Serum and Intestinal Immune Response to Rotavirus Enteritis in ...

4 downloads 1349 Views 766KB Size Report
Oct 4, 1982 - virus-specific antibody of each immunoglobulin class per milligram of total ..... ance, and C. Lloyd, Medical Illustrator, The Adelaide Chil- o.
INFECTION AND IMMUNITY, May 1983, p. 447-452 0019-9567/83/050447-06$02.00/0

Vol. 40, No. 2

Serum and Intestinal Immune Response to Rotavirus Enteritis in Children GEOFFREY P. DAVIDSON,* RUSSELL J. HOGG, AND CHELLAM P. KIRUBAKARAN Gastroenterology Unit, The Adelaide Children's Hospital Inc., North Adelaide, S.A. 5006, Australia Received 4 October 1982/Accepted 3 January 1983

This study was designed to assess the serum and mucosal immune response to naturally acquired rotavirus enteritis in children. Serum and duodenal secretions were collected 1 week and again 4 to 5 weeks after the onset of illness from 10 children. In two of these children, the procedure was repeated 12 to 15 months later. Another 10 children with bacterial enteritis were studied as controls. The antibody response in serum included a significant elevation of rotavirus-specific immunoglobulin M (IgM) in acute-phase samples (P < 0.05), but not in convalescent-phase samples, when compared with controls. Rotavirus-specific IgG and IgA levels were significantly elevated in convalescent-phase serum when compared with acute-phase serum (P < 0.025), but not in control serum. Rotavirusspecific IgA levels in convalescent duodenal secretions were significantly raised when compared with both acute-phase and control samples (P < 0.01). Rotavirusspecific IgM levels were elevated in acute-phase duodenal secretions (P < 0.05), but not in convalescent-phase secretions. In two children, the secretory IgA response had disappeared by 12 months. These studies demonstrate the presence of rotavirus-specific antibody in duodenal secretions which may be important for protection against reinfection and may be capable of being stimulated by oral vaccination.

Rotavirus is now recognized as the major etiological agent of infectious diarrhea in infants and young children throughout the world (9). In developed countries, rotavirus causes low rates of mortality (2) but high rates of morbidity (10). In developing countries, however, rotavirus is probably a major cause of mortality, especially in children less than 2 years old. Black et al. (1) have recently produced evidence to support this in a study in Bangladesh in which they showed that children with rotavirus were more likely to suffer from dehydration and also more likely to visit a treatment center than children with other causes of diarrhea. They calculated a mortality rate of 6.5 per 1,000. In considering vaccine development, it is important to know (i) the types of organisms causing diarrhea and their relative contributions to death and disability, and (ii) the mechanisms of host resistance to these organisms. Little specific information on intestinal immunity to rotavirus is available. Breast feeding is known to provide protection, and this has been related to the level of rotavirus-specific immunoglobulin A (IgA) in breast milk (12). Colostrum containing anti-rotavirus immunoglobulin has been shown to prevent infection in lambs (19). Recent evidence from experimental studies in adult volunteers suggests that local intestinal IgA antibody

directed against type 2 rotavirus is of importance in resistance to the illness (10). These studies all support the contention that antibody in the lumen of the small intestine is important in resistance to rotavirus enteritis. In the present study, the serum and secretory intestinal antibody responses to rotavirus enteritis were measured with a sensitive enzyme-linked immunosorbent assay. The main aims were to determine whether children developed a local intestinal response to infection, the duration of such a response, and whether this intestinal response was mirrored in the serum. MATERIALS AND METHODS Patients. The 10 children studied all showed features of acute enteritis requiring hospital admission, and a diagnosis of rotavirus infection was confirmed by electron microscopic examination of stool specimens. Their ages ranged from 2 to 24 months (mean, 12.7 months). Controls were 10 children studied earlier with proven salmonella enteritis whose ages ranged from 10 to 73 months (mean, 33.3 months). They were not matched in any way with the patients. Samples of intestinal fluid were collected from the duodena of patients through a double lumen tube. Fluoroscopy was used to position the tube. The aspirate was collected onto ice, divided into portions, and stored at -70°C until assayed. The samples were taken within 1 week of the onset of symptoms, 30 to 45 days 447

448

INFECT. IMMUN.

