ARTICLES. An Epizootic of Lymphoplasmacytic Gastritis Attributed to Helicobacter pylori Infection in Cynomolgus Monkeys (Macaca fascicularis)

Transcription

1 Vet Pathol 36:1 13 (1999) ARTICLES An Epizootic of Lymphoplasmacytic Gastritis Attributed to Helicobacter pylori Infection in Cynomolgus Monkeys (Macaca fascicularis) J. F. REINDEL, A. L. FITZGERALD, M. A. BREIDER, A. W. GOUGH, C. YAN, J. V. MYSORE, AND A. DUBOIS Departments of Pathology and Experimental Toxicology (JFR, ALF, MAB, AWG), and Biometrics (CY), Parke-Davis Pharmaceutical Research Division of Warner-Lambert Co., Ann Arbor, MI; and Digestive Diseases Division, Department of Medicine, Uniformed Services University of the Health Sciences, Bethesda, MD (JVM, AD) Abstract. An epizootic of subclinical lymphoplasmacytic gastritis occurred in cynomolgus monkeys maintained at our research facility. Gastric pathology data and histologic sections of 63 adolescent monkeys ( years old) sacrificed during the epizootic were reviewed. Localized to multifocal reddening of the gastric mucosa was noted grossly in 7 of 44 (16%) monkeys harboring Helicobacter pylori, but not in any of 19 monkeys in which these bacteria were not seen. Gastritis, characterized by accentuation of lymphoplasmacytic infiltrates in antral and to a lesser degree cardiac mucosa, occurred in 42 of 63 (67%) monkeys evaluated and in 42 of 44 (93%) monkeys in which H. pylori was observed microscopically. Two monkeys with H. pylori infection had infiltrate scores that overlapped with the upper limit of scores of H. pylori-negative animals. Coincident with accentuated infiltrates were gastric gland epithelial hyperplasia, reduction in mucin content of surface and gland epithelia, and comparatively minor infiltrates of neutrophils in superficial lamina propria and gastric glands. Antral mucosa thickness often exceeded 1.5 to 2 times normal. Antral mucosal erosions occurred in 7 of 44 (16%) monkeys with H. pylori. Argyrophilic bacteria morphologically consistent with H. pylori were present in antral and less commonly cardiac mucosal glands. Intensity of bacterial colonization correlated with lymphoplasmacytic infiltrates (r.754) and hyperplasia (r.7), although responses were quite variable. These bacteria were not detected in fundic mucosa except in instances where parietal cells were substantially depleted in glands coincident with localized increases in lamina propria inflammatory cell infiltrates. Helicobacter heilmannii-like organisms (HHLOs) were present in fundic glands of all 63 monkeys; colonization was often pronounced. Scores for fundic mucosal inflammation did not correlate with presence or intensity of colonization with HHLOs (r.5). Rather, fundic inflammation scores positively correlated with the antral inflammation scores (r.548). Bacteria morphologically, biochemically, and genetically consistent with H. pylori were cultured from gastric mucosal specimens confirming bacterial identification. These findings demonstrate that adolescent cynomolgus monkeys are susceptible to natural infection with H. pylori and develop many morphologic hallmarks of H. pylori-related gastritis in humans. Key words: Animal model; bacteria; gastritis; Gastrospirillum hominis; Helicobacter; Helicobacter heilmannii; Helicobacter pylori; monkey; nonhuman primate; spirochetes; stomach. Spirochetes or spiral-shaped bacteria have long been recognized as constituents of gastric microflora in various mammalian species. 8,29,36,48,55,61,62 More recently, certain gastric spiral-shaped organisms have been established as pathogens responsible for subclinical as well as clinically apparent gastrointestinal disease in animals and humans. 38,4 These pathogenic spiralshaped organisms include several species of the Helicobacter genus including H. pylori (formerly Campylobacter pylori), H. mustelae, H. canis, H. felis, H. acinonyx, H. muridarum, and H. heilmannii (also called Gastrospirillum hominis or Gastrospirillum-like organisms). 38 Most Helicobacter species do not have strict host specificity. Rather, certain species have been isolated from multiple host species and experimental transmission among animal species has been shown. 19,22,31,37,41,52 Suspected zoonotic transmission has also been reported. 31,58 Helicobacter species have been isolated from saliva, 23 gastric fluid, 28 and feces 26,57 of infected animals or humans. These secretions or excretions likely play a role in natural transmission of these organisms. 1

