AJP - GI Journal of Applied Physiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Gastrointest Liver Physiol 294: G1335-G1343, 2008. First published April 3, 2008; doi:10.1152/ajpgi.00265.2007
0193-1857/08 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
294/6/G1335    most recent
00265.2007v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Schjoldager, K. T.-B. G.
Right arrow Articles by Olsen, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Schjoldager, K. T.-B. G.
Right arrow Articles by Olsen, J.

MUCOSAL BIOLOGY

Cellular cross talk in the small intestinal mucosa: postnatal lymphocytic immigration elicits a specific epithelial transcriptional response

Katrine T.-B. G. Schjoldager,1 Henrik R. Maltesen,1 Sophie Balmer,1 Leif R. Lund,2 Mogens H. Claesson,3 Hans Sjöström,1 Jesper T. Troelsen,1 and Jørgen Olsen1

1Department of Cellular and Molecular Medicine, University of Copenhagen; 2The Finsen Laboratory, Rigshospitalet; and 3Department of International Health, Immunology and Microbiology, University of Copenhagen, Copenhagen, Denmark

Submitted 12 June 2007 ; accepted in final form 27 March 2008


   ABSTRACT
TOP
 ABSTRACT
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
During the early postnatal period lymphocytes migrate into the mouse small intestine. Migrating infiltrative lymphocytes have the potential to affect the epithelial cells via secreted cytokines. Such cross talk can result in the elicitation of an epithelial transcriptional response. Knowledge about such physiological cross talk between the immune system and the epithelium in the postnatal small intestinal mucosa is lacking. We have investigated the transcriptome changes occurring in the postnatal mouse small intestine using DNA microarray technology, immunocytochemistry, and quantitative real-time RT-PCR analysis. The DNA microarray data were analyzed bioinformatically by using a combination of projections to latent structures analysis and functional annotation analysis. The results show that infiltrating lymphocytes appear in the mouse small intestine in the late postweaning period and give rise to distinct changes in the epithelial transcriptome. Of particular interest is the expression of three genes encoding a mucin (Muc4), a mucinlike protein (16000D21Rik), and ATP citrate lyase (Acly). All three genes were shown to be expressed by the epithelium and to be upregulated in response to lymphocytic migration into the small intestinal mucosa.

intestinal development; projections to latent structures; intraepithelial lymphocytes; intestinal gene expression; DNA microarray

IN THE ADULT MOUSE SMALL INTESTINAL epithelium, intraepithelial lymphocytes (IELs) are present at ~10% of the number of villus epithelial cells (for a review, see Ref. 18). The small intestinal IELs are mostly CD8-positive T cells (18). The migration of immune cells into the small intestinal mucosa is a postnatal event, which has been studied in detail in the mouse (27) and rat (3). Migration of IELs into the small intestinal epithelium occurs at weaning in the mouse and depends on epithelially secreted chemokine ligand 25 (CCL25) (27). The number of IELs in the rat small intestinal epithelium also increases markedly during or immediately after weaning (3). At approximately the same time the crypt size increases in the rat small intestine (20). These studies in rodents therefore suggest that the physiological crypt hyperplasia seen in the postnatal intestinal development may be a physiological response to the migration of immune system cells into the small intestinal mucosa. Such a response might be driven by cytokines released from IELs and growth factors released from cytokine stimulated stromal cells. If this is the case, changes in the epithelial transcriptome should accompany the migration of lymphocytes into the small intestinal epithelium. Massive changes in the transcriptomes of epithelial cells do indeed take place during development from the fetal to the adult small intestine where biological processes related to immune function and digestion are upregulated, whereas cell cycle-related processes are downregulated (31). The altered expression of small intestinal digestive enzymes around weaning is also well documented (19). It is therefore highly probable that some of the gene expression changes that occur in the intestinal epithelial cells in the early postnatal period are driven by factors released from IELs. Cellular studies support this hypothesis. For instance, IFN-{gamma} has been shown to downregulate the epithelial expression of sodium/hydrogen exchangers (28), and interleukin-2 has been shown to increase the expression of transforming growth factor-β in intestinal epithelial cells (8). Thus the available circumstantial evidence suggests that a physiological epithelial gene expression response might exist for the migration of mature T lymphocytes into the small intestine but the magnitude, the characteristics of this response and the relationship to physiological crypt hyperplasia are unknown at present. Hyperplasia of the small intestinal crypts is observed in Crohn's disease and in celiac disease (for reviews see Refs. 7, 16, 25, 26). Both diseases are characterized by increased lymphocytic infiltration, and basic knowledge about the normal epithelial response to migration of lymphocytes into the small intestinal epithelium would also be valuable for our understanding of these pathological conditions.

The aim of the present work was to prove the existence of a small intestinal epithelial transcriptional response to physiological postnatal lymphocytic migration into the small intestinal mucosa, to give a first characterization of this response and finally to establish its importance in physiological crypt hyperplasia. The gene expression changes in the mouse ileum from the preweaning period, the periweaning period, and the postweaning period were measured by high-density oligonucleotide array hybridization and analyzed by multivariate dimension reduction technique [projections to latent structures (PLS) analysis] combined with functional annotation analysis. This analysis was supplemented by expression studies of selected genes in wild-type C57BL/6 and RAG2-deficient mice. In summary, the results demonstrate that the small intestinal epithelium has a gene expression response to the process of postnatal lymphocytic migration into the ileal epithelium but that crypt hyperplasia occurs independently of this response.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Animals. C57BL/6 mice were kept on standard nutritional intake and fed ad libitum in the animal facility at the Panum Institute. B6.129S6-Rag2tm1Fwa N12 (RAG2-deficient) mice were obtained from Taconic (Taconic Europe, Lille Skensved, Denmark). Mice were killed by cervical dislocation and the abdomen was immediately opened by use of surgical scissors. The small intestine was located, and for RNA purification purposes a 1-cm fragment was cut from the distal part of the ileum and snap-frozen in 2-methylbutane cooled by dry ice. For isolation of intestinal epithelial cells, a 10-cm fragment of the ileum was immediately placed in ice-cold PBS. For immunocytochemistry, fragments (ranging between 0.5 and 1 cm in length) of the distal part of the ileum were placed in 4% paraformaldehyde (4°C) for 24 h. The tissue specimens were subsequently placed in 60% ethanol for storage (4°C) until dehydration and paraffin embedding.

