Original Article
Enhanced Local Production of Complement Components in the Small Intestines of Patients with Crohn's Disease
N Engl J Med 1990; 322:1345-1349May 10, 1990
- Abstract
- Abstract
There is evidence that complement components may be formed locally in inflammatory lesions containing monocytes and macrophages. To investigate the role of complement in Crohn's disease we measured jejunal-fluid concentrations of the complement components C4, C3, and factor B by perfusion of a closed segment of the jejunum in 22 patients with Crohn's disease thought to be limited to the terminal ileum.
The mean (±SEM) jejunal-fluid C4 concentration was 2.0±0.3 mg per liter, significantly higher than the mean level in 35 healthy controls (0.7±0.1 mg per liter; P<0.001). The mean C3 concentration was 1.0±0.1 mg per liter in the patients and 0.7±0.1 mg per liter in the controls (P<0.05). The factor B levels were similar in the two groups. Calculated rates of intestinal secretion of these components showed differences of the same magnitude. Leakage of protein from plasma was not increased. The jejunal-fluid serum ratios of these complement proteins indicated that their appearance in the lumen of the jejunum was due at least in part to local mucosal synthesis. The increased jejunal secretion of C4, but not C3 or factor B, paralleled the clinical activity of Crohn's disease. Values were normal in first-degree relatives of the patients (n = 13), patients with celiac disease (n = 8), and patients with ulcerative colitis (n = 4).
We conclude that increased secretion of complement by clinically unaffected jejunal tissue in patients with Crohn's disease reflects the systemic nature of this disorder and may be due to the stimulated synthesis of complement by activated intestinal monocytes and macrophages. (N Engl J Med 1990; 322:1345–9.)
Article Activity
- Article
EVIDENCE of extrahepatic synthesis of complement components by cells of the monocyte—macrophage series has been obtained in vitro.1 2 3 4 5 6 Enhanced local synthesis of complement components may also occur in vivo, in tissues invaded by macrophages and monocytes. The concentration of locally synthesized complement should depend on the number of these cells and on other factors that regulate the production and secretion of complement proteins.5 , 6 We studied the jejunal secretion of the complement components C4, C3, and factor B in patients with Crohn's disease. The rate of secretion of complement proteins was measured with the aid of a recently developed intestinal-tube system with two occluding balloons.7 This system allows the perfusion of a well-defined "closed" segment of the jejunum.
The intestinal inflammation of Crohn's disease is characterized by the lymphoid infiltration of mononuclear and polymorphonuclear leukocytes. Immune complexes deposited in the intestinal mucosa may activate complement components, thereby contributing to the inflammatory lesions found in this disorder.8 9 10 We found that the jejunal secretion of C4 was greatly enhanced in patients with Crohn's disease and was linked to the level of disease activity.
Methods
We studied 22 patients with biopsy-verified Crohn's disease of the terminal small bowel (15 women and 7 men, with a mean age of 34 years; range, 18 to 62). The level of disease activity was assigned a score according to the Crohn's disease—activity index, in which a remission is indicated by an index score below 150.11 The mean duration of disease was 6.5 years (range, 1 to 16). Two patients had extraintestinal manifestations of the disease (erythema nodosum and arthralgia). The patients were compared with 35 healthy controls similar to them in age and sex (20 women and 15 men, with a mean age of 30 years; range, 21 to 55). We also studied 13 healthy first-degree relatives of the patients (9 women and 4 men, with a mean age of 47 years; range, 37 to 65), 4 patients with ulcerative colitis (1 woman and 3 men, with a mean age of 47 years; range, 37 to 65), and 8 patients with celiac disease (3 women and 5 men, with a mean age of 45 years; range, 36 to 68). The study was approved by the ethics committee of the medical faculty of Uppsala University.
