Contrary to rodents, in which functional changes leading to the acquisition of adult intestinal functions occur in a highly synchronized manner following weaning (1,2), the functional development of the human intestine is much less coordinated and mainly occurs during the fetal period (3,4). Indeed, the small intestine crypt-villus structure starts to develop under a proximal-distal gradient between 8 and 11 weeks of gestation, so that by 15 to 20 weeks, the morphology of the mucosa essentially resembles that of the newborn (3,5). Furthermore, several digestive enzymes associated with the enterocytic brush border, such as the disaccharidases and peptidases, and intestinal lipid processing and lipoprotein synthesis are present at mid-gestation at levels ranging from 70% to 100% of those of the adult intestine (5–8).
During the first 2 trimesters of gestation, the development of the human colon closely parallels that of the small intestine (5,6). The formation of the crypt-villus axis begins between 11 and 14 weeks of gestation under a distal-proximal gradient. Between 15 and 20 weeks, the colonic mucosa is characterized by a well-defined crypt-villus architecture similar to that of the small intestine (5,6). All brush-border digestive enzymes are present in the fetal colon, although their activities appear to be lower than in the small intestine (9,10). Similarly, the fetal colon at mid-gestation is found to be functional for lipid esterification, apolipoprotein synthesis, and lipoprotein assembly (11,12). Interestingly, the villi remain present in the colon until 30 weeks of gestation, at which point final maturation of the mucosa begins with the disappearance of villus structures and the loss of sucrase-isomaltase (13). Even though little is known about the physiological significance of the colonic villi between 15 and 30 weeks of gestation, it is generally assumed that these 2 morphologically related gut segments must perform similar physiological functions.
The degree of maturity of the intestinal mucosa during the second trimester is of crucial importance for physicians caring for immature infants. Indeed, medical advances have resulted in the increased survival of premature babies with an underdeveloped gut. However, the immature intestine is susceptible to inappropriate response to nutrients and microbes, leading to severe complications such as necrotizing enterocolitis, pneumatosis coli, and allergic colitis (14–21). Efforts toward the identification of potential treatments that could prevent the onset of pathogenesis (22–26) appear promising, although their particular effects on specific intestinal developing functions and segments are still incompletely understood.
A better knowledge of functional intestinal development is essential for the design of optimal treatments ensuring the survival and growth of premature infants. As a first step toward this ultimate goal, we have used the technology of microarray analysis conjugated with the Ingenuity Pathways Analysis (IPA) software (Ingenuity Systems, http://www.ingenuity.com) to systematically evaluate the gene-expression profiles of the developing human ileum and colon at mid-gestation. Our results show that >11% of the genes are differentially expressed in the ileum and colon. More important, IPA analysis of the differentially expressed genes revealed important functional differences between the 2 segments, emphasizing the immature status of the colon.
SUBJECTS AND METHODS
Three small intestines (ileum) and 3 colons from 4 fetuses ranging from 17 to 20 weeks of age (after fertilization) were obtained from normal elective pregnancy termination. No tissue was collected from cases associated with known abnormality or death. Only specimens obtained rapidly (<30 minutes) were used and stored at −80°C in RNAlater (Qiagen, Mississauga, Canada). Samples of normal adult ileum were obtained from Quebec Transplant (Quebec City, Canada). Samples of adult colon were obtained from patients who had undergone surgical treatment for colon adenocarcinoma. For each patient, samples from nondiseased areas (at least 10 cm distant from the lesion) corresponding to the resection margin were obtained from the Cooperative Human Tissue Network (Midwestern Division, Ohio State University, Columbus), which is funded by the National Cancer Institute. The project was in accordance with a protocol approved by the institutional human subject review board for the use of human material.
RNA Extraction and Amplification
RNA was extracted with TRIzol (Invitrogen, Burlington, Canada) according to the manufacturer's instructions and stored at −80°C. Quality of RNA was verified on agarose gel and by spectrophotometric assay. Poly(A) RNA was amplified using the TargetAmp 1-Round aRNA Amplification Kit (Epicentre Biotechnologies, Madison, WI).