Definition of time points and samples for analysis. Samples for microarray analysis were taken from wild-type C57BL/6 mice of age 4, 7, 9, 11, 24, and 32 days. Samples from three individual mice were obtained at each time point, and for each of these individual samples a DNA microarray hybridization experiment was conducted. The samples used for microarray analysis were also used for quantitative real-time RT-PCR analysis, but additional samples were obtained for quantitative real-time RT-PCR analysis from 24-day-old (5 additional samples) and from 32-day-old (3 additional samples) C57BL/6 mice. Samples from five individual B6.129S6-Rag2tm1Fwa N12 24-day-old and 32-day-old mice were also obtained.

The mothers were removed from the offspring at day 21. The samples from the 4-, 7-, 9-, and 11-day-old mice are referred to as the preweaning time point group. The samples from the 24-day-old mice are referred to as the periweaning time point group, and the samples obtained from the 32-day-old mice are referred to as the postweaning time point group.

IEC isolation. Intestinal epithelial cells (IEC) were isolated essentially as described previously (31). In brief, 10-cm ileal segments were flushed with ice-cold PBS. The intestines were then inverted on plastic sticks and placed in a 15-ml tube with a screw cap containing ice-cold chelating buffer (27 mM trisodium citrate, 5 mM Na2HPO4, 96 mM NaCl, 8 mM KH2PO3, 1.5 mM KCl, 59 mM D-sorbitol, 44 mM sucrose, 0.5 mM DTT, pH 7.3). The tubes were incubated (15–16 h, 4°C) and rotated (1 h, 4°C). This procedure leads to isolation of complete villi as well as villus fragments. Crypt cells and remaining villus fragments were shaken off by use of a handheld minishaker (1 min). The fractions were combined and the cell suspension containing the total epithelium was centrifuged at 3,500 g (10 min, 4°C). The cell pellet was resuspended in 5 ml of chelating buffer. The suspension was pipetted up and down until ~90% of the suspension consisted of single cells. The cells were pelleted (3,500 g, 10 min, 4°C) and resuspended in 2 ml of PBS + 0.1% BSA. An aliquot of cells was kept for total RNA quantification, and 300 µl of washed anti-CD8 (lyt-2) Dynabeads (Invitrogen, Carlsbad, CA) was added to the remaining cells and incubated (with rocking) for 0.5 h (4°C). The CD8+ cells were isolated in a Dynabead magnet and the supernatant was removed. CD8-negative cells were pelleted (10,000 g, 10 min, 4°C) from the supernatant.

Immunohistochemistry. Terminal ileum tissue biopsies were fixed in 4% paraformaldehyde and embedded in paraffin. Thin sections (7 µm) were cut and mounted on glass slides (SuperFrostPlus, Menzel-Glaser, Braunschweig, Germany).

The tissue sections were deparaffinized in xylene (5 min) and rehydrated through graded concentrations of ethanol to PBS (1 min in each of the following solutions: 99% ethanol, 96% ethanol, 70% ethanol).

Unmasking of antigens was performed by boiling the sections (10 min) in 0.01 M citrate buffer (pH 6). After being cooled to room temperature the sections were blocked with blocking buffer (50 mM Tris, pH 7.4, 150 mM NaCl, 0.5% ovalbumin, 0.1% gelatin, 0.05% Tween-20, and 0.2% Teleostan gelatin) for 30 min.

Sections were incubated with either a rabbit antibody against CD3 (ab-16669-500, Abcam, Cambridge, UK), a goat antibody against CCL5 (AF478, R&D Systems, Abingdon, UK), or a rabbit antibody against ATP citrate lyase (ab40793, Abcam) diluted 1/100 in blocking-buffer overnight at 4°C. The sections were washed three times (10 min) with blocking buffer and incubated (30 min, room temperature) with a goat antibodies against rabbit IgG (A11008 [GenBank] , Alexa Fluor-488 or A11037 [GenBank] , Alexa Fluor-594, Invitrogen) diluted 1/200 in blocking buffer. Finally, the sections were washed twice (5 min) in PBS and mounted in Fluorescent Mounting Medium (Dako Cytomation, Glostrup, Denmark). The sections were analyzed under a Leica DM 4000 B microscope equipped with a Leica DC300 FX camera.

Western blotting. Pieces of mouse ileum were homogenized in 200 mM Tris·HCl (pH 7.5), 1.5 M NaCl, 10 mM EDTA, 10 mM EGTA, 10% Triton supplemented with a protease inhibitor cocktail (Sigma-Aldrich, St. Louis, MO). Protein concentration was determined by using the Bio-Rad protein assay kit (Bio-Rad Laboratories, Hercules, CA). A 10-µg protein extract was loaded in each lane on a 10% Nupage gel (Invitrogen). Following electrophoresis proteins were transferred to an Immobilon-P filter (Millipore, Billerica, MA) by semidry blotting. The filter was incubated in blocking buffer (PBS supplemented with 2% nonfat dry milk and 2% Tween-20) for 1 h at room temperature. ATP citrate lyase antibody (ab40793, Abcam) was added, and the incubation continued overnight at 4°C. The filter was washed in blocking buffer (room temperature) and incubated with a horseradish peroxidase-conjugated donkey anti-rabbit IgG. Following a wash in blocking buffer, bound antibody was visualized by use of the Pierce ECL Western blotting substrate (Pierce Biotechnology, Rockford, IL).