A segment of the intestine was perfused by means of a small-diameter tube ( 16 French, or 5.3 mm in external diameter) that was 175 cm long and contained six channels (LOC-I-GUT, Pharmacia, Uppsala)7; two latex balloons 10 cm apart were attached to the tube. The insertion of the tube, gastric drainage, inflation of the balloons, and rinsing of the closed intestinal segment were performed as previously described.7 The intestinal segment had a volume in vivo of 80 to 100 ml and was perfused at a rate of 3 ml per minute with a solution containing 10 mmol of glucose, 5.4 mmol of potassium chloride, 120 mmol of sodium chloride, 2 mmol of disodium hydrogen phosphate, 1 g of polyethylene glycol (molecular weight, 4000) and 35 mmol of mannitol per liter. Aprotinin (Bayer, Leverkusen, Federal Republic of Germany) was added to the perfusion solution to inhibit possible proteolytic activity in the effluent.7 The duration of perfusion was 180 minutes; samples of the perfusate were collected on ice at 20-minute intervals. 14C-labeled polyethylene glycol was used as a volume marker, and phenolsulfonphthalein as a marker of the patency of the proximal balloon.7 The instilled volume of perfusion fluid was 60 ml per 20 minutes, and the volume of the recovered effluent was on average 59 and 57 ml per 20 minutes in controls and patients, respectively. The samples were kept frozen at — 70°C until they were analyzed. Since some of the assays might be affected by the presence of even small amounts of proteases, 2 mmol of phenylmethylsulfonyl fluoride per liter (Sigma Chemical, St. Louis) was added to all samples before they were thawed on ice and analyzed.
Complement components C3, C4, and factor B were measured by inhibition enzyme-linked immunosorbent assays. The assay for C3 has been described elsewhere.12 Similar assays were developed for C4 and factor B. C3 (3 μg per milliliter), C4 (1 μg per milliliter), and factor B (1 μg per milliliter) were preadsorbed to the surface of wells of microtiter plates in a volume of 200 μl of phosphate-buffered saline overnight at 4°C. Fifty microliters of each sample was diluted 20-fold in working buffer supplemented with 2 mmol of phenylmethylsulfonyl phosphate per liter, and 50 μl of constant doses of rabbit IgG antibodies against C3c, C4, or factor B was incubated at room temperature for 60 minutes. The anti—C3c antibodies were conjugated with horseradish peroxidase. The working buffer used was phosphate-buffered saline containing 0.1 percent polysorbate (Tween 20, vol/vol) and 0.1 percent bovine serum albumin (wt/vol). One hundred microliters of swine antirabbit IgG conjugated with horseradish peroxidase was allowed to bind to the bound anti—complement factor antibodies (DAKO-Immunoglobulins, Roskilde, Denmark) for 60 minutes at room temperature. The samples from patients and controls were measured in sequence to avoid day-to-day variation. The applicability of the assays to jejunal fluid was confirmed by dilution and recovery experiments, which demonstrated activity parallel to that of the standard curves and the complete recovery of added proteins. Nonreduced samples of jejunal fluid were also analyzed by sodium dodecyl sulfate–polyacrylamide-gel electrophoresis (SDS-PAGE) in 12 percent gels on a Phast-system apparatus (Pharmacia),13 which was followed by Western blotting.14 C3 and factor B were prepared as previously described.15 , 16 The C4 was a gift from Dr. Gösta Eggertsen (Department of Clinical Chemistry, Huddinge University Hospital, Huddinge, Sweden). Albumin was measured by radioimmunoassay (Pharmacia).
Measurements were performed on two randomly selected samples of perfusate from each subject. Samples were discarded if they had a phenolsulfonphthalein concentration more than 5 percent of the concentration of the marker when infused into the stomach, a [14C]polyethylene glycol concentration less than 80 percent of the concentration of the marker when infused into the jejunal segment, or a recovered volume (per 20-minute period) less than 80 percent of the volume infused into the jejunal segment during the same period. The results are expressed as the arithmetric mean of the two samples.
