INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE SMALL INTESTINE has a remarkable capacity to adapt to the loss of small bowel surface area. This response occurs after surgical resection due to vascular injury or abdominal trauma or because of loss of functional intestine resulting from a variety of small bowel disorders such as Crohn's disease or radiation enteritis. The morbidity resulting from short bowel syndrome is a major source of health care costs, particularly when long-term total parenteral nutrition is required to sustain life (4, 5). The critical importance of the adaptive response in facilitating discontinuation from total parenteral nutritional support underscores the need for better therapies to hasten and enhance this process. These treatments might also prove effective in a variety of other intestinal disorders, by enhancing small bowel recovery after injury (e.g., ischemia or caused by inflammation in Crohn's disease, etc.).

In rodent models of small bowel adaptation after massive resection, the remnant intestine exhibits increased crypt cell proliferation, resulting in enhanced villus height and crypt depth followed by increased absorption of fluid, electrolytes, and nutrients. The adaptive response is quite complex and includes alterations in processes as diverse as cellular proliferation, apoptosis, epithelial migration, and epithelial cell gene expression (6, 14, 20, 21, 24). This process is also temporally regulated. As early as 16 h after resection, there is an increase in crypt cell proliferation rate (27), and by 48 h enhanced expression of a variety of enterocyte-specific genes is seen (6, 20, 21). This is followed by villus epithelial hyperplasia and increased enterocyte migration rate by 1-2 wk (7). Much progress has been made in identifying growth factors and peptides that initiate enhanced crypt cell proliferation (such as epidermal growth factor, keratinocyte growth factor, and glucagon-like peptide-2) (8, 9, 22, 25). However, the intracellular pathways that mediate growth factor effects in the crypt cell and the genetic regulation of other facets of this complex response have not yet been identified.

To begin to elucidate the molecular mechanisms that regulate the adaptive response, we have used subtractive hybridization techniques to isolate novel cDNAs that are specifically regulated during this process (6, 21). In this study, we report the cloning and characterization of a novel rat cDNA [termed "ileal remnant repressed" (IRR)-219] that encodes a protein homologous to a recently identified human IgG Fc binding protein (IgGFcBP) (13), a mucinlike intestinal protein that specifically binds the Fc portion of IgG. IRR-219 mRNA is only expressed in goblet cells, and it is differentially regulated during early gut adaptation (at 48 h after resection) in small bowel and colon. These findings suggest that the goblet cell may play distinctive roles in small bowel and colonic adaptation.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Rat model of intestinal adaptation. For intestinal resection and control surgery, male Sprague-Dawley rats (225-260 g) were subjected to either 70% small intestinal resection or transection control surgery as previously described (6, 20, 21). For 70% small bowel resection, the bowel was divided 5 cm distal to the ligament of Treitz and 15 cm proximal to the ileocecal valve, and the remnant jejunum was anastomosed to the ileum after removal of the intervening intestine. In the transected group, the bowel was divided 5 cm distal to the ligament of Treitz and then reanastomosed end to end. Rats were killed and tissues harvested at 48 h after surgery (n = 7 per experimental and control group), or at 2, 4, 8, 16, 24, and 168 (1 wk) h postoperatively (n = 3 per group).

Tissue harvesting. Tissues were frozen in liquid nitrogen for RNA preparation or were fixed in 4% paraformaldehyde in 1× phosphate-buffered saline for in situ hybridization analysis. For regional analysis of IRR-219 expression, segments of intestine were isolated from duodenum, jejunum, ileum, and proximal and distal colon. For analysis of IRR-219 expression during adaptation, the adaptive intestine was divided into two segments. The portion "proximal" to the anastomosis consisted of duodenum and 5.0 cm of jejunum, and the portion "distal" to the anastomosis consisted of the remnant ileum. For tissue-specific expression analyses, liver, kidney, lung, heart, brain, and testes were harvested from 8-wk-old male Sprague-Dawley rats. To examine the developmental regulation of IRR-219 expression in the gut, segments were harvested on fetal days 13 and 14 (whole intestine) and days 18, 20, and 21 (jejunum) and postnatal weeks 2, 3, 4, 6, and 8 [proximal and midintestine (representing jejunum) or distal intestine (representing ileum)].