Crypt length measurements. Sections of small intestinal tissue from 24-day-old (n = 7) and 32-day-old (n = 6) C57BL/6 mice and from 24-day-old (n = 4) and 32-day-old (n = 4) B6.129S6-Rag2tm1Fwa N12 mice were stained with hematoxylin and eosin according to standard histochemical procedures. Following staining, the sections were analyzed under a Leica DM 4000 B microscope equipped with a Leica DC300 FX camera. Each section was photographed, so that the entire section was covered. With the distance measurement tool implemented in the ImageManager500 software, an average of eight crypt lengths were measured per mouse. For each mouse the average crypt length was calculated and used to calculate a mean crypt length for each time point. The change in the mean crypt length from 24 to 32 days was analyzed statistically by a Student's t-test.

RNA extraction and analysis. Total RNA was isolated by using the E.Z.N.A. total RNA kit (Omega BioTek, Doraville, GA). Frozen tissue was homogenized [either with an Ultra-Turrax motor T8 (IKA-WERKE, Staufen, Köngen, Germany) or a mini-BeadBeater-8 (Biospec Products, Bartlesville, OK)] directly in lysis buffer. A phenol-chloroform extraction step was introduced for improved RNA quality. The aqueous phase was loaded on a HiBindRNA spin column followed by an on-column DNase I digestion, according to the manufacturer's protocol (Omega BioTek). RNA quality and quantity were analyzed by capillary electrophoresis on the Agilent 2100 bioanalyzer (Agilent Technologies, Santa Clara, CA). Probe synthesis and hybridization with the 430 2.0 GeneChips (Affymetrix, Santa Clara, CA) were performed at the microarray core facility (Rigshospitalet, Copenhagen, Denmark) according to previous protocols (17, 31).

Primers (Table 1) for Tmed5, Itsn2, Acly, and Ccl5 were found in the Primer Bank (33).


View this table:
[in this window]
[in a new window]

  		Table 1. Primers used for quantitative real-time RT-PCR

 
Primers for 1600029d21riken, Erbb2ip, Gzma, and Muc4 were designed by using PRIMER3 software (29). The Actb mRNA was used for normalization and the Apoa4 transcript as a specific marker for villus epithelium. All primers were synthesized by MWG-Biotech (Ebersberg, Germany).

For validation, the primers were initially used in a PCR reaction with Taq polymerase (Fermentas), using mouse intestinal cDNA as template, and the PCR products were run in a 3% agarose gel to confirm the expected size. The products were gel purified with E.Z.N.A.'s gel extraction kit (Omega BioTek) and used as standards.

One microgram total RNA was used for first-strand cDNA synthesis using RevertAid H minus first-strand cDNA synthesis kit (Fermentas) following the manufacturer's instructions. For real-time RT-PCR SYBR Premix Ex Taq (Perfect Real Time) (Takara Bio, Shiga, Japan) was used. The reactions were mixed in a 20-µl LightCycler capillary tube (Roche Diagnostics, Hvidovre, Denmark) and 50 ng of first-strand cDNA was used as template for the reaction. To standardize the reactions a serial 10-fold dilution of 1 µl of gel-purified PCR products were used as templates in independent reactions. Melting curves were inspected after each run to rule out the occurrence of unwanted amplified PCR fragments such as primer dimers. Furthermore, transcription levels in each sample were normalized to the β-actin level in the same sample.

Microarray analysis and multivariate modeling. Scanning of the GeneChips was performed with the Affymetrix GeneChip scanner, and normalized expression values were calculated with the Bioconductor open-source software (15) by using the implemented Robust Multi Array analysis (RMA) (22) for the R environment for statistical computing (4). The complete dataset was deposited in the Gene Expression Omnibus under the series accession number GSE8065 but can also be found in the supplementary file 1.

The data set was filtered by removing probe sets where the lower quartile of the RMA expression measures was below 5.5. This reduced the data set to 16,045 probe sets. The filtered data set was subjected to an initial principal component analysis (PCA) (for a review of PCA, see Ref. 35) using the SIMCA-P 11 software (Umetrics, Umeå, Sweden). The PCA model was derived by using the autofitting option in SIMCA-P 11, which led to a two-component model that explained 47% of the total variance. On the basis of the PCA results it was decided to build a multivariate model defining three distinct gene expression patterns (a preweaning, a periweaning, and a postweaning pattern) using PLS (also called partial least squares regression analysis) analysis (for reviews of PLS, see Refs. 9 and 13). A two-component PLS model was then built using the SIMCA-P 11 software. The model explained 82% of the variance. The SIMCA-P 11 cross-validation function was applied to check for overfitting, which did not occur in the model. The probe sets with the highest positive regression values for each dependent variable (preweaning type, periweaning type, or postweaning type) are the most important probe sets for defining these types of gene expression patterns. The probe sets with the most negative regression coefficients are those that are of the least importance for defining these gene expression patterns. It is therefore possible to carry out a functional annotation analysis by comparing the annotation associated with the probe sets with the highest positive regression coefficients with the annotation of the probe sets with the most negative regression coefficients for each dependent variable. This functional annotation analysis is carried out for each dimension in the PLS model. In the present work the functional annotation analysis was conducted using the GoSurfer software (36). We have previously demonstrated the feasibility of this strategy in relation to PCA (2). The probe sets selected for the analysis for each dependent variable (preweaning, periweaning, and postweaning) were the 500 genes with the highest regression coefficients and the 500 probe sets with the most negative regression coefficients (supplementary file 2).

The PLS model suggested that lymphocyte-driven gene expression changes occurred between day 24 and day 32, and a t-test was performed comparing the 24-day data with the 32-day data (supplementary file 3). This calculation was performed using the R environment for statistical computing (4). A significance level of 0.001 was chosen.


   RESULTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Identification and functional interpretation of the postnatal ileal transcriptomic changes. On the basis of the available literature (3, 27), it was expected that T lymphocyte-induced transcriptomic changes would occur in a relatively short time interval around the time of weaning, which starts 3 wk after birth. Terminal ileum was therefore collected from mice at 4, 7, 9, 11 days (preweaning), 24 days (periweaning), and 32 days (postweaning) of age. A genomewide gene expression analysis was carried out by hybridization to the Affymetrix mouse genome 430A 2.0 GeneChips using probes generated from RNA isolated from total ileum. Unsupervised PCA suggested (not shown) the presence of three different gene expression patterns corresponding to a preweaning pattern (4–11 days), a periweaning pattern (24 days) and a postweaning pattern (32 days). This information was used to build a multivariate gene expression model by PLS analysis. Visualization of the samples in a scores plot of the first two PLS components (Fig. 1) revealed that most of the variation (which is by definition represented in the first PLS component) is important for distinguishing between the preweaning pattern and the two later time point patterns. A smaller part of the variation (represented by the second PLS component) is important for distinguishing between the periweaning pattern and the postweaning pattern. The preweaning group of samples has very little variation along the second PLS component whereas the two later time point groups diverge toward positive (24 days) and negative (32 days) values. The probe sets that are important for defining the position of the samples in the scores plot can be identified as the probe sets with the highest positive PLS regression coefficients onto the three time point groups for each PLS component. A functional annotation analysis was conducted on these probe sets. The complete analysis is listed in Table 2, and a simplified version is depicted in Fig. 1. Of particular interest with respect to lymphocyte migration into the intestine is the finding that the Gene Ontology (GO) terms 42110 (T cell activation) and 50863 (regulation of T cell activation) were found to be overrepresented in the annotation of probe sets with the highest positive regression coefficients of the second PLS component axis for the postweaning pattern. The transcriptome analysis therefore suggests that alterations in the small intestinal gene expression related to appearance of T lymphocytes in the epithelium occur from the periweaning time point to the postweaning time point. The CD3 antigen, epsilon peptide, was found among the genes annotated with the GO term 50863 (regulation of T cell activation) that were important for separating the 32- and 24-day samples along the second PLS component axis. An immunocytochemical analysis was therefore carried out with an anti-CD3{varepsilon} antibody (Fig. 2). This analysis confirmed that CD3{varepsilon}-positive cells are found in the ileal villus epithelium and stroma at day 32 but not at day 24.


Figure 1
View larger version (13K):
[in this window]
[in a new window]

  		Fig. 1. Projections to latent structures (PLS) model of gene expression changes occurring in the mouse ileum from postnatal day 4 to day 32. Genomewide gene expression analysis was carried out by using Affymetrix GeneChips and hybridization probes derived from ileal RNA. The derived expression measures were filtered (lower quartile expression measure above 5.5) and subjected to principal component analysis (PCA) (see MATERIALS AND METHODS for details). The PCA analysis suggested the existence of 3 distinct types of expression patterns: a preweaning pattern (4–11 days), a periweaning pattern (24 days), and a postweaning pattern (32 days). On the basis of the PCA results, a PLS model was built with the purpose of identifying the probe sets (genes) that are characteristic for each of the 3 expression patterns. The projections of each of the 18 samples in a PLS scores plot of the first 2 components are shown. It can be seen that the first PLS component axis separates the preweaning samples from the later time point samples whereas the second PLS component axis separates all 3 types of sample from each other. The regression coefficients for the probe sets onto each expression pattern type were used to perform a functional annotation analysis. The result of the complete analysis is shown in Table 2. A representative Gene Ontology (GO) term associated with the genes characteristic for each expression pattern type is written on the scores plot. The smaller inserted axes indicate the PLS component axis with which each of the representative GO term is associated. The 4- to 11-day group is mainly projected along the first PLS component and only GO terms for this axis are accordingly shown for this group. Specifically, it can be seen that probe sets annotated with the term "T cell activation" define the position of the 32-day samples along the second PLS component axis. Thus migration of T cells into the epithelium appears to occur between day 24 and day 32.

 

View this table:
[in this window]
[in a new window]

  		Table 2. Gene Ontology terms overrepresented in the annotation associated with the probe sets with the highest positive PLS regression coefficients onto the 3 time groups

 

Figure 2
View larger version (6K):
[in this window]
[in a new window]

  		Fig. 2. CD3{varepsilon}-positive cells in the ileal mucosa. Sections of mouse day 24 (A) and day 32 (B) ileum were incubated with a rabbit antibody raised against mouse CD3{varepsilon}. Bound antibody was visualized by using an Alexa-488 conjugate. CD3{varepsilon}-positive cells are seen (arrows point to 3 examples of CD3{varepsilon}-positive intraepithelial cells) in day 32 epithelium and stroma whereas such cells are absent from day 24 epithelium. Optical magnification: x20.

 
The terms 9607 (response to biotic stimulus), 6952 (defense response), 6955 (immune response), and 51707 (response to other organism) were found to be overrepresented in the annotation of the probe sets with the highest positive regression coefficients of the first PLS component for both the periweaning and the postweaning patterns. Examples of probe sets from this list annotated with the term "immune response" are chemokine (C-C) ligand 25 (CCL25), CD14, and immunoglobulin heavy chain. CCL25 is expressed by small intestinal epithelial cells and stimulates the migration of T lymphocytes into the epithelium (27), CD14 is a marker for monocytes/macrophages, and immunoglobulin mRNAs are expressed in B lymphocytes. Thus the migration of monocytes and B lymphocytes into the intestinal mucosa might precede the migration of T lymphocytes.

Identification of genes with increased ileal expression between the early and late postweaning time points. A Student's t-test was conducted to identify probe sets with differential expression between the 24-day and 32-day time points. In total, 112 genes turned out to be significantly differentially expressed (supplementary file 3) by this approach. Eight genes were selected for validation by quantitative real-time RT-PCR. Four of these genes (ATP citrate lyase, Acly; chemokine ligand 5, Ccl5; RIKEN cDNA, Rik1600029d21; and Erbb2 interacting protein isoform 2, Erbb2ip) correspond to probe sets that were both found differentially expressed by the t-test approach and that also appeared on the list of probe sets with high positive regression coefficients for the second PLS component for the 32-day time point. Three of the genes (SH3 domain protein 1B, Itsn2; mucin 4, Muc4; and transmembrane emp24 protein transport domain, Tmed5) were selected from the list of significantly expressed genes according to the t-test only, and one additional gene (Granzyme A, Gzma) was selected from the regression coefficient list only. All the selected genes demonstrated a significant upregulation in their ileal expression from 24-day-old to 32-day-old mice when measured by quantitative real-time RT-PCR (Fig. 3, A and B).