DAVIDSON, HOGG, AND KIRUBAKARAN

later, and (for two children) after 12 to 15 months. Duodenal secretions were tested by light microscopy, microbial culture, and electron microscopy for evidence of pathogens. Total immunoglobulin levels. Concentrations of IgG, IgA, and IgM in serum and duodenal secretions were estimated by the radial immunodiffusion method of Fahey and McKelvey (4) with commercially available plates (LC-Partigen and Tri-Partigen, Hoechst Australia Pty. Ltd., Melbourne, Australia). The concentration of IgA in duodenal secretions was estimated with 11S IgA standards. Preparation of standard curves. Purified human IgG, IgA, and IgM was kindly supplied by J. T. LaBrooy, Royal Adelaide Hospital, Adelaide, Australia. To construct standard curves, it was necessary to determine the binding affinities of each immunoglobulin class for the polyvinyl microtiter wells. For this purpose, immunoglobulins were radiolabeled with 125I by the chloramine-T method, and unbound iodine was separated by Sephadex G-25 filtration. The proportion of protein-bound 125I in the final preparations was monitored by the trichloroacetic acid precipitation technique. Known quantities of radiolabeled immunoglobulin were adsorbed to polyvinyl microtiter trays by dilution in carbonate-bicarbonate coating buffer (pH 9.6) and incubation for 18 h at 4°C. The proportion of immunoglobulin remaining bound at this and subsequent stages of the assay was determined by counting the remaining radioactivity in each well of the microtiter tray. The microtiter plates were incubated with enzyme-labeled anti-globulins and substrate in the same way as the test plates as outlined below. This procedure demonstrated that at the nanogram per milliliter level 51, 39, and 38% of the original quantities of IgG, IgA, and IgM, respectively, remained bound to the microtiter wells at the completion of the assay. Absorbance values were related to immunoglobulin binding data in constructing standard curves. Enzyme-linked immunosorbent assay for rotavirusspecific antibody. The enzyme immunoassay used to detect anti-rotavirus antibodies was essentially that described by Ghose et al. (5) with several modifications. Briefly, Cooke polyvinyl microtiter trays (catalog no. 220-29) were incubated with (per well) either 100 ,ul of a previously standardized dilution of the tissue culture-raised simian rotavirus (SA-11) in 0.1 M carbonate-bicarbonate coating buffer (pH 9.6) or 100 ,ul of buffer alone for 18 h at 4°C. Simian rotavirus has been proven antigenically similar to human rotavirus (18, 20). After incubatiori, the plate was washed three times in phosphate-buffered saline with 0.15% Tween 20 (polyoxyethylene sorbitan monolaurate, Sigma P1397). Each wash consisted of a 3-min soaking and an emptying of the wells. Doubling dilutions of serum and duodenal secretions from 1/50 and 1/3, respectively, were made in phosphate-buffered saline containing 0.15% Tween 20 and 2% bovine serum albumin (Sigma RIA grade VII) (PBS-Tween-BSA). A 75-,ul portion of each dilution was added to each test and control well and was incubated for 18 h at 4°C. All dilutions were tested in duplicate. The incubation procedure was chosen to avoid enzymatic degradation of immunoglobulin in duodenal juice samples. After further washing, 50 p1d (per well) of a predeter-

mined dilution of affinity-purified alkaline phosphatase-labeled anti-immunoglobulin heavy-chain antisera (Kirkegaard & Perry Laboratories, Gaithersburg, Md.) diluted in PBS-Tween-BSA was incubated for 3 h at 37°C. After a further washing sequence, 50 ml of pnitrophenylphosphate substrate, diluted to 1 mg/ml in 0.1 M carbonate-bicarbonate buffer (pH 9.8) containing 1 mM MgCl2, was added to each well and incubated for 60 min at 37°C. The enzyme-substrate reaction was stopped by the addition of 25 ,ul of 3 M NaOH. The absorbance values were determined from each well with a through-the-plate absorbance monitor, and antibody concentrations were estimated from the standard curve. The coefficient of variation was 7.9o within the assay and 18.3% between assays. The results are expressed as nanograms of rotavirus-specific antibody of each immunoglobulin class per milligram of total immunoglobulin. This was done to compensate partially for dilution factors which are present in duodenal secretions. Results were analyzed by the Wilcoxon matched pairs test and the MannWhitney U test for matched and unrelated samples, respectively.

RESULTS The duodenal juice was negative for rotavirus antigen. Bacterial cultures of duodenal juice showed no significant microbial growth. Table 1 shows the rotavirus-specific antibody levels for all patients and controls and also for the followup samples. Figure 1 shows the IgG results for both serum and secretions. The antibody levels showed a skewed distribution with a few very high responses, making the median the most appropriate way of illustrating the information. Rotavirus-specific IgG levels in convalescent-phase serum (sample 2) were significantly elevated when compared with both acute-phase (sample 1) and control serum (P < 0.025). IgG levels in two patients 12 to 15 months after infection were also significantly elevated. The levels at this time were higher than those seen at 6 weeks. Rotavirus infection failed to stimulate a specific IgG class response in duodenal secretions. IgM levels were significantly elevated in the acute-phase serum specimens when compared with convalescent-phase, control, and 12-month follow-up samples (P < 0.05) (Fig. 2). Rotavirusspecific IgM levels were elevated when compared with control values (P < 0.05). Levels were below the limits of detection in two patients seen over 12 months later. Rotavirus-specific IgA levels were significantly elevated in convalescent-phase serum when compared with acute-phase serum (P < 0.025), but not when compared with controls (Fig. 3). Rotavirus-specific IgA levels in convalescentphase duodenal secretions were significantly raised when compared with both acute-phase and control samples (P < 0.05). The rotavirus-

VOL. 40, 1983

IMMUNE RESPONSE TO ROTAVIRUS ENTERITIS

449

TABLE 1. Levels of rotavirus-specific antibody Source of samples

Serum from patient: 1 2 3 4 5 6 7 8

9 10 Duodenal aspirate from patient: 1 2 3

Antibody level' Rotavirus enteritis patients Sample 1 Sample 2 12 mo IgG IgA IgM IgG IgA IgM IgG IgA

0.19