Probe Preparation and Reference Pool
Probes were prepared as previously described (27). Briefly, first-strand cDNA synthesis from 1 μg of aRNA was primed with 3-μg random hexamers (Invitrogen) by heating at 70°C for 10 minutes, snap-cooling on ice for 30 seconds, and incubating at room temperature for an additional 5 to 10 minutes. Reverse transcription was performed in the presence of 500 μmol/L dATP, dCTP, and dGTP, 300 μmol/L 5-aminoallyl-dUTP (Sigma, Oakville, Canada) and 200 μmol/L dTTP, 1× first-strand buffer, 10 mmol/L dithiothreitol, and 400 U Superscript II (Invitrogen) in a volume of 40 μL at 42°C for 3 hours to overnight. cDNA was purified on QIAquick columns (Qiagen) according to the manufacturer's directions. For all of the microarray experiments, a standardized reference RNA pool (mixture of equimolar aliquots of RNA from the colon adenocarcinoma cell lines Caco-2/15 [at −1, 10, and 25 days of confluence] and HT29, the lung cancer cell line A549, and the ovarian cancer cell line SKOV3) (27) was labeled with Cy3 dye, whereas test samples were labeled with Cy5 dye. Coupling reactions were quenched by the addition of 35 μL of 0.1 mol/L sodium acetate, pH 5.2, and unincorporated dye was removed using QIAquick columns. The labeling efficiency was determined by analyzing the whole undiluted sample in a spectrophotometer using a 50-μL microcuvette (Beckman, Mississauga, Canada).
Hybridization and Image Processing
Human single spot microarrays comprising 19,208 human clones were used (University Health Network, Microarray Centre, Toronto, Canada; http://www.microarrays.ca). A total of 12 slides (3 independent biological samples of each intestinal segment; 2 slides per biological sample) were used. Slides were prehybridized in 0.1% bovine serum albumin, 5× saline-sodium citrate (SSC), 0.1% sodium dodecyl sulfate (SDS) for 45 minutes, washed by dipping in MilliQ water twice and 2-propanol once, and air-dried. Fluorescent cDNA probes were lyophilized and resuspended in 30 μL of hybridization buffer (50% formamide, 5× SSC, 0.1% SDS). To the combined Cy3 and Cy5 samples, 20-μg Cot1 DNA and 20-μg poly(A)+ DNA were added and the samples were denatured at 95°C for 5 minutes, followed by snap cooling on ice for 1 minute. Room-temperature probes were applied to a prehybridized array, covered with another slide rather than a glass coverslip, and placed in a humidified hybridization chamber (Corning, Lowell, MA). Hybridizations were carried out at 42°C for 16 to 20 hours, followed by 5-minute washings in 1× SSC, 0.2% SDS at 42°C, 0.1× SSC, 0.2% SDS at room temperature, and 0.1× SSC at room temperature, twice. Arrays were scanned using a ScanArray Express dual-color confocal laser scanner (Perkin Elmer, Waltham, MA). Data were collected in Cy3 and Cy5 channels and stored as paired TIFF images.
Spots were identified and local background was subtracted using the Quantarray software (Perkin-Elmer). A quality control (QC) filter was used to remove questionable array features. Two criteria for spot rejection were a spot shape that deviated from a circle and a low signal-to-noise ratio. Hybridization intensity data were normalized as previously described (27), and log2 ratio (gene/reference pool) was used for statistical analysis. Statistical significance was assessed by the Wilcoxon-Mann-Whitney test (P < 0.05), and hierarchical clustering analysis was performed using the TMEV 4.2 software (TIGR_MultiExperiment Viewer) (28). All software is available at the J. Craig Venter Institute Web site, http://www.jcvi.org/.