Figure 3
View larger version (17K):
[in this window]
[in a new window]

  		Fig. 3. Expression of selected transcripts in day 24 and day 32 ileum of C57BL/6 and B6.129S6-Rag2tm1Fwa N12 (RAG2-deficient) mice as measured by real-time RT-PCR. Scatterplots showing copy numbers of Acly, Ccl5, Erbb2ip, and Gzma (A) and Itsn2, Muc4, 1600029D21Rik, and Tmed5 (B) measured in ileal tissue from C57BL/6 and RAG2-deficient mice. All measurements are normalized to expression level of β-actin (Actb). The horizontal line marks the mean. D, days. *P < 0.05, **P < 0.01; n.s., not significant.

 
Identification of epithelial transcripts that are induced in response to postnatal T lymphocytic infiltration. To investigate whether the upregulation in ileal expression seen for the eight selected transcripts between day 24 and day 32 depended on the presence of infiltrating mature lymphocytes, we analyzed their expression in mice with a targeted disruption of both Rag2 alleles. The Rag2 gene encodes the RAG2 recombinase enzyme involved in V(D)J rearrangements (14), and mice lacking this enzyme cannot generate mature B and T lymphocytes. The Tmed5 and Itsn2 transcripts were differentially expressed in the ileum of RAG2-deficient mice between the 24- and 32-day time points (Fig. 3B) whereas the ileal expression of the remaining six transcripts (Acly, Ccl5, Erbb2ip, Gzma, Muc4, and 1600029D21Rik) did not change from day 24 to day 32 in the RAG2-deficient mice (Fig. 3, A and B). The upregulation of these transcripts therefore depends on the presence of mature lymphocytes in the epithelium.

To determine to what extent the transcripts found to be upregulated are a part of an epithelial response to the presence of lymphocytes, we determined the expression of the eight selected transcripts in CD8-positive and -negative cells derived from isolated ileal epithelium of C57BL/6 mice. For epithelial isolation we used the method originally developed for rat small intestinal epithelium by Flint et al. (11) depending on citrate as a chelating agent. This epithelial isolation procedure is carried out at low temperature, and high-quality RNA can be obtained from cells isolated by this procedure. Our laboratory has previously described an optimized version of this protocol (31) for mouse small intestine. Microscopic examination of the muscular and connective tissue remaining after epithelial isolation show that lamina propria remains intact after the release of the epithelium and that lamina propria lymphocytes are not released together with the epithelium.

The apolipoprotein A4 transcript (Apoa4) is involved in lipoprotein synthesis and is only expressed in the epithelial cells; it was therefore selected as an epithelial marker transcript. As seen in Table 3, there is a fivefold enrichment in the amount of Apoa4 transcript found in RNA isolated from the CD8-depleted cells compared with the amount of Apoa4 transcript found in RNA isolated from the CD8-positive cells. The presence of Apoa4 transcripts in the CD8-positive cell fraction is due to epithelial contamination, which therefore at the most constitutes 20%. This contamination might occur because clumps of epithelial cells attached to IEL are pulled down by the anti-CD8 magnetic beads. Alternatively, epithelial clumps may simply be trapped during the bead-isolation procedure. In contrast, the CD8-negative fraction appears to contain very few contaminating lymphocytes. This can be seen because the Gzma transcript, which is expressed in T cells and natural killer cells (32), is enriched by 50-fold in the CD8-positive fraction. The ratio of CD8-positive IEL to epithelial cells was estimated by measuring the total RNA content in the epithelium before fractionation in the CD8-negative and CD8-positive fractions. The result showed that the CD8-positive cells on average constitute 7% of the cells in the mouse ileal epithelium. On the basis of the measurements in Table 3 we calculated, for each transcript, the increase in copy number found in total ileal RNA that would occur if the ileal epithelium was populated with lymphocytes constituting 7% of the cells. As seen in Table 3, Ccl5 and Gzma transcripts carried by infiltrating T lymphocytes would cause a major increase in the level of these transcripts in the ileum (3.2- and 4.4-fold, respectively). For all other transcripts the calculated fold increase due to transcripts carried by lymphocytes was found to be close to one. The transcripts Acly, Erbb2ip, Muc4, and Rik1600029d21 have a measured increase in copy number ranging from 1.7-fold to 4.9-fold in total ileal RNA from the 24-day ileum to the 32-day ileum. Thus the increase in ileal expression observed for these transcripts cannot be due to transcripts carried by infiltrating lymphocytes but is caused by an increased epithelial expression.


View this table:
[in this window]
[in a new window]

  		Table 3. Quantitative RT-PCR measurements of selected mRNAs in the mouse ileum

 
To validate these findings at the protein level histological sections of mouse ileum from 24- and 32-day-old mice were stained with antibodies against CCL5 and ATP citrate lyase (Fig. 4). None of the antibodies stained the ileum from 24-day-old mice. The CCL5 antibody strongly stained few cells scattered in the stroma and epithelium around the crypt-villus junction. This is in agreement with the quantitative RT-PCR measurements, which suggest that CD8-positive infiltrating T lymphocytes are the cause of a large increase in the ileal Ccl5 expression in day 32 mice. Some weak CCL5 staining was also observed of the villus epithelial cells from 32-day-old mice, indicating that the epithelium might also express CCL5 at this stage. The ATP citrate lyase antibody strongly stained the villus epithelial cells from the 32-day-old mice. The staining was mainly subapical and in agreement with a cytoplasmic localization of ATP citrate lyase. The increase in small intestinal ATP citrate lyase expression was also confirmed by Western blotting (Fig. 5). Thus it can be concluded that the ileal epithelial expression of ATP citrate lyase is increased during the aging of the mice from 24 to 32 days.