The erythrocyte sedimentation rate was determined according to the Westergren method; this variable, the leukocyte count, and hemoglobin, orosomucoid (α1-acid glycoprotein, found in plasma), folic acid, serum zinc, and serum iron levels were measured at the Department of Clinical Chemistry, University Hospital, Uppsala.
Absolute secretion rates were calculated after correction for the loss of l4C activity, according to standard equations, and the rate of secretion was expressed as the amount secreted per centimeter of bowel per hour.17 Results are expressed as means ±SEM. Statistical analysis was performed with Student's two-tailed t-test for unpaired values. Differences were considered to be significant if the P value was less than 0.05.
Results
Eleven of the 22 patients with Crohn's disease of the terminal small bowel had active disease; their average score for the disease-activity index was 251 (range, 170 to 400). The 11 patients with inactive disease (i.e., with a score <150) had previously undergone surgery involving ileal or ileocecal resection. Values for laboratory indicators of disease activity are summarized in Table 1Table 1
Laboratory Values in Patients with Crohn's Disease, According to Crohn's Disease-Activity Index11 Score.*. Two patients with active disease were being treated with metronidazole at the time of the study. None of the patients had recently received or were receiving corticosteroid or immunosuppressive therapy.The serum concentrations of C3 and C4 were similar in the patients with inactive Crohn's disease and the healthy controls, but these concentrations were significantly increased (P<0.05) in the patients with active Crohn's disease as compared with the healthy controls (Table 2Table 2
Serum Concentrations of Complement Components and Albumin in Patients with Crohn's Disease and Healthy Controls.*). The serum concentrations of factor B tended to be increased in the patients, and their serum albumin levels were slightly decreased (Table 2). Serum values were not available for 4 of the 11 patients with inactive disease because the serum samples for the test day were missing.The jejunal-fluid concentrations of the measured complement components in the patients with Crohn's disease and the controls are shown in Table 3Table 3
Jejunal-Fluid Concentrations of Complement Components in Patients with Crohn's Disease, Their First-Degree Relatives, Patients with Ulcerative Colitis, Patients with Celiac Disease, and Healthy Controls.*. The concentrations of C3 were increased approximately 40 percent in the patients (P<0.05), and those of C4 approximately 200 percent (P<0.001); however, the patients' factor B concentrations were similar to those of the controls. To compensate for differences in flow rates during perfusion, the secretion rates of C3, C4, and factor B were calculated (Table 3). The differences between the jejunal-fluid concentrations of the components in the patients and those in the controls were similar to the differences in the secretion rates. When the patients were stratified according to the level of disease activity, the jejunal secretion rate of C4 was found to be increased fourfold in those with active disease and twofold in those with inactive disease. The increase in the secretion rate of C3 was similar in the two subgroups (Fig. 1Figure 1
Rates of Secretion of C4, C3, Factor B, and Albumin in Jejunal Fluid from Healthy Controls (I), Patients with Inactive Crohn's Disease (II), and Patients with Active Crohn's Disease (III).).Five of the patients were studied again 11 to 17 months after jejunal perfusion. Two of the patients had had active disease during the first investigation and were in remission during the second investigation, after they had undergone ileocecal resection. Their jejunal secretion rates for C4 decreased from 32 and 33 μg per centimeter per hour to 17 and 18 μg per centimeter per hour, respectively. During the first investigation, a patient who had been in remission after ileocecal resection had a secretion rate for C4 of 20 μg per centimeter per hour; during the second investigation, when recurrent disease had developed in the distal ileum, the rate increased to 63 μg per centimeter per hour. Two patients had similar levels of disease activity at follow-up; their secretion rates for C4 remained stable, averaging 72 and 58 μg per centimeter per hour during the two periods of investigation. The patients' secretion rates for C3 and factor B varied within narrow limits at the second investigation, and the variations were not related to their levels of disease activity.