Functional analyses were performed with the use of IPA 7.6 software (Ingenuity Systems, http://www.ingenuity.com). The reference lists containing differentially expressed genes with gene identifiers and corresponding expression values were uploaded into the IPA application. Each gene identifier was mapped to its corresponding gene object in the Ingenuity Knowledge Base. IPA allows filtering to consider only functions and interactions in protein networks and/or pathways that are known for the defined species and tissue or cell line range. The filters were set for human species and cell lines for the core analyses. The Fisher exact test was used to calculate a P value determining the probability that each biological function assigned to that dataset would be the result of chance alone.
To focus on any specific gene and its pathway, another tool, canonical pathway analysis, is available. This uses KEGG annotated pathway maps to project the genes from a given pathway that are shown to be active. The most active and important pathways are summarized in Table 1.
Data Validation by Real-time qPCR
All of the reactions were performed in an Mx3000P real-time PCR system (Stratagene, Cedar Creek, TX) as previously described (29). Fluorescence data were acquired after each annealing step, and amplification efficiencies ranged from 91% to 105%. The 2× SYBRGreen qPCR Master mix (Stratagene) was mixed with the appropriate primers and high-quality sterile water. The genes investigated were aquaporin 3 (AQP3), potassium large-conductance calcium-activated channel, subfamily M, alpha member 1 (KCNMA1), lactase phlorizin hydrolase (LPH), ornithine aminotransferase (OAT), sodium channel, nonvoltage-gated 1, gamma (eNAC), solute carrier family 9 (sodium/hydrogen exchanger), member 3 (NHE3), and sucrase-isomaltase (SI). Supplemental Table 1 (http://links.lww.com/MPG/A39) summarizes the primers used in this study. Primers were generated using the primer formation software Primer3 (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) with attention given to avoiding primer-dimer formation by stringent use of the maximum 3′ self-complementarity function of the Primer3 program. Differences in gene expression between ileum and colon were calculated using R = (Etarget)ΔCttarget/(Ereference)ΔCtreference (30). Normalization was done with mannosidase α, class 1B (MAN1B1), but similar results were obtained using β-2 microglobulin (B2M) and 5-methyltetrahydrofolate-homocysteine methyltransferase (MTR) as reference genes (29). Statistical significance was evaluated by using unpaired t test, and P < 0.05 was considered to be significant.
To determine the level of functional similarity throughout the human intestine, we compared the gene-expression profiles of the developing ileum and colon using a cDNA microarray containing 19,208 cDNA clones. Among these, 18,387 were compared (genes present in all of the tested samples were included in the analysis). Wilcoxon-Mann-Whitney test (P < 0.05) applied to the dataset revealed that >11% of the genes (2071 genes, 1538 with known IDs) were differentially expressed. There were 1138 genes with increased and 933 with decreased expression in the ileum compared to colon (see Supplemental Table 2 [http://links.lww.com/MPG/A40] for the gene list). The original data discussed in this publication have been deposited in the National Center for Biotechnology Information's Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) and are accessible through GEO Series accession number GSE20983.
Unsupervised hierarchical clustering of all of the samples was applied based on the expression profile of genes relative to a reference pool. Results are displayed in a color-coded matrix (Fig. 1) in which samples are ordered on the horizontal axis and genes on the vertical axis on the basis of the similarity of their expression profiles. The clustering sample tree obtained distinguished between the 2 intestinal segments. Hierarchical clustering also sorted genes into 2 large clusters: genes upregulated in either the ileum or the colon, as exemplified in Figure 1. The complete list of the 2071 differentially expressed genes with ratios is provided in Supplemental Table 2 (http://links.lww.com/MPG/A40).