Cloning, sequencing, and database analysis of IRR-219 cDNA. The IRR-219 cDNA was cloned by subtractive hybridization techniques as previously described (6). Briefly, an adaptive, size-selected cDNA library was synthesized from rat ileal mRNA isolated from tissues harvested 48 h after small bowel resection. High specific activity subtracted cDNA probes were generated as follows. Ileal poly(A)+ RNA from animals subjected to 70% intestinal resection was labeled by reverse transcription. Ileal poly(A)+ RNA from the transected control animals was biotinylated, and subtraction was performed by hybridizing radiolabeled adaptive probe with the biotinylated RNA (1:15 ratio, hybridized at 65°C for 48 h). Biotinylated RNA (free and bound to cDNA) was removed by extraction with streptavidin. Two rounds of subtraction were performed. The subtracted probe and a control probe [created by subtracting radiolabeled cDNA from transected animal (control) ileum with biotinylated ileal adaptive mRNA] were used to screen the adaptive cDNA library. Clones that exhibited differential hybridization to the subtracted compared with the control probe were selected for sequence analysis by the Sanger method using Sequenase 2.0 as described previously (6). The nonredundant Genbank/European Molecular Biology Laboratory databases were searched on the National Center for Biotechnology Information server, using the BLASTn, TBLASTn, and BLASTx algorithms.

Northern blot hybridization. Total RNA was prepared from various intestinal segments and extraintestinal tissues, and Northern blot hybridization was performed as previously described (6, 20, 27). IEC-18 and Caco-2 cell lines were grown as described (21) and harvested at 3 days after plating (IEC-18) or 21 days after confluence (Caco-2) for total RNA isolation. Briefly, Northern blots containing 20 µg of total RNA were incubated with IRR-219 cDNA probe, rat intestinal trefoil factor (RITF) cDNA probe (gift of Daniel Podolsky, Massachusetts General Hospital; Ref. 23), or rat -actin cDNA probe, radiolabeled by the random oligonucleotide primer method. Blots were stringently washed at 55°C in 0.1× sodium chloride sodium citrate (SSC) and exposed for various times to Kodak BioMax MS films and intensifying screens. The relative abundance of each mRNA per sample was determined by NIH Image 1.55 analysis of digitized images obtained with a UMAX PS-2400X scanner using UMAX Magicscan version 1.2. Results were normalized for differences in RNA loading by digitized image analysis of 18S rRNA bands as visualized by ethidium bromide staining of RNA gels.

In situ hybridization. Tissues were subjected to in situ hybridization analyses as previously described (10, 11). The 1,700-bp IRR-219 cDNA was subcloned into the Bluescript II SK vector, linearized by Hpa I and transcribed to produce a 341-bp antisense probe, or cut with Nde 1 and hybridized with a 506-bp sense control probe, labeled with ³⁵S-UTP. Paraformaldehyde-fixed frozen cryostat sections (6-8 µm thick) were hybridized overnight to the radiolabeled antisense or sense control RNA probes (5 × 10⁶ cpm/ml of hybridization solution) and then washed stringently in 0.1× SSC at 55°C. Slides were dipped in Kodak NTB-2 photoemulsion and exposed at 4°C for 6 days.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sequence analysis of clone IRR-219 reveals homology to a human IgG Fc binding protein. As previously described, subtractive and differential hybridization cloning techniques were used to identify genes that are differentially regulated during intestinal adaptation (6). IRR-219, a cDNA demonstrating suppressed expression in adaptation, was identified after three rounds of hybridization screening of an adaptive ileal cDNA library derived from ileal RNA harvested 48 h after 70% resection. The partial nucleotide and amino acid sequences are indicated in Fig. 1. Nucleic acid homology searches revealed 80% sequence identity to the 3'-end of a human IgG Fc binding protein (known as IgGFcBP). IgGFcBP is a very large (encoded by a 17-kb mRNA), mucinlike protein that binds the Fc portion of IgG (13). It is secreted with mucin and is expressed only in intestine and placenta. Protein database analysis also revealed homology of IRR-219 to the rat and human mucins MUC2 and human mucin MUC5b (~30% amino acid sequence identity; data not shown). Northern blots of jejunal, ileal, or colonic total RNA hybridized with the IRR-219 cDNA demonstrated a single band specifying a large mRNA of >10.0 kb.

View larger version (52K): [in this window] [in a new window]   Fig. 1.   Partial nucleotide (A) and amino acid (B) sequence of IRR-219 and homology to the human IgG Fc binding protein (IgGFcBP). The nonredundant Genbank/European Molecular Biology Laboratory databases were searched on the National Center for Biotechnology Information server, using the BLASTn, TBLASTn, and BLASTx algorithms. A: 219, ileal remnant repressed (IRR)-219; IgG, IgGFcBP. B: 219, IRR-219; IgGFc, IgGFcBP; +, conserved amino acids.