2 2 Reindel, Fitzgerald, Breider, Gough, Yan, Mysore, and Dubois Vet Pathol 36:1, 1999 Gastric pathology associated with Helicobacter infections ranges from absence of noticeable mucosal injury to severe gastric disease characterized by mucosal mononuclear to mixed cell inflammation, erosions, and ulcerations. Such differences in gastric pathology may be influenced by the Helicobacter species 18 and strain, 12,14,22,52 and the mammalian host. 2,22,46 H. pylori-related disease is perhaps the most thoroughly investigated of the gastric diseases caused by Helicobacter species. 2 Pathogenic factors with proposed roles in H. pylori-related gastric disease include various bacterial products such as vacuolating toxin, 5,24,56 a protein that inhibits acid secretion, 6 and several bacterial enzymes including urease, catalase, oxidase, protease, and phospholipase. 12,49 Host factors including mucosal immunologic reactions 21,51,59 and cytokine release from gastric mucosal cells 51,54 also likely play a role in disease development. The pathogenesis of gastric disease with other Helicobacter species is not clearly defined. H. pylori and H. heilmannii (or H. heilmannii-like organisms [HHLOs]) are the Helicobacter species most commonly identified in the gastric mucosa of nonhuman primates and humans. 4,45 Dual infections with these organisms have been reported. 15,18,45,47 These organisms can be differentiated based on bacterial morphology and location within the gastric mucosa. H. pylori has a comma, loose spiral, or gull-wing shape, measures approximately 2 4 m in length, and resides largely in the antral mucosa. Bacterial isolation andinvitrocultureofh. pylori are possible under microaerophilic conditions. In contrast, HHLOs are approximately 7 1 m long and have a tightly wound spiral shape consisting of approximately 4 12 coils. 38 HHLOs typically colonize fundic mucosa and are visible within parietal cells. 15 Attempts to culture H. heilmannii in vitro have generally not been successful. 27,5 Thus, diagnoses of H. heilmannii infection in animals are based on appearance of morphologically consistent organisms in histologic sections. Whether organisms currently classified as H. heilmannii represent one or more bacterial species is unclear. 5 Gastric disease has been attributed to both organisms, but incidence and intensity of gastric disease tends to be greatest in primates infected with H. pylori. 7,15,18,25,34,35 An additional Helicobacter species, H. nemestrinae, has been isolated from the stomach of a pigtailed macaque and was considered to be nonpathogenic. 3 Several reports have documented the natural occurrence of Helicobacter-related infections, particularly H. pylori, in nonhuman primates. Rhesus monkeys (Macaca mulatta) are commonly referenced, perhaps because they are used extensively in biomedical research. 1,15,16,22,32,43,47 Single reports have also documented the natural occurrence of H. pylori or H. pylori-like organism (HPLO) infections in pigtailed macaques (Macaca nemestrina), 4 baboons (Papio papio), 6 and cynomolgus monkeys (Macaca fascicularis). 41 The paucity of reports for these other monkey species might suggest that occurrence of infection in these species is unusual. In two limited surveys of rhesus and cynomolgus monkeys, infection was common in rhesus monkeys but was not detected in cynomolgus monkeys. 22,43 Furthermore, attempts to infect cynomolgus monkeys with a human-derived strain of H. pylori were not successful. 22 Such findings suggested that cynomolgus monkeys may be less susceptible to H. pylori infection than rhesus monkeys. In studies by other investigators, small numbers of cynomolgus monkeys were experimentally exposed to monkey-derived or human-derived H. pylori strains; gastritis developed indicating that cynomolgus monkeys were susceptible to infection. 41,52 To date, evaluation of gastric pathology associated with Helicobacter-related infection in monkeys has been restricted to endoscopic biopsy material or small numbers of necropsied monkeys. Our attention was drawn to this epizootic because an unusually high frequency of subclinical gastritis occurred in cynomolgus monkeys from several toxicology studies conducted at our facility. The pattern of gastritis was similar in xenobiotic-treated and control monkeys, suggesting a spontaneous rather than chemical-induced disease. Organisms consistent with H. pylori were identified by histology and culture. Our recent report that the sequence of the entire 16S rrna gene of one of these isolates was % homologous with isolates from humans demonstrates that these bacteria are H. pylori. 9 Additionally, gastric spirochetes consistent with H. heilmannii were observed in histologic specimens. This report characterizes the gastric lesions occurring in monkeys involved in this naturally occurring epizootic of H. pylori-related gastritis. Materials and Methods Colony-bred, adolescent cynomolgus monkeys originating from the Philippines were obtained from Hazelton Research Products (Alice, TX) and kept in individual cages at our laboratory which is accredited by the American Association for Accreditation of Laboratory Animal Care. All monkeys were serologically negative for Herpesvirus simiae, simian retrovirus, simian immunodeficiency virus, and simian T- lymphotrophic virus-1. Monkeys were routinely subjected to a series of evaluations that included physical examinations, fecal examinations, clinical chemistry hematology assessments, and tuberculin tests to monitor health status. Clinical evidence of gastric disease was not apparent in monkeys before or during study periods. All monkeys were approximately years old at time of necropsy. The present report is based on the evaluation of 63 monkeys derived from four separate studies conducted over a 6-