Differentially expressed genes upregulated in the ileum and colon were subjected to IPA analysis. We then performed a comparative analysis of the cellular, molecular, and physiological functions identified in each segment. Fifty-three of the IPA functions were found to be significantly expressed in the ileum and/or the colon at mid-gestation. The complete list of IPA functions and corresponding genes is provided in Supplemental Table 3 (http://links.lww.com/MPG/A41). To illustrate the data, we plotted the negative logarithm of P values calculated by IPA for each functional category found in the ileum against the negative logarithm of P values of the corresponding categories in the colon, allowing the visualization of which functions are more relevant for each segment (Fig. 2). More than 73% (39/53 functions) of the significant categories identified in the ileum were shared with the colon. These functions included various biological processes such as “cell-to-cell signaling and interaction,” “cellular development,” “cellular movement,” and “connective tissue development and function,” and also some important metabolic functions such as lipid metabolism. However, some important metabolic functions were significantly expressed exclusively in the ileum (12/53) such as “amino acid metabolism,” “vitamin and mineral metabolism,” and “cell-mediated immune response” (Fig. 2; Table 2), reflecting an apparent advanced level of development of the ileum. Surprisingly few functions were identified (2/53) that were significantly expressed exclusively in the colon (Fig. 2; Table 2 and Supplemental Table 3, http://links.lww.com/MPG/A41). Furthermore, 5 categories were found to be overexpressed in the colon compared to the ileum (Fig. 2) such as “cell cycle,” “cell death,” “cellular assembly and organization,” and “cell morphology” (Table 2) consistent with the apparent relative functional immaturity of the colon in comparison with the ileum at mid-gestation.
To further investigate this possibility, we performed another comparative analysis by IPA using canonical pathways. Seventy-nine of 282 of the IPA canonical pathways were found to be significantly expressed in the ileum and/or the colon at mid-gestation. The complete list of IPA canonical pathways and corresponding genes is provided in Supplemental Table 4 (http://links.lww.com/MPG/A38). As illustrated in Figure 3, this approach led to the identification of several canonical pathways exclusive to each segment. Indeed, among the 79 canonical pathways significantly expressed, 22 of them were found to be exclusively represented in the ileum, including “arginine and proline metabolism,” “phenylalanine metabolism,” and “tyrosine metabolism” (Table 1). Among the genes involved in arginine metabolism, we identified 2 genes, OAT and arginosuccinate lyase (ASL), that were overexpressed by the developing small intestine. The net production of ornithine by enterocytes is a prerequisite for arginine biosynthesis. By qPCR, we found that only OAT was significantly overexpressed by the fetal ileum (Table 3). In the adult, we observed that the expression of ASL was higher in both segments, whereas the difference between ileum and colon remained nonsignificant (Table 3). Furthermore, 53 canonical pathways were exclusively represented in the colon (Fig. 3). These pathways included “actin cytoskeleton signaling,” “Myc mediated apoptosis,” “14-3-3–mediated signaling,” and “AMPK signaling” (Table 1), reflecting major active processes playing a role in the morphological development of the colon. Surprisingly, only 4 canonical pathways (Fig. 3) were expressed at significant levels in both intestinal segments, for instance the pathway related to “ephrin receptor signaling” (Table 1).
These results identify the major differences in terms of overall pathways governing the development of these 2 intestinal segments at mid-gestation. We selected a number of specific biochemical markers expressed by the adult small and large intestines to molecularly define differences between developing and mature segments. Sucrase-isomaltase and LPH are hydrolases specific to the small intestine, whereas NHE3 and DRA are exchangers expressed by both segments. Aquaporin 3, eNAC, and KCNMA1 are transmembrane channels highly enriched in the adult colon, accurately representing the functions attributed to this segment. Expression of these markers in fetal and adult intestinal tissues was evaluated by qPCR, and relative amounts found in each segment are presented in Table 3. As expected, fetal and adult small intestines express similar amounts of SI and LPH, and although the fetal colon also expresses these 2 markers, they are present in much smaller quantities. However, KCNMA1 and DRA were already expressed by the 2 fetal segments, whereas none of the other genes encoding exchangers and channels were found to be present in both developing tissues. In the adult, all of the exchangers and channels were highly expressed in the colon, whereas the expression of DRA decreased in parallel with maturation of the small intestine. All together, these results point out the major differences between these 2 intestinal segments and reflect the immaturity of the colon at the functional level when compared with the ileum.