Figure 4
View larger version (27K):
[in this window]
[in a new window]

  		Fig. 4. Expression of CCL5 and ATP citrate lyase in day 24 and day 32 ileum of C57BL/6 mice. Sections of mouse ileum were stained with antibodies against either Ccl5 or Acly. Bound antibody was visualized by using an Alexa-594 conjugate. The CCL5 antibody stains single cells scattered in the stroma (2 examples are marked with white arrows) and villus epithelium in the ileum of 32-day-old mice. In addition, some staining is also observed in the villus epithelial cells in 32-day-old mice. The ATP citrate lyase antibody displays a subapical cytoplasmic staining (exemplified by 3 white arrows) in the ileal epithelium from 32-day-old mice. No staining is observed in the ileum from 24-day-old mice with either antibody. Optical magnification: x20.

 

Figure 5
View larger version (14K):
[in this window]
[in a new window]

  		Fig. 5. Western blot of ATP citrate lyase. Protein extracts from the ileum of 24-day-old and 32-day-old mice were separated on a 10% SDS-PAGE gel followed by transfer to a filter, which was probed with an antibody against ATP citrate lyase. A single band with mobility slightly slower than the 117-kDa marker protein is observed only with the extract from 32-day-old mice. The mobility of the reactive band is compatible with a predicted size of 120 kDa for the mouse ATP citrate lyase.

 
Physiological ileal crypt hyperplasia is not dependent on T lymphocyte migration into the intestinal epithelium. To evaluate whether the lymphocytic migration into the small intestinal epithelium induced any morphological changes in the ileal epithelium, hematoxylin and eosin-stained ileal sections from C57BL/6 and RAG2-deficient mice were studied by light microscopy. A significant 1.3-fold increase in mean crypt length was observed between day 24 and day 32 ileum in both mouse strains (Fig. 6). Thus, although an increase in crypt length in the mouse ileum does occur over the same time period when T lymphocytes enter the epithelium, functional T cells are not involved in the physiological crypt hyperplasia.


Figure 6
View larger version (6K):
[in this window]
[in a new window]

  		Fig. 6. Crypt length increase from the periweaning to postweaning period. The increase in crypt length (µm) measured in histological hematoxylin and eosin-stained small intestinal sections of 24-day-old and 32-day-old mice from wild-type (C57BL/6) and RAG2-deficient mice on C57BL/6 background (RAG2/C57BL/6). The horizontal line marks the mean. **P < 0.01.

 

   DISCUSSION
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
In the present work we have used a multivariate dimension reduction technique (PLS) to translate the collected genomewide gene expression profiles of postnatal mouse small intestinal development into an easily, functionally interpretable result. Most notable, biological processes related to immune function are activated in the ileum during the preweaning to periweaning transition. This might occur because the diet of the pups is gradually changed from milk to solid food, resulting in a changed antigenic profile of the diet and in the colonization by gut microbes. T cell activation becomes evident in the gene expression profile during the transition from day 24 to day 32. Accordingly, the results of our investigation show that T lymphocytes expressing CD3 migrate into the ileal epithelium in the time interval between the periweaning and the postweaning period between day 24 and day 32 in mice. This is slightly later than reported previously (27). We were not, however, able to detect CD3-positive cells in the ileal epithelium by immunocytochemistry in the 24-day-old mice. One possible explanation for the discrepancy might be that the proximal small intestine was analyzed in the study by Marsal et al. (27), whereas the distal part of the small intestine was used in the present study. Evaluation of eight transcripts that were found to be upregulated from day 24 to day 32 showed that for four of these the upregulation was explained by their altered epithelial expression. The transcripts Erbb2ip, Muc4, and 1600029d21Rik are all expressed at comparable levels per cell in intestinal epithelial cells and in T lymphocytes, whereas the transcript Acly is expressed at slightly higher levels per cell in T lymphocytes than in intestinal epithelial cells. Lymphocytes constitute less that 10% of the epithelial cells (we estimate 7% in the present work), and lymphocytic expression levels for all four transcripts are too low to explain the increase seen in their expression from day 24 to day 32 in the ileum. Moreover, the increase in epithelial ATP citrate lyase expression in the ileum of 32-day-old mice was confirmed by immunocytochemistry. The conclusion is therefore that the lymphocytic migration into the epithelium elicits an epithelial transcriptional response. Moreover, the response was due to lymphocytic migration into the epithelium because it was not seen in the RAG2-deficient mice.

Potential functions of the epithelial transcriptional response to intestinal lymphocytic migration into the small intestine. The Muc4 gene encodes a membrane-bound mucin that is known to be highly expressed in the gastrointestinal tract. The mucins have a protein core, which is heavily glycosylated by O- and N-bound sugar chains. The intestinal mucins are important for intestinal epithelial integrity and for the mucosal colonization by microflora (for a review, see Ref. 24). Interestingly, the Muc2 gene encodes another mucin that is also found on the list of genes that are upregulated from day 24 to day 32 as predicted by the PLS analysis (supplementary file 2). The 1600029D21Rik gene is predicted to encode a 285-amino acid protein with unknown function. We therefore searched the Pfam database (10) and found that the predicted protein showed some similarity to the Pfam protein family named "Mucin" (entry no. PF01456), which is a family of trypanosomal mucinlike proteins. The predicted amino acid sequence was also submitted to the Signal P 3.0 prediction server (1), which predicted (with very high probability) the presence of an NH2-terminal signal sequence. Thus the 1600029D21Rik transcript is likely to encode a secreted soluble protein with mucinlike properties. Also taking the Muc2 transcript into account, it seems reasonable to propose that one consequence of the lymphocytic migration into the ileal epithelium is the upregulation of a subset of epithelial mucins. Interestingly, it has been demonstrated that the expression of the Muc2 protein is increased in the ileal epithelium of patients with active celiac disease (12). This might therefore represent an increase of the normal physiological response to the migration of lymphocytes into the ileal epithelium.