Expression of IRR-219 mRNA is restricted to the intestine and colon. The tissue-specific expression of IRR-219 was examined by Northern blot hybridization (Fig. 2). IRR-219 mRNA was found in small intestine and colon but was undetectable in stomach, liver, kidney, brain, lung, heart, and testes harvested from 8-wk-old rats. The regional intestinal expression of IRR-219 was examined by isolating total RNA from intestinal segments including duodenum, jejunum, ileum, and proximal and distal colon. IRR-219 mRNA expression was most abundant in the ileum and jejunum, with lower levels in the duodenum and proximal and distal colon (Fig. 2A). However, unlike the human IgGFcBP (13), IRR-219 is not expressed in rat placenta (Fig. 2B). In addition, two epithelial cell lines, IEC-18, a proliferating cryptlike cell line derived from rat ileum, and human Caco-2, a colon cancer cell line that exhibits enterocytic characteristics on achieving confluence, do not express IRR-219.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   IRR-219 mRNA is expressed in the small bowel and colon. A: total RNA was prepared from 8-wk-old rat duodenum (D), jejunum (J), ileum (I), proximal colon (PC), distal colon (DC), liver (L), kidney (K), lung (P), heart (H), brain (B), and testes (T). Stomach not shown. Northern blot hybridization was performed using a radiolabeled IRR-219 cDNA probe, as described in METHODS. B: total RNA was prepared from 11-day rat placenta (lane 1), rat ileum (lane 2), IEC-18 cells harvested at 3 days after plating (lane 3) and Caco-2 cells harvested at 21 days after confluence (lane 4).

IRR-219 expression is regulated during small bowel and colonic adaptation. To assess the regulation of IRR-219 mRNA expression during gut adaptation, Northern blot hybridization analyses were performed. Because IRR-219 mRNA is expressed in small bowel and colon, both tissues were analyzed for the response after small bowel resection (Fig. 3). Cloning experiments were designed to isolate cDNAs that are regulated during "early" gut adaptation (e.g., at 48 h after resection), so this time point was the first examined by Northern blot hybridization. At 48 h after resection surgery, the crypt cell proliferation rate is increased (27), although villus height and crypt depth have not yet changed. The adaptive ileum also exhibits increased expression of a variety of enterocytic genes, including the immediate-early gene, PC4/TIS7, a sensitive marker of early adaptation (6, 20, 21). Intestinal adaptation was documented in the tissues used for these studies by showing that expression of PC4/TIS7 was increased in adaptive compared with sham-resected control ileum (21).

View larger version (32K): [in this window] [in a new window]   Fig. 3.   Regulation of IRR-219 expression during intestinal adaptation after small bowel resection. Rats were subjected to 70% small intestinal resection or sham-control surgery and killed at 48 h after surgery. Duodenal-jejunal or ileal segments (n = 7/group) or colonic segments (n = 3/group) were removed for total RNA preparation, and Northern blot hybridization was performed as described in METHODS. A: representative Northern blot hybridization. B: relative IRR-219 mRNA levels in sham-resected (SHR) vs. resected (RE) small bowel or colon. Data are expressed as mean densitometric units per group. *P < 0.04 vs. SHR.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Temporal regulation of IRR-219 expression during small bowel adaptation. Rats were subjected to intestinal resection or sham control surgery and killed at 2, 4, 8, 16, 24, 48 and 168 h (1 wk) postoperatively (n = 3/group). IRR-219 mRNA expression was quantitated by dot blot hybridization of total RNA prepared from each segment at each time point. Data are expressed as the ratio of relative mRNA levels from resected/sham-resected rats.

Ontogeny of IRR-219 mRNA expression in gut. The developmental regulation of IRR-219 expression was examined in whole intestine harvested from fetuses on gestation days 13 and 14 and from jejunum on fetal days 18, 20, and 21. Segments of proximal and midintestine (jejunum) and distal intestine (ileum) were also isolated at postnatal weeks 2, 3, 4, 6, and 8. IRR-219 mRNA was first detected from fetal day 18 onward in jejunum (data not shown) and continued to exhibit low levels of expression in small bowel at postnatal weeks 2, 3, and 4 (Fig. 5). By week 6, mRNA levels were markedly upregulated and abundant expression of IRR-219 mRNA was noted.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   The expression of IRR-219 mRNA is developmentally regulated. Representative autoradiogram from Northern blot hybridization of total RNA prepared from intestinal segments including duodenum (DU) proximal (P, representing proximal jejunum), mid (M, representing midjejunum), and distal (D, representing ileum) intestine, harvested from 2-, 3-, 4-, 6-, and 8-wk-old rats. Northern blots were prepared and probed as described in Fig. 2.