The amount of albumin recovered was similar in the patients and controls; their mean jejunal-fluid concentrations were 25±4 and 27±4 mg per liter, respectively. The mean ratio of albumin in jejunal fluid to albumin in serum was 7 × 10–4 in the patients and controls, indicating a similar degree of leakage from the intestinal mucosa. The corresponding ratios for C4 in the patients and controls were 7 × 10–3 and 3 × 10–3; for C3, 8 × 10–4 and 7 × 10–4; and for factor B, 2 × 10–3 and 3 × 10–3.
To assess the state of activation of the complement proteins in jejunal fluid, we tested samples of perfusate from five patients with active Crohn's disease and high levels of C3 in the perfusate, by means of SDS-PAGE with Western blotting that employed antibodies to C3c and C3d. Multiple protein bands located close together that had molecular weights similar to that of C3c were observed in all tested samples (Fig. 2Figure 2
Results of SDS-PAGE with Western Blotting of Nonreduced Samples of Perfusate from Five Patients with Crohn's Disease and High Concentrations of C3 in Jejunal Fluid (Lanes 1 through 5).). Furthermore, low-molecular-weight proteins were detected in three samples; these bands were specifically identified with the anti—C3c antibodies (data not shown). No difference was observed in the fragmentation pattern when phenylmethylsulfonyl fluoride was added to the perfusate directly after lavage of the intestine or after testing by SDS-PAGE with Western blotting.To determine the specificity of these observations to Crohn's disease, we analyzed perfusion fluid collected from patients with celiac disease and patients with ulcerative colitis. The concentrations of C4, C3, and factor B were similar to those measured in the healthy controls (Table 3). Finally, we investigated the jejunal secretion of complement proteins in first-degree relatives of patients with Crohn's disease. Their concentrations of C4, C3, and factor B did not differ from those of the controls (Table 3).
Discussion
These studies indicate that in patients with Crohn's disease, the small intestine is a site of enhanced synthesis of C4 and C3, but not factor B. Although the liver is the principal site of synthesis of several of the serum complement proteins, extrahepatic synthesis of these proteins has been demonstrated in vitro.18 The extrahepatic cells capable of producing complement proteins are the cells of the reticuloendothelial system1 2 3 4 5 6 and epithelial cells of various origins.19 In vitro, the synthesis of complement by macrophages is greatly enhanced after stimulation with inflammatory factors.5 , 6 Our observations in patients with Crohn's disease are therefore suggestive of enhanced mucosal production of complement as a result of intestinal inflammation. However, the increased jejunal secretion of complement components in Crohn's disease does not seem to be a nonspecific finding in inflammatory bowel disease, since patients with ulcerative colitis and celiac disease had normal secretion of complement.
The area of the jejunum that we investigated with perfusion studies was morphologically uninvolved by Crohn's disease. Therefore, these results suggest that a subclinical inflammatory process may occur in the entire small bowel in this disorder. Previous investigators have also emphasized the systemic nature of the disease.20 21 22 23 24 25 26 The observation that the C3 found in the jejunal fluid had different molecular sizes suggests that complement proteins are fragmented in the intestine. The fact that proteinase inhibitor was present in the perfusate used in the perfusion studies, that the perfusion fluid was collected on ice, and that serine proteinase inhibitors added to the perfusate had no effect on further fragmentation of complement suggests that the fragmentation occurs in the intestinal wall. This may reflect a physiologic process inducing inactivation of the complement system in the intestine. It is also possible that complement fragments in the jejunal lumen induce the activation of inflammatory cells.27 28 29 30 The signs of enhanced jejunal synthesis of complement — or at least of C4 — seem to be linked to the clinical activity of Crohn's disease. However, the observation that patients in remission also had increased rates of secretion of complement components may suggest that enhanced jejunal production of complement is not only a secondary event. The finding that clinically unaffected relatives of patients with Crohn's disease have an increased intestinal permeability to polyethylene glycol31 led us to perform jejunal-perfusion studies in the relatives of our patients. However, we have not observed any change in the jejunal secretion of complement in the relatives studied to date.