3 Vet Pathol 36:1, 1999 H. pylori Gastritis in Cynomolgus Monkeys 3 Table 1. Incidence of antral lesion scores relative to Helicobacter pylori status. Monkeys evaluated Neutrophilic infiltrates Mucosal erosions Scores Mononuclear infiltrates Epithelial hyperplasia Lymphoid follicles H. pylori Animals % * H. pylori Animals % * * Percentage of all monkeys evaluated; all other percentage values are relative to number of monkeys with each specific H. pylori status. month period. Monkeys included both controls and monkeys receiving proprietary drugs that caused no gastric toxicity. At study termination, animals were euthanatized, exsanguinated, and necropsied. Generally, stomachs were rapidly removed from the carcass and immersion-fixed in 1% neutral buffered formalin. Within 4 minutes, stomachs were incised along the greater curvature and placed in fresh formalin fixative for 24 hours. Following fixation, 2-cm transmural sections were trimmed from antral (pyloric), fundic (corpus), and cardiac regions and processed in paraffin. Sections were cut and stained with hematoxylin and eosin (HE) and one or more silver stains (Warthin Starry, Genta, 3 or modified Steiner s) for detection of spiral bacteria. Immunostaining with a commercially available anti-h. pylori antibody (Signet, Dedham, MA) was performed on selected specimens. Briefly, sections were deparaffinized, rehydrated, and treated with polyclonal rabbit anti-h. pylori antibody (prediluted) for 1 hour at room temperature, followed by sequential treatment with biotinylated link antibody (Vector, Burlingame, CA), the Avidin Biotin Complex kit (Vectastain ABC kit, Vector), and diaminobenzidine (Sigma, St. Louis, MO) substrate. All sections were counterstained with hematoxylin. For negative controls, the primary antibody was omitted. For a subset of monkeys in this study, gastric mucosal sections were collected prospectively for bacterial culture before tissue fixation. Gastric mucosal samples collected from four monkeys with histologically confirmed gastric disease were bacteriologically cultured for H. pylori. In a subsequent survey of monkeys from this colony, eight necropsied monkeys with histologically confirmed gastric disease were sampled and cultured. In both surveys, sections of mucosa and mucosal scrapings were collected from the antral region of all monkeys subjected to necropsy, placed in sterile-filtered 3% glycerol in saline, chilled on ice, and frozen at 7 C. Once histopathologic assessment of tissues confirmed the presence of gastric disease and HPLOs in gastric Table 2. Comparison of antral lesion score severity relative to Helicobacter pylori status. Mononuclear infiltrates Neutrophilic infiltrates Epithelial hyperplasia Lymphoid follicles Bacteria H. pylori (antrum) H. heilmannii (antrum) H. heilmannii (fundus) H. pylori Mean 6* 1* 5* 6* Range H. pylori Mean Range * Statistically different from H. pylori-negative group by Wilcoxon test at P.1. mucosa of specific animals, frozen samples were thawed, homogenized, and microaerophilic culture was conducted using Campylobacter chocolatized blood agar plates supplemented with trimethoprim, vancomycin, amphotericin B, and polymyxin B (Remel, Lenexa, KS) at 37 C in an atmosphere of 9% N 2,5%O 2, and 5% CO 2 (microaerophilic conditions). Bacterial isolates consistent with H. pylori in shape, colony morphology, enzymatic activity, and gram-negative staining grew within 7 1 days. Single colony isolates were subcultured on sheep blood agar plates supplemented with trypticase soya agar (Remel) and confirmed for enzymatic activity and Gram stain. Bacteria were frozen in 3% glycerol Brucella broth solution for subsequent characterization. HE-stained sections of each stomach region were scored in a blinded fashion for lesion severity and distribution. Bacterial colonization was generally not readily apparent in HE preparations. Thus, Genta-stained sections from each stomach region were used for evaluating bacterial colonization status and intensity scores. Intensity was graded from to 4 ( normal; 1 minimal; 2 mild; 3 moderate; 4 marked) for mucosal hyperplasia, mononuclear cell infiltration, neutrophil infiltration, erosions, and presence of H. pylori