IRR-219 mRNA is detected exclusively in intestinal goblet cells. To identify the intestinal cells that express IRR-219, in situ hybridization analyses were performed using neonatal and adult rat intestine (Fig. 6) from rats at fetal day 18 and at postnatal weeks 2, 4, and 8. Beginning with the onset of villus morphogenesis, IRR-219 mRNA was expressed exclusively in goblet cells (Fig. 6). Relatively low IRR-219 mRNA levels are expressed in goblet cells during the suckling and weaning period (Fig. 6, B and C), and a marked increase in expression occurs in the adult intestine (Fig. 6D), consistent with the Northern blot hybridization results shown in Fig. 5. mRNA expression was not detected in enterocytes, Paneth cells, or enteroendocrine cells in the lamina propria or muscle layers. A sense control probe revealed no specific signal (Fig. 7A), and IRR-219 mRNA was not present in other tissues such as liver (Fig. 7B). Furthermore, during small bowel adaptation, the decrease in IRR-219 mRNA expression occurred specifically in ileal goblet cells (Fig. 8).

View larger version (119K): [in this window] [in a new window]   Fig. 6.   IRR-219 mRNA is expressed in goblet cells. In situ hybridization analysis was performed on sections of fetal day 18 (A), and postnatal week 2 (B), 4 (C), and 8 (D) proximal ileum to determine the cell-specific expression of IRR-219 mRNA. Sections are viewed by darkfield microscopy. A: IRR-219 mRNA is present in goblet cells on fetal day 18, coincident with villus morphogenesis. Multiple small lumens are present, and small thumblike villi are beginning to emerge. Arrows depict scattered goblet cells with silver grains, indicating IRR-219 mRNA expression. B: ileum from 2-wk-old rat pups. Villi and crypts are well-formed. Many goblet cells are present and scattered silver grains overly these cells. Two representative goblet cells are depicted by the arrows. C: ileum from a 4-wk-old rat pup. Two representative goblet cells that contain white grains are depicted by arrows. D: ileum from 8-wk-old rat. IRR-219 is abundantly expressed in goblet cells (arrows). Magnification, ×200.

View larger version (112K): [in this window] [in a new window]   Fig. 7.   Specificity of IRR-219 expression by in situ hybridization analyses. To determine the specificity of the signal detected by the antisense IRR-219 probe, a sense IRR-219 cRNA was hybridized to sections of adult rat ileum, as depicted in A. Goblet cells are devoid of white grains. White cells in the lamina propria are macrophages that reflect light under darkfield microscopy. The antisense IRR-219 probe was also hybridized to liver (B), which does not express IRR-219, as shown in Fig. 2. No specific signal was seen on either tissue. (A, ×400; B, ×200).

View larger version (138K): [in this window] [in a new window]   Fig. 8.   In situ hybridization analysis of IRR-219 expression during intestinal adaptation. Cell-specific expression of IRR-219 mRNA was detected in ileal sections from resected (A) or sham-resected control (B) rats killed at 48 h after surgery. Representative goblet cells are depicted by arrows. Sections are viewed by darkfield microscopy (×200).

Expression of goblet cell-specific RITF is unaltered in adaptive intestine at 48 h after resection. To determine the specificity of the alteration in IRR-219 mRNA levels at 48 h after intestinal resection, the expression pattern of another goblet cell-specific gene, RITF, was examined by Northern blot hybridization in adaptive vs. sham-resected control ileum and colon (Fig. 9). RITF is expressed exclusively in goblet cells of colon and small bowel and is secreted with mucin (23). Unlike IRR-219, RITF mRNA levels were unchanged in adaptive vs. control small bowel and colon.

View larger version (15K): [in this window] [in a new window]   Fig. 9.   Rat intestinal trefoil factor (RITF) mRNA levels are unchanged in adaptive (resected, RE) compared with control SHR small bowel and colon. Total RNAs were prepared as in Fig. 3, and Northern blot hybridization was performed as described in METHODS using a radiolabeled RITF cDNA probe. No significant differences were found comparing SHR to RE proximal small intestine, ileum, or colon.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The intestinal adaptive response is a complex process characterized by alterations in crypt cell proliferation, epithelial cell migration, and apoptosis (7, 14, 24, 26). To identify the molecular pathways responsible for initiating the adaptive response, we have used subtractive hybridization techniques to identify genes that are regulated in the early adaptive period (6, 21). This methodology led to the identification of several groups of cDNAs that encode proteins that affect nutrient absorption, the cell cycle, and protein synthesis and processing (6, 21). In this study, we report the cloning and expression analysis of a unique cDNA, IRR-219, that is expressed only in goblet cells and is differentially regulated in the small bowel and colon during adaptation after intestinal resection.