The fourfold increase in the rate of jejunal secretion of C4 observed in the patients with active Crohn's disease cannot be explained by increased permeability of the mucosal membrane, since albumin, a marker of plasma leakage into the jejunum, appeared in similar concentrations in the jejunal fluid of patients and controls. The serum concentrations of C4 were slightly increased in the patients, but this too cannot explain the concentrations observed in jejunal fluid, in view of the jejunal-fluid:serum ratios for albumin and C4. In vitro studies have documented that stimulated macrophages may increase the synthesis of C4 by 5 to 10 times per cell.5 Thus, our observations in patients with Crohn's disease are compatible with an enhanced jejunal synthesis of C4, possibly by activated mucosal macrophages. However, the possibility cannot be excluded that the recovered amounts of C4 differ in health and disease because of differences in the rate of catabolism of C4 in the jejunum. The measured amounts of C3 in jejunal fluid in the patients were also elevated, although less so than the amounts of C4. The observation that the amounts of C3 recovered during jejunal perfusion tended to be higher in the patients with inactive disease than in those with active disease may reflect differences in the rate of degradation of C3 that are dependent on the level of disease activity. The measured levels of complement in jejunal fluid from the healthy controls were relatively higher than the levels that could be expected to result from passive leakage from plasma. Therefore, local intestinal synthesis of complement may also occur in healthy persons.
The capacity to increase concentrations of complement components at a site of inflammation may be critical to complement-dependent functions in host defenses. However, in Crohn's disease such a mechanism may not be beneficial, but may in fact contribute to the inflammatory events occurring in the intestine and thus to tissue damage. Recent immunohistochemical studies have revealed deposits of terminal complement complex in muscularis mucosae and submucosal vessels in inflamed intestinal tissue from patients with Crohn's disease.32
Supported by grants from the Swedish Medical Research Council, the Tore Nilson Foundation, the Swedish Life Insurance Companies' Trust for Medical Research, and Pharmacia, Sweden.
Source Information
From the Departments of Surgery (O.A., L.K.), Clinical Immunology (B.N., K.N.-E.), and Internal Medicine (R.H.), University Hospital, and the Department of Research and Development, Pharmacia (B.O.), Uppsala, Sweden. Address reprint requests to Dr. Ahrenstedt at the Department of Surgery, University Hospital, S-751 85 Uppsala, Sweden.
- References
References
1
Einstein
CrossRef | Web of Science | Medline2
Rubin
Web of Science | Medline3
Littleton
Web of Science | Medline4
Wyatt
Web of Science | Medline5
Cole
Web of Science | Medline6
Einstein
CrossRef | Web of Science | Medline7
Knutson
Web of Science | Medline8
Baklien
CrossRef | Web of Science | Medline9
Hodgson
Web of Science | Medline10
Lake
Web of Science | Medline11
Best 12
Nilsson
Web of Science | Medline13
Laemmli
CrossRef | Web of Science | Medline14
Nilsson
CrossRef | Web of Science | Medline15
Hammer
Web of Science | Medline16
Kerr
Web of Science | Medline17
Stern
Web of Science | Medline18
Colten
CrossRef | Medline19
Teisberg
CrossRef | Web of Science | Medline20
McCallum
CrossRef | Web of Science | Medline21
Nugent
Full Text | Web of Science | Medline22