and HHLOs. Distribution characteristics for hyperplasia, infiltrates, and erosions were assigned numerical scores of 1 or 2 for multifocal or diffuse, respectively. Lymphoid follicles were evaluated as atrophic (1), quiescent (2), or active (3), and densities per region were given intensity scores. Products of intensity and distribution scores were calculated for determination of the lesion score. Pearson correlation coefficients (r values; P.1) were determined between the scores for intensity of H. pylori and HHLO colonization and intensity of several subjectively quantified histologic parameters including mononuclear infiltrates, neutrophilic infiltrates, mucosal epithelial hyperplasia, mucosal erosions, and lymphoid hyperplasia. Comparisons of histologic parameters between HPLO-positive and - negative groups were made using the Wilcoxon (rank sum) test with P values at.1. Results Background gastric morphology Tables 1 and 2 summarize incidence and intensity of gastric pathologic findings relative to H. pylori sta-

4 4 Reindel, Fitzgerald, Breider, Gough, Yan, Mysore, and Dubois Vet Pathol 36:1, 1999 Fig. 1. Gastric antral mucosa from a normal monkey (Fig. 1A) and a monkey in which H. pylori were identified (Fig. 1B). Mucosa in Fig. 1B is thickened due to diffuse epithelial hyperplasia and mononuclear infiltrates evenly distributed throughout the lamina propria. Notice the reduction in normal undulations of the antral surface and the prominent lymphoid follicles in Fig. 1B. HE stain. Bar 15 m. tus. These bacteria were not evident in histologic sections from 19 of 63 (3%) monkeys. Gastric antral mucosa from a monkey in which H. pylori were not detected is shown in Figs. 1A, 2A, and 3A. Gastric morphology of H. pylori-negative animals was generally consistent with findings in monkeys euthanatized in the preepizootic period. These monkeys had minimal infiltrates of small mononuclear cells sprinkled throughout the gastric lamina propria. Such infiltrates were predominantly located in antral and cardiac

5 Vet Pathol 36:1, 1999 H. pylori Gastritis in Cynomolgus Monkeys 5 Fig. 2. Superficial antral mucosa from a normal (Fig. 2A) and H. pylori-infected monkey (Fig. 2B). Notice infiltrates of small mononuclear cells in the lamina propria. The epithelium of HPLO-infected monkey has less prominent epithelial mucin. Small mononuclear cells and cell debris were also present within surface and gland epithelia. HE stain. Bar 28 m. Fig. 3. Antral glands from a normal (Fig. 3A) and H. pylori-infected monkey (Fig. 3B). Notice reduction in epithelial mucin, enlargement of epithelial cell nuclei, presence of intraepithelial lymphocytes, and lamina propria infiltrates of small mononuclear cells. HE stain. Bar 28 m.

6 6 Reindel, Fitzgerald, Breider, Gough, Yan, Mysore, and Dubois Vet Pathol 36:1, 1999 mucosa; minor mononuclear aggregates were uncommonly seen in superficial fundic mucosa. A small number of eosinophils and histiocytes were also present in the gastric lamina propria. Deep mucosal or submucosal lymphoid follicles were observed in all gastric regions but were most common and prominent near the gastroesophageal and gastroduodenal junctions. Such follicles were generally quiescent. Mononuclear infiltrate scores (intensity distribution) for H. pylori-negative animals ranged from to 2 for the sampled gastric regions. Seven of 19 (37%) monkeys had scores of 1 or 2. Four of 19 (21%) animals had focal to diffuse minimal epithelial hyperplasia in gastric glands of the pyloric antrum. Because of the potential for patchy colonization of H. pylori in gastric mucosa, infection in these animals could not be ruled out with certainty. Mucosal lymphoid follicle scores were less in animals in which these bacteria were not observed. The range of scores for this group, however, overlapped the lower range of scores for H. pylori-positive animals. Microbiologic findings Bacteria consistent with H. pylori were cultured under microaerophilic conditions from one of four monkeys with gastritis sampled in the present survey and from five of eight monkeys with gastritis in a subsequent survey of monkeys of this colony. In all instances, gram-negative, curved to gull-wing-shaped bacteria grew within 7 1 days to form pinhead-sized colonies with water-spray morphology. Bacteria were 3 6 m in length and had urease, oxidase, and catalase activity. The DNA sequence of the 16sRNA gene for one of these isolates was % homologous with published DNA sequences of the 16S rrna gene for human isolates of H. pylori. 