The incidence of preterm deliveries has increased during the past few decades. Providing nutrients to these infants is essential because prenatal nutrient stores are low and infants have a high metabolic rate (31). The immature nature of the gastrointestinal tract limits the route of nutrient delivery, however, and preterm infants commonly survive on parenteral nutrition, which is associated with major complications, including an increased risk of sepsis, intestinal atrophy, and an increased susceptibility to inflammatory stimuli (32). Therefore, it is urgent that our understanding of the human fetal gastrointestinal tract be improved, and identifying key differences between the small intestine and the colon at mid-gestation should contribute to the development of appropriate therapeutic approaches in the field of neonatal care and nutrition.
Analysis of a gene-expression profile revealed that approximately 60% of the top significant cellular, molecular, and physiological functions of the small intestine such as “cellular growth and cell proliferation,” “small molecule biochemistry,” “carbohydrate metabolism,” and “lipid metabolism” were shared with the colon, an observation consistent with the apparent similar morphological and functional characteristics of the small and large intestine throughout the first 2 trimesters of gestation (5,6,9–12). For example, it has already been established that the human fetal small and large intestines have the capacity to elaborate lipoproteins for the transport of newly synthesized lipids (12), the colon having a limited capacity compared to the small intestine. Other gene-expression profile studies in rodents have also demonstrated the gene enrichment in lipid metabolism during intestinal cell maturation (33,34).
As first suggested by hierarchical clustering analysis, the gene-expression profile established by IPA analysis revealed a surprisingly high level of transcriptional dissimilarity between the developing ileum and colon in 35% of the top significant cellular, molecular, and physiological functions and 85% of the canonical pathways. These data suggest that despite the fact that the fetal small intestine and colon share morphological and functional characteristics during the first 2 trimesters, these 2 intestinal segments are distinct organs.
For the small intestine, 3 specifically expressed functions were particularly obvious: “amino acid metabolism,” “vitamin and mineral metabolism,” and “lymphoid tissue structure and development.” The fact that these functions were exclusively detected in the ileum confirms that many of the primary functions of this segment were already established at this period of gestation (5–8). In the same direction, analysis of the canonical pathways (Table 1) revealed that the metabolism of amino acids, particularly arginine/proline, phenylalanine, and histidine, is exclusively represented in the small intestine. Consistent with this observation, we found that OAT, an enzyme involved in arginine synthesis in enterocytes, is overexpressed in the small intestine compared with the colon (see Table 3 and Supplemental Table 2, http://links.lww.com/MPG/A40). The importance of this metabolism has been previously demonstrated in premature infants in whom a severe deficiency in arginine results in several metabolic complications (35,36). A recent study found that the human neonatal small intestine expresses arginine-synthesizing enzymes such as OAT (37) and this expression is conserved in the adult, thus corroborating our results. We have shown that this expression pattern for OAT is exclusive to the ileum even in the adult intestine. In the porcine fetal small intestine, both metabolic and molecular studies have indicated that underdevelopment of intestinal arginine synthesis may be responsible for hypoargininemia in preterm neonates (38,39). Enterocytes are responsible for the majority of endogenous citrulline and arginine synthesis, and this metabolic pathway is crucial for maintaining arginine homeostasis in both the fetus and neonate (40,41). Thus, it appears that the small intestine has already established such basic mechanisms as those related to amino acid metabolism, vitamin and mineral metabolism, and the ability to respond to an inflammatory signal, unlike the developing colon.
IPA analysis on the data produced from the colon at mid-gestation did not identify tissue- or segment-specific cell and molecular functions. Predominantly expressed functions in the colon included “cell cycle,” “cellular assembly and organization,” “cell morphology,” and “cell death,” suggesting that primary biological functions occurring in the colon at this developmental stage are related more to organ development and structural morphogenesis than functional metabolism. Consistent with this suggestion, canonical pathways represented exclusively in the colon were “actin cytoskeleton signaling,” “Myc-mediated apoptosis,” “14-3-3–mediated signaling,” and “AMPK signaling” (Table 1). These results uncovered marked functional differences between the human small and large intestine at mid-gestation. It is worth noting that our results with colonic functional markers (Table 3) indicated that the colon is still governed by morphological concerns and does not appear to have initiated pathways related to functional maturation when compared with the adult, whereas the small intestine is already relatively functionally mature.