Sequence analysis of the IRR-219 cDNA revealed 80% nucleotide identity with the 3'-end of the human IgGFcBP (13), suggesting that it is the rat homologue of this gene. The human IgGFcBP is distinct from both the IgG Fc receptor located on the apical surface of the enterocyte that mediates transcytosis of milk immunoglobulin and from the IgG Fc receptor on leukocytes. This very large protein is expressed only in intestinal and colonic goblet cells, which secrete it with mucus into the gut lumen (15, 16). The mRNA is also expressed in human placenta (13). The physiological function of this protein is not known. Its ability to specifically bind the Fc portion of IgG and its secretion with mucus are consistent with presumptive roles such as protection of IgG molecules from bacterial degradation, promotion of multivalent IgG formation, and combination with antigen-IgG complexes to prevent antigen invasion into the mucosa (13).

Our in situ hybridization data support the supposition that IRR-219 is the rat homologue of the human IgG Fc binding protein, because IRR-219 mRNA is expressed exclusively in goblet cells of the small bowel and colon, as is the human protein (13). However, previous data indicated that this binding site is not present in rodents (16), because horseradish peroxidase-labeled IgG stained only human goblet cells and not rat or rabbit small and large bowel epithelium. The region containing the putative Fc binding site is not present in the partial IRR-219 cDNA clone. In addition, nonstringent Southern blot hybridization of genomic DNA from various animal species detected IgGFcBP sequence only in humans and primates (13). The probe used for Southern blot hybridization was derived from a sequence upstream of the region represented in our rat IRR-219 cDNA, suggesting that IRR-219 may be a homologue with greater sequence divergence in the 5' region.

The role of IRR-219 in the normal and adapting intestine remains to be clarified. From 24 h to 1 wk after small bowel resection, IRR-219 mRNA levels were decreased in the adaptive small intestine compared with the sham-operated control yet were increased in the adapting colon. In contrast, the expression of another goblet cell-specific gene product, RITF, was unchanged, indicating the specificity of this response. The increase in colonic mRNA levels is consistent with a cytoprotective role for IRR-219, similar to one postulated for the human IgGFcBP, in response to the increased exposure of the colonic mucosa to antigens, bile acids, undigested and partially digested food, acid, and other enteral contents. The observation that IRR-219 mRNA levels did not similarly increase in the adaptive ileum suggests that innate differences between small bowel and colonic goblet cells, or differences in their luminal environments, might affect this response. For example, the heterogeneity in goblet cell populations between small bowel and colon can be demonstrated by variations in goblet cell sialomucin and sulfomucin content (1). Luminal contents also vary in the postresective intestine, e.g., there is a relatively greater change in luminal bile salt concentration in the postresective cecum/colon compared with small bowel. Intestinal isograft experiments in which fetal intestine is grown to maturity in the subcutaneous space of a syngeneic host in the absence of luminal contents (11, 17-19) may prove useful to begin to answer these questions.

If IRR-219 is the rat IgG Fc binding protein, then factors that regulate expression of the human IgGFcBP might also be expected to affect IRR-219 expression in gut. On the basis of the putative role for IgGFcBP in the intestinal immune response, cytokine regulation of binding site expression was examined in HT29-N2 cells, a subclone of HT29 cells that differentiate into goblet cells when grown in media containing galactose (12). Tumor necrosis factor- (TNF-) was shown to decrease binding site expression independent of effects on cell proliferation, whereas interferon- had no effect. Other cytokines were not studied. The regulation and effects of TNF- in small bowel adaptation after resection are unknown. An inflammatory response is not typically seen after small bowel resection, so whether TNF- or other cytokines might affect IRR-219 expression during adaptation remains to be clarified.

In conclusion, a goblet cell-specific gene has been cloned that is expressed in a unique pattern during intestinal adaptation after resection. Few markers of colonic adaptation have been identified, making IRR-219 potentially useful in furthering our understanding of the large bowel's response to loss of small bowel surface area. Further determination of its relationship to the human IgG Fc binding protein awaits isolation of the full-length cDNA. Its regulation in other models of intestinal injury and repair (such as irradiation or infection) may help further define its function in small bowel and colon.