Ménard
Full Text | Web of Science | Medline23
Beitman
CrossRef | Web of Science | Medline24
Bonniere
CrossRef | Web of Science | Medline25
Hällgren
CrossRef | Web of Science | Medline26
Knutson
CrossRef | Web of Science | Medline27
Hugli
CrossRef | Medline28
Seya
Web of Science | Medline29
Hoeprich
Web of Science | Medline30
Melchers
CrossRef | Web of Science | Medline31
Hollander
Web of Science | Medline32
Halstensen
CrossRef | Web of Science | Medline
- Citing Articles (15)
Citing Articles
1
Sae-Hae Kim, Dae-Im Jung, In-Young Yang, Ju Kim, Kyung-Yeol Lee, Tomonori Nochi, Hiroshi Kiyono, Yong-Suk Jang. (2011) M cells expressing the complement C5a receptor are efficient targets for mucosal vaccine delivery. European Journal of Immunology 41:11, 3219-3229
CrossRef2
Sebastian Zeissig, Theresa Bergann, Anja Fromm, Christian Bojarski, Frank Heller, Ute Guenther, Martin Zeitz, Michael Fromm, Jörg–Dieter Schulzke. (2008) Altered ENaC Expression Leads to Impaired Sodium Absorption in the Noninflamed Intestine in Crohn's Disease. Gastroenterology 134:5, 1436-1447
CrossRef3
Alberto José León, José Antonio Garrote, Eduardo Arranz. (2006) Citocinas en la patogenia de la enfermedad inflamatoria intestinal. Medicina Clínica 127:4, 145-152
CrossRef4
K. M. BLAKE, S. O. CARRIGAN, A. C. ISSEKUTZ, A. W. STADNYK. (2004) Neutrophils migrate across intestinal epithelium using beta2 integrin (CD11b/CD18)-independent mechanisms. Clinical and Experimental Immunology 136:2, 262-268
CrossRef5
Michael O Ogundele. (2001) Role and significance of the complement system in mucosal immunity: Particular reference to the human breast milk complement. Immunology and Cell Biology 79:1, 1-10
CrossRef6
Arvind Sahu, John D Lambris. (2000) Complement inhibitors: a resurgent concept in anti-inflammatory therapeutics. Immunopharmacology 49:1-2, 133-148
CrossRef7
J. Laufer, R. Oren, I. Goldberg, A. Horwitz, J. Kopolovic, Y. Chowers, J. H. Passwell. (2000) Cellular localization of complement C3 and C4 transcripts in intestinal specimens from patients with Crohn's disease. Clinical and Experimental Immunology 120:1, 30-37
CrossRef8
JOSEPH LAUFER, RACHEL OREN, IRIS GOLDBERG, ARNON AFEK, JURI KOPOLOVIC, JUSTEN HOWARD PASSWELL. (1999) Local Complement Genes Expression in the Mammary Gland: Effect of Gestation and Inflammation. Pediatric Research 46:5, 608
CrossRef9
H. Farkas, L. Gyeney, E. Nemesánszky, G. Káldi, F. Kukán, I. Masszi, J. Soós, M. Bély, E. Farkas, G. Füst, L. Varga. (1999) Coincidence of hereditary angioedema (HAE) with Crohn's Disease. Immunological Investigations 28:1, 43-53
CrossRef10
Hans Lennern
s. (1998) Human intestinal permeability. Journal of Pharmaceutical Sciences 87:4, 403-410
CrossRef11
Andoh, Fujiyama, Sakumoto, Uchihara, Kimura, Koyama, Bamba. (1998) Detection of complement C3 and factor B gene expression in normal colorectal mucosa, adenomas and carcinomas. Clinical and Experimental Immunology 111:3, 477-483
CrossRef12
HANS LENNERNÄS. (1997) Human Jejunal Effective Permeability and Its Correlation with Preclinical Drug Absorption Models. Journal of Pharmacy and Pharmacology 49:7, 627-638
CrossRef13
Per Brandtzaeg. (1995) Autoimmunity and ulcerative colitis: Can two enigmas make sense together?. Gastroenterology 109:1, 307-312
CrossRef14
Robert A. Kozol. (1992) Neutrophil recruitment to the gastrointestinal tract. Journal of Surgical Research 53:3, 310-315
CrossRef15
Podolsky, Daniel K., . (1991) Inflammatory Bowel Disease. New England Journal of Medicine 325:13, 928-937
Full Text