Despite many similarities throughout the first 2 trimesters of gestation, the identification of basic differences between the small intestine and colon is not without precedent. For instance, a complementary pattern of expression of the 2 extracellular matrix molecules tenascin-C and SPARC/BM40/osteonectin has been shown at the epithelial–mesenchymal interface of the small intestine and colon, respectively (42,43). The distinct expression of Bcl-2 homologs along the length of the developing fetal intestine (44) is another example. Furthermore, the demonstration that cell proliferation and expression of functional markers in human fetal intestinal and colonic explants were differentially affected by hormones and growth factors (45–47) suggests that these 2 intestinal segments are under different regulatory mechanisms (3,7). It has been previously demonstrated that after weaning the migration of lymphocytes into the mouse ileum could affect epithelial gene expression (48). Whether the human colonic mucosa must be specifically exposed to lymphocytes to initiate and/or achieve maturation needs to be further explored.
In summary, our data revealed that the developing ileum was similar, morphologically and functionally, to the adult ileum, but that the same cannot be claimed for the developing colon. In fact, we showed that although the fetal colon expressed a variety of biochemical intestinal markers it could not be considered fetal ileum and even less to be mature colon, but more as an intestinal segment in a transitional state. More important, the present study provides new insight into our knowledge of the basic molecular and cellular mechanisms governing the morphological and functional development of the human gastrointestinal tract, particularly regarding the major differences in terms of maturity in the small intestine and colon. Indeed, our data disclose an unexpected level of immaturity of the colon in mid-gestation fetuses that needs to be considered for the design of optimal treatments ensuring the survival and growth of premature infants.
The authors thank Nuria Basora and Elizabeth Herring for reviewing the manuscript. We also acknowledge the excellent collaboration of Drs A. Poulin and F. Jacot of the Département de la Santé Communautaire of the Centre Hospitalier Universitaire de Sherbrooke in providing tissue specimens for this study. We also acknowledge the excellent collaboration of Québec Transplant in obtaining fresh specimens of normal adult small intestine and of the Cooperative Human Tissue Network, funded by the National Cancer Institute, in obtaining human colon samples.
1. Henning SJ. Postnatal development: coordination of feeding, digestion, and metabolism. Am J Physiol 1981; 241:G199–G214.
2. Ménard D, Calvert R. Fetal and Postnatal Development of the Small and Large Intestine: Patterns and Regulation. Boca Raton, FL: CRC Press; 1991.
3. Menard D. Functional development of the human gastrointestinal tract: hormone- and growth factor-mediated regulatory mechanisms. Can J Gastroenterol 2004; 18:39–44.
4. Montgomery RK, Mulberg AE, Grand RJ. Development of the human gastrointestinal tract: twenty years of progress. Gastroenterology 1999; 116:702–731.
5. Ménard D. Growth-promoting Factors and the Development of the Human Gut. New York: Raven Press; 1989.
6. Ménard D, Beaulieu JF. Human Intestinal Brush Border Membrane Hydrolases. Norwell: Kluwer Academic; 1994.
7. Levy E, Menard D. Developmental aspects of lipid and lipoprotein synthesis and secretion in human gut. Microsc Res Tech 2000; 49:363–373.
8. Auricchio S, Stellato A, De Vizia B. Development of brush border peptidases in human and rat small intestine during fetal and neonatal life. Pediatr Res 1981; 15:991–995.
9. Menard D, Pothier P. Differential distribution of digestive enzymes in isolated epithelial cells from developing human fetal small intestine and colon. J Pediatr Gastroenterol Nutr 1987; 6:509–516.