The Erbb2ip transcript (also known as Erbin) encodes a protein that was first described as an interactor of the ErbB2 receptor. Recently, however, the Erbb2ip-encoded protein was demonstrated to interact with the Nod2 protein and thereby inhibit Nod2-mediated NF-{kappa}B activation (23). Through the Erbb2ip-encoded protein, the migration of lymphocytes into the small intestinal epithelium might therefore modulate the epithelial Nod2/NF-{kappa}B network.

The Acly gene encodes the ATP citrate lyase, which splits cytosolic citrate into acetyl-CoA and oxaloacetate. This reaction provides acetyl-CoA for fatty acid and cholesterol synthesis (6). Cholesterol is an important component of the specialized microvillar membrane of the differentiated small intestinal enterocyte (5), and the stimulation of Acly expression by lymphocytic and epithelial cross talk might therefore be important for the establishment of the correct lipid composition of the microvillar membrane from adult mice.

Recruitment of IEL lymphocytes and their activation of the epithelial transcriptional response. The Ccl25 transcript was already found to be upregulated at day 24 in the gene expression analysis and this finding is in accordance with the proposed function of CCL25 as an important chemoattractant for T lymphocytes in the small intestinal epithelium (27). The Ccl5 transcript was found to be highly expressed in the T lymphocytes migrating into the small intestine, which is consistent with the fact that the Ccl5 transcript was originally identified as a transcript of activated T lymphocytes (for a review, see Ref. 34). As a secreted soluble signaling molecule the CCL5 protein is a candidate for eliciting the epithelial transcriptional response. On the basis of the expression analysis, the known chemokine receptors binding to CCL5 (CCR1, CCR3, and CCR5; for a review, see Ref. 30) are not, however, expressed in the ileum of 32-day-old mice. Recently, the orphan G protein-coupled receptor GPR75 was reported to be activated by CCL5 (21). GPR75 does not appear to be expressed in the mouse ileum on the basis of our gene expression analysis, but other orphan G protein-coupled receptors are. Specifically, GPR48 and GPR160 appear to be highly expressed on the basis of our gene expression analysis. Both are without known ligands and it is worth investigating whether one of these receptors might respond to CCL5.

Conclusion. In the present work we demonstrate that the migration of lymphocytes into the ileal epithelium elicits an epithelial transcriptional response. For one of the epithelially induced transcripts, Acly, the induction was also confirmed at the protein level by immunocytochemistry. The ATP citrate lyase is a cytoplasmic enzyme that provides cytosolic acetyl-coA for lipid and cholesterol synthesis, which is needed for the establishment of the correct lipid composition of the enterocyte brush border membrane. Thus our results might suggest that the lymphocytes promotes lipid synthesis in the enterocyte by inducing expression of Acly. This can now be investigated in future experiments by analyzing brush border lipids from wild-type mice and RAG2 knockout mice.


   GRANTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
This work was supported by grants from The Alfred Nielsen and Wife's Foundation, The Danish Medical Research Council, The Novo Nordic Foundation, and the Lundbeck Foundation.


   ACKNOWLEDGMENTS
 
We thank Liselotte Laustsen for excellent technical assistance and Susanne Smed for valuable assistance with the Affymetrix GeneChip hybridizations and scannings.


   FOOTNOTES
 

Address for reprint requests and other correspondence: J. Olsen, Dept. of Cellular and Molecular Medicine, Univ. of Copenhagen, The Panum Institute Bldg. 6.4, Blegdamsvej 3, DK2200 Copenhagen N, Denmark (e-mail: jolsen{at}imbg.ku.dk)

