Burlen-Defranoux, T. Mota-Santos, A. Coutinho, S. Tonegawa. 1991. Extrathymic origin of intestinal intraepithelial lymphocytes bearing T-cell antigen receptor . Proc. Natl. Acad. Sci. USA 88: 43-47. [Abstract/Free Full Text]
15. De Geus, B., M. Van den Enden, C. Coolen, L. Nagelkerken, P. Van der Heijden, J. Rozing. 1990. Phenotype of intraepithelial lymphocytes in euthymic and athymic mice: implications for differentiation of cells bearing a CD3-associated T cell receptor. Eur. J. Immunol. 20: 291-298. [Medline]
16. Whetsell, M., R. L. Mosley, L. Whetsell, F. V. Schaefer, K. S. Miller, J. R. Klein. 1991. Rearrangement and junctional-site sequence analysis of T-cell receptor genes in intestinal intraepithelial lymphocytes from murine athymic chimeras. Mol. Cell. Biol. 11: 5902-5909. [Abstract/Free Full Text]
17. Guy-Grand, D., M. Malassis-Seris, C. Briottet, P. Vassalli. 1991. Cytotoxic differentiation of mouse gut thymo-dependent and independent intraepithelial T lymphocytes is induced locally: correlation between functional assays, presence of perforin and granzyme transcripts and cytoplasmic granules. J. Exp. Med. 173: 1549-1552. [Abstract/Free Full Text]
18. Guy-Grand, D., C. Vanden Broecke, C. Briottet, M. Malassis-Seris, 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: 505-510. [Medline]
19. Lin, T., G. Matsuzaki, H. Yoshida, N. Kobayashi, H. Kenai, K. Omoto, K. Nomoto. 1994. CD3^–CD8⁺ intestinal intraepithelial lymphocytes (IEL) and the extrathymic development of IEL. Eur. J. Immunol. 24: 1080-1087. [Medline]
20. Page, S. T., L. Y. Bogatzki, J. A. Hamerman, C. H. Sweenie, P. J. Hogarth, M. Malissen, R. M. Perlmutter, A. M. Pullen. 1998. Intestinal intraepithelial lymphocytes include precursors committed to the T cell receptor lineage. Proc. Natl. Acad. Sci. USA 95: 9459-9464. [Abstract/Free Full Text]
21. Hamad, M., M. Whetsell, J. Wang, J. R. Klein. 1997. T cell progenitors in the murine small intestine. Dev. Comp. Immunol. 21: 435-442. [Medline]
22. Lefrancois, L., L. Puddington. 1995. Extrathymic intestinal T-cell development: virtual reality?. Immunol. Today 16: 16-21. [Medline]
23. Kisielow, P., H. von Boehmer. 1995. Development and selection of T cells: facts and puzzles. Adv. Immunol. 58: 87-209. [Medline]
24. Kanamori, Y., K. Ishimaru, M. Nanno, K. Maki, K. Ikuta, H. Nariuchi, H. Ishikawa. 1996. Identification of novel lymphoid tissues in murine intestinal mucosa where clusters of c-kit⁺ IL-7R⁺ Thy1⁺ lympho-hemopoietic progenitors develop. J. Exp. Med. 184: 1449-1459. [Abstract/Free Full Text]
25. 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: 687-694. [Abstract/Free Full Text]
26. Saito, H., Y. Kanamori, T. Takemori, H. Nariuchi, E. Kubota, H. Takahashi-Iwanaga, T. Iwanaga, H. Ishikawa. 1998. Generation of intestinal T cells from progenitors residing in gut cryptopatches. Science 280: 275-278. [Abstract/Free Full Text]
27. Ishikawa, H., H. Saito, K. Suzuki, T. Oida, Y. Kanamori. 1999. New gut associated lymphoid tissue "cryptopatches" breed murine intestinal intraepithelial T cell precursors. Immunol. Res. 20: 243-250. [Medline]
28. Kuo, S., A. El Guindy, C. M. Panwala, P. M. Hagan, V. Camerini. 2001. Differential appearance of T cell subsets in the large and small intestine of neonatal mice. Pediatr. Res. 49: 543-551. [Medline]
29. Suzuki, K., T. Oida, H. Hamada, O. Hitotsumatsu, M. Watanabe, T. Hibi, H. Yamamoto, E. Kubota, S. Kaminogawa, H. Ishikawa. 2000. Gut cryptopatches: direct evidence of extrathymic anatomical sites for intestinal T lymphopoiesis. Immunity 13: 691-702. [Medline]
30. Oida, T., K. Suzuki, M. Nanno, Y. Kanamori, H. Saito, E. Kubota, S. Kato, M. Itoh, S. Kaminogawa, H. Ishikawa. 2000. Role of gut cryptopatches in early extrathymic maturation of intestinal intraepithelial T cells. J. Immunol. 164: 3616-3626. [Abstract/Free Full Text]
31. 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: 333-341. [Abstract/Free Full Text]
32. Lambolez, F., O. Azogui, A. M. Joret, C. Garcia, H. von Boehmer, J. Di Santo, S. Ezine, B. Rocha. 2002. Characterization of T cell differentiation in the murine gut. J. Exp. Med. 195: 437-449. [Abstract/Free Full Text]
33. Taylor, R. T., A Lugering, K. A. Newell, I. R. Williams. 2004. Intestinal cryptopatch formation in mice requires lymphotoxin and the lymphotoxin