10. Sebastio G, Hunziker W, O'Neill B, et al
. The biosynthesis of intestinal sucrase-isomaltase in human embryo is most likely controlled at the level of transcription. Biochem Biophys Res Commun 1987; 149:830–839.
11. Basque JR, Levy E, Beaulieu JF, et al
. Apolipoproteins in human fetal colon: immunolocalization, biogenesis, and hormonal regulation. J Cell Biochem 1998; 70:354–365.
12. Levy E, Loirdighi N, Thibault L, et al
. Lipid processing and lipoprotein synthesis by the developing human fetal colon. Am J Physiol 1996; 270:G813–G820.
13. Raul F, Lacroix B, Aprahamian M. Longitudinal distribution of brush border hydrolases and morphological maturation in the intestine of the preterm infant. Early Hum Dev 1986; 13:225–234.
14. Caicedo RA, Schanler RJ, Li N, et al
. The developing intestinal ecosystem: implications for the neonate. Pediatr Res 2005; 58:625–628.
15. Kliegman RM. The relationship of neonatal feeding practices and the pathogenesis and prevention of necrotizing enterocolitis. Pediatrics 2003; 111:671–672.
16. Neu J, Caicedo RA, Schanler RJ, et al
. Neonatal necrotizing enterocolitis: an update. Acta Paediatr Suppl 2005; 94:100–105.
17. Fell JM. Neonatal inflammatory intestinal diseases: necrotising enterocolitis and allergic colitis. Early Hum Dev 2005; 81:117–122.
18. Nanthakumar NN, Fusunyan RD, Sanderson I, et al
. Inflammation in the developing human intestine: a possible pathophysiologic contribution to necrotizing enterocolitis. Proc Natl Acad Sci U S A 2000; 97:6043–6048.
19. Hoehn T, Stover B, Buhrer C. Colonic pneumatosis intestinalis in preterm infants: different to necrotising enterocolitis with a more benign course? Eur J Pediatr 2001; 160:369–371.
20. Leonidas JC, Hall RT. Neonatal pneumatosis coli: a mild form of neonatal necrotizing enterocolitis. J Pediatr 1976; 89:456–459.
21. Travadi JN, Patole SK, Gardiner K. Pneumatosis coli, a benign form of necrotising enterocolitis. Indian Pediatr 2003; 40:349–351.
22. Nanthakumar NN, Young C, Ko JS, et al
. Glucocorticoid responsiveness in developing human intestine: possible role in prevention of necrotizing enterocolitis. Am J Physiol Gastrointest Liver Physiol 2005; 288:G85–G92.
23. Amin HJ, Zamora SA, McMillan DD, et al
. Arginine supplementation prevents necrotizing enterocolitis in the premature infant. J Pediatr 2002; 140:425–431.
24. Di Lorenzo M, Bass J, Krantis A. Use of L-arginine in the treatment of experimental necrotizing enterocolitis. J Pediatr Surg 1995; 30:235–240.
25. Clark JA, Doelle SM, Halpern MD, et al
. Intestinal barrier failure during experimental necrotizing enterocolitis: protective effect of EGF treatment. Am J Physiol Gastrointest Liver Physiol 2006; 291:G938–G949.
26. Nair RR, Warner BB, Warner BW. Role of epidermal growth factor and other growth factors in the prevention of necrotizing enterocolitis. Semin Perinatol 2008; 32:107–113.
27. Tremblay E, Auclair J, Delvin E, et al
. Gene expression profiles of normal proliferating and differentiating human intestinal epithelial cells: a comparison with the Caco-2 cell model. J Cell Biochem 2006; 99:1175–1186.
28. Saeed AI, Sharov V, White J, et al
. TM4: a free, open-source system for microarray data management and analysis. Biotechniques 2003; 34:374–378.
29. Dydensborg AB, Herring E, Auclair J, et al
. Normalizing genes for quantitative RT-PCR in differentiating human intestinal epithelial cells and adenocarcinomas of the colon. Am J Physiol Gastrointest Liver Physiol 2006; 290:G1067–G1074.