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


   REFERENCES
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Bendtsen JD, Nielsen H, von Heijne G, Brunak S. Improved prediction of signal peptides: SignalP 3.0. J Mol Biol 340: 783–795, 2004.[CrossRef][Web of Science][Medline]
  2. Csillag C, Nielsen OH, Borup R, Nielsen FC, Olsen J. Clinical phenotype and gene expression profile in Crohn's disease. Am J Physiol Gastrointest Liver Physiol 292: G298–G304, 2007.[Abstract/Free Full Text]
  3. Cummins AG, Steele TW, Labrooy JT, Shearman DJ. Maturation of the rat small intestine at weaning: changes in epithelial cell kinetics, bacterial flora, and mucosal immune activity. Gut 29: 1672–1679, 1988.[Abstract/Free Full Text]
  4. Dalgaard P. Introductory Statistics with R. New York: Springer-Verlag, 2002.
  5. Danielsen EM, Hansen GH. Lipid rafts in epithelial brush borders: atypical membrane microdomains with specialized functions. Biochim Biophys Acta 1617: 1–9, 2003.[Medline]
  6. Devlin TM. Textbook of Biochemistry with Clinical Correlations. Hoboken, NJ: Wiley-Liss, 2006.
  7. Dieterich W, Esslinger B, Schuppan D. Pathomechanisms in celiac disease. Int Arch Allergy Immunol 132: 98–108, 2003.[CrossRef][Web of Science][Medline]
  8. Dignass AU, Podolsky DK. Interleukin 2 modulates intestinal epithelial cell function in vitro. Exp Cell Res 225: 422–429, 1996.[CrossRef][Web of Science][Medline]
  9. Eriksson L, Antti H, Gottfries J, Holmes E, Johansson E, Lindgren F, Long I, Lundstedt T, Trygg J, Wold S. Using chemometrics for navigating in the large data sets of genomics, proteomics, and metabonomics (gpm). Anal Bioanal Chem 380: 419–429, 2004.[CrossRef][Web of Science][Medline]
  10. Finn RD, Mistry J, Schuster-Bockler B, Griffiths-Jones S, Hollich V, Lassmann T, Moxon S, Marshall M, Khanna A, Durbin R, Eddy SR, Sonnhammer EL, Bateman A. Pfam: clans, web tools and services. Nucleic Acids Res 34: D247–D251, 2006.[Abstract/Free Full Text]
  11. Flint N, Cove FL, Evans GS. A low-temperature method for the isolation of small-intestinal epithelium along the crypt-villus axis. Biochem J 280: 331–334, 1991.[Web of Science][Medline]
  12. Forsberg G, Fahlgren A, Horstedt P, Hammarstrom S, Hernell O, Hammarstrom ML. Presence of bacteria and innate immunity of intestinal epithelium in childhood celiac disease. Am J Gastroenterol 99: 894–904, 2004.[CrossRef][Web of Science][Medline]
  13. Geladi P, Kowalski BR. Partial least-squares regression—a tutorial. Anal Chim Acta 185: 1–17, 1986.[CrossRef][Web of Science]
  14. Gellert M. V(D)J recombination: RAG proteins, repair factors, regulation. Annu Rev Biochem 71: 101–132, 2002.[CrossRef][Web of Science][Medline]
  15. Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S, Ellis B, Gautier L, Ge Y, Gentry J, Hornik K, Hothorn T, Huber W, Iacus S, Irizarry R, Leisch F, Li C, Maechler M, Rossini AJ, Sawitzki G, Smith C, Smyth G, Tierney L, Yang JY, Zhang J. Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 5: R80, 2004.[CrossRef][Medline]
  16. Green PH, Jabri B. Celiac disease. Annu Rev Med 57: 207–221, 2006.[CrossRef][Web of Science][Medline]
  17. Hansen MB, Skov L, Menné T, Olsen J. Gene transcripts as potential diagnostic markers for allergic contact dermatitis. Contact Dermatitis 53: 100–106, 2005.[CrossRef][Web of Science][Medline]
  18. Hayday A, Theodoridis E, Ramsburg E, Shires J. Intraepithelial lymphocytes: exploring the Third Way in immunology. Nat Immunol 2: 997–1003, 2001.[CrossRef][Web of Science][Medline]
  19. Henning SJ. Ontogeny of enzymes in the small intestine. Annu Rev Physiol 47: 231–245, 1985.[CrossRef][Web of Science][Medline]
  20. Herbst JJ, Sunshine P. Postnatal development of the small intestine of the rat. Changes in mucosal morphology at weaning. Pediatr Res 3: 27–33, 1969.[Web of Science][Medline]
  21. Ignatov A, Robert J, Gregory-Evans C, Schaller HC. RANTES stimulates Ca2+ mobilization and inositol trisphosphate (IP3) formation in cells transfected with G protein-coupled receptor 75. Br J Pharmacol 149: 490–497, 2006.[CrossRef][Web of Science][Medline]
  22. Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res 31: e15, 2003.[Abstract/Free Full Text]
  23. Kufer TA, Kremmer E, Banks DJ, Philpott DJ. Role for erbin in bacterial activation of Nod2. Infect Immun 74: 3115–3124, 2006.[Abstract/Free Full Text]
  24. Lievin-Le Moal V, Servin AL. The front line of enteric host defense against unwelcome intrusion of harmful microorganisms: mucins, antimicrobial peptides, and microbiota. Clin Microbiol Rev 19: 315–337, 2006.[Abstract/Free Full Text]
  25. MacDonald TT, Bajaj-Elliott M, Pender SL. T cells orchestrate intestinal mucosal shape and integrity. Immunol Today 20: 505–510, 1999.[CrossRef][Web of Science][Medline]
  26. MacDonald TT, Monteleone G, Pender SL. Recent developments in the immunology of inflammatory bowel disease. Scand J Immunol 51: 2–9, 2000.[CrossRef][Web of Science][Medline]
  27. Marsal J, Svensson M, Ericsson A, Iranpour AH, Carramolino L, Marquez G, Agace WW. Involvement of CCL25 (TECK) in the generation of the murine small-intestinal CD8alpha alpha+CD3+ intraepithelial lymphocyte compartment. Eur J Immunol 32: 3488–3497, 2002.[CrossRef][Web of Science][Medline]
  28. Rocha F, Musch MW, Lishanskiy L, Bookstein C, Sugi K, Xie Y, Chang EB. IFN-{gamma} downregulates expression of Na+/H+ exchangers NHE2 and NHE3 in rat intestine and human Caco-2/bbe cells. Am J Physiol Cell Physiol 280: C1224–C1232, 2001.[Abstract/Free Full Text]
  29. Rozen S, Skaletsky H. Primer3 on the WWW for general users and for biologist programmers. Methods Mol Biol 132: 365–386, 2000.[Medline]
  30. Sebastiani S, Albanesi C, De Pità O, Puddu P, Cavani A, Girolomoni G. The role of chemokines in allergic contact dermatitis. Arch Dermatol Res 293: 552–559, 2002.[Web of Science][Medline]
  31. Stegmann A, Hansen M, Wang Y, Larsen JB, Lund LR, Ritié L, Nicholson JK, Quistorff B, Simon-Assmann P, Troelsen JT, Olsen J. Metabolome, transcriptome and bioinformatic cis-element analyses point to HNF-4 as a central regulator of gene expression during enterocyte differentiation. Physiol Genomics 27: 141–155, 2006.[Abstract/Free Full Text]
  32. Trapani JA. Granzymes: a family of lymphocyte granule serine proteases. Genome Biol 2: REVIEWS3014, 2001.[Medline]
  33. Wang X, Seed B. A PCR primer bank for quantitative gene expression analysis. Nucleic Acids Res 31: e154, 2003.[Abstract/Free Full Text]
  34. Ward SG, Westwick J. Chemokines: understanding their role in T-lymphocyte biology. Biochem J 333: 457–470, 1998.[Web of Science][Medline]
  35. Wold S. Principal component analysis. Chemometr Intell Lab 2: 37–52, 1987.[CrossRef]
  36. Zhong S, Storch KF, Lipan O, Kao MC, Weitz CJ, Wong WH. GoSurfer: a graphical interactive tool for comparative analysis of large gene sets in Gene Ontology space. Appl Bioinformatics 3: 261–264, 2004.[CrossRef][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
294/6/G1335    most recent
00265.2007v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Schjoldager, K. T.-B. G.
Right arrow Articles by Olsen, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Schjoldager, K. T.-B. G.
Right arrow Articles by Olsen, J.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2008 by the American Physiological Society.