30. Pfaffl MW, Tichopad A, Prgomet C, et al
. Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper–Excel-based tool using pair-wise correlations. Biotechnol Lett 2004; 26:509–515.
31. Commare CE, Tappenden KA. Development of the infant intestine: implications for nutrition support. Nutr Clin Pract 2007; 22:159–173.
32. Neu J. Gastrointestinal development and meeting the nutritional needs of premature infants. Am J Clin Nutr 2007; 85:629S–634S.
33. Mariadason JM, Nicholas C, L'Italien KE, et al
. Gene expression profiling of intestinal epithelial cell maturation along the crypt-villus axis. Gastroenterology 2005; 128:1081–1088.
34. Stegmann A, Hansen M, Wang Y, et al
. Metabolome, transcriptome, and bioinformatic cis-element analyses point to HNF-4 as a central regulator of gene expression during enterocyte differentiation. Physiol Genomics 2006; 27:141–155.
35. Zamora SA, Amin HJ, McMillan DD, et al
. Plasma L-arginine concentrations in premature infants with necrotizing enterocolitis. J Pediatr 1997; 131:226–232.
36. Becker RM, Wu G, Galanko JA, et al
. Reduced serum amino acid concentrations in infants with necrotizing enterocolitis. J Pediatr 2000; 137:785–793.
37. Kohler ES, Sankaranarayanan S, van Ginneken CJ, et al
. The human neonatal small intestine has the potential for arginine synthesis; developmental changes in the expression of arginine-synthesizing and -catabolizing enzymes. BMC Dev Biol 2008; 8:107.
38. Wu G. Synthesis of citrulline and arginine from proline in enterocytes of postnatal pigs. Am J Physiol 1997; 272:G1382–G1390.
39. Wu G, Meininger CJ, Knabe DA, et al
. Arginine nutrition in development, health and disease. Curr Opin Clin Nutr Metab Care 2000; 3:59–66.
40. Wu G, Morris SM Jr. Arginine metabolism: nitric oxide and beyond. Biochem J 1998; 336(pt 1):1–17.
41. Bertolo RF, Brunton JA, Pencharz PB, et al
. Arginine, ornithine, and proline interconversion is dependent on small intestinal metabolism in neonatal pigs. Am J Physiol Endocrinol Metab 2003; 284:E915–E922.
42. Desloges N, Simoneau A, Jutras S, et al
. Tenascin may not be required for intestinal villus development. Int J Dev Biol 1994; 38:737–739.
43. Lussier C, Sodek J, Beaulieu JF. Expression of SPARC/osteonectin/BM4O in the human gut: predominance in the stroma of the remodeling distal intestine. J Cell Biochem 2001; 81:463–476.
44. Vachon PH, Cardin E, Harnois C, et al
. Early acquisition of bowel segment-specific Bcl-2 homolog expression profiles during development of the human ileum and colon. Histol Histopathol 2001; 16:497–510.
45. Arsenault P, Menard D. Influence of hydrocortisone on human fetal small intestine in organ culture. J Pediatr Gastroenterol Nutr 1985; 4:893–901.
46. Menard D, Corriveau L, Arsenault P. Differential effects of epidermal growth factor and hydrocortisone in human fetal colon. J Pediatr Gastroenterol Nutr 1990; 10:13–20.
47. Menard D, Arsenault P, Pothier P. Biologic effects of epidermal growth factor in human fetal jejunum. Gastroenterology 1988; 94:656–663.
48. Schjoldager KT, Maltesen HR, Balmer S, et al
. Cellular cross talk in the small intestinal mucosa: postnatal lymphocytic immigration elicits a specific epithelial transcriptional response. Am J Physiol Gastrointest Liver Physiol 2008; 294:G1335–G1343.
Keywords:Copyright 2011 by ESPGHAN and NASPGHAN
colon; digestive functions; gene expression; mid-gestation; premature; small intestine