9 H. pylori-associated gastric pathology Grossly, multifocal to locally extensive reddening of the antral fundic mucosa was observed in 7 of 44 (16%) monkeys in which H. pylori were observed microscopically. More commonly, however, gross abnormalities were not recorded. Mucosal discoloration correlated microscopically with gastric mucosal erosions, congestion, and/or hemorrhage. Microscopically, animals in which H. pylori were identified in gastric mucosa had lesions that predominated in the antral mucosa (Figs. 1B, 2B, 3B). Antral lymphoplasmacytic infiltrates were significantly increased (P.1) and were generally confined to the lamina propria. Submucosa and tunica muscularis were generally spared of substantial inflammatory reactions. Antral infiltrate scores ranged from 4 to 8 in 42 of 44 (95%) monkeys; the 2 remaining monkeys had a score of 2, which overlapped with the upper limit of scores from animals in which H. pylori were not seen. Increased mononuclear infiltrate scores correlated with microscopic evidence of these bacteria in antral sections. Infiltrates resulted in lamina propria expansion with separation of gastric glands. In areas with intense infiltrates, the mucosa was 1.5- to twofold thicker than normal (Fig. 1A, B). Intraepithelial lymphocytes were increased and plasma cells with Russel bodies were common throughout the lamina propria. Pyknotic nuclear fragments and cell debris were often increased in upper glandular and surface epithelium (Fig. 2A, B). Epithelial hypertrophy and hyperplasia in expanded proliferative zones of antral glands were coincident with the lamina propria inflammation. Mucin-producing cells of the surface, pits, and deep portions of antral glands consistently had reductions in the size of mucin globules (Fig. 3A, B). Nuclei of mucus cells were enlarged with reduced polarity. Comparatively minor neutrophilic infiltrates were observed in the superficial lamina propria, within epithelia between surface and glandular epithelial cells (transmigration), and within lumens of upper gastric glands in 8% of infected animals (Fig. 4). Such infiltrates were not common throughout sections and were often present only in a few glands per section. Disruption of gland integrity, mucosal erosions, and hemorrhage were occasionally observed where mucosal thickening, inflammatory infiltrates, and epithelial hyperplasia were pronounced. Reepithelialized antral erosions were observed in 7 of 44 (16%) H. pyloripositive monkeys. Erosions were typically focal in nature and did not penetrate the full thickness of the mucosa. Intensities of mucosal inflammation and epithelial hyperplasia were greatest near the gastroduodenal junction. Mucosal and submucosal lymphoid follicles were enlarged with active follicular centers. Significant differences in lymphoid follicle scores were evident between H. pylori-positive and H. pylori-negative groups. Cardiac mucosal lesions paralleled those of antral mucosa, although such changes and scores tended to be lower than in the antrum. Changes included glandular epithelial hyperplasia, increased lymphoplasmacytic infiltrates, decreased size of mucin globules in surface and glandular epithelial cells, and prominent lymphoid follicles. Neutrophilic infiltrates were uncommon and mucosal erosions were observed in only one monkey. The fundic mucosa had comparatively minor alterations relative to other mucosal regions. A noticeable increase in infiltrates of small mononuclear cells was present in the superficial lamina propria below the surface epithelia and between gastric pits of the fundic body, in 27 of 44 (61%) animals with H. pylori. These minimal to mild cellular infiltrates always coincided

7 Vet Pathol 36:1, 1999 H. pylori Gastritis in Cynomolgus Monkeys 7 with more pronounced infiltrates in the antral mucosa. In transition zones between the fundic body and the antral or cardia mucosa, inflammatory infiltrates and proliferative epithelial changes occurred in parallel with changes in the antrum and cardiac regions (Fig. 5). Inflammation and hyperplasia were most conspicuous in superficial mucosa. A positive correlation occurred between intensity of infiltrates in fundic mucosa and antral mucosa (r.548), suggesting that fundic inflammation represented an extension of processes from bordering regions. When inflammation in these transition zones was marked, parietal cells were segmentally reduced or absent, particularly in upper portions of glands. In most animals, significant microscopic changes were not evident in gastric glands of the fundic body, despite often marked infection with HHLOs. Silver stains revealed two morphologically distinct spiral bacteria in the gastric mucosa. Both were immunocytochemically positive with the anti-h. pylori antibody. The smaller and less conspicuous, commaor gull-wing-shaped bacteria were m in length,.5 m wide, and consistent morphologically with H. pylori (Fig. 6A, B). These bacteria were evident in the antral mucosa of 44 of 63 (7%) monkeys. Among these H. pylori-positive monkeys, H. pylori were concurrently found in the cardiac mucosa of 15 of 44 (34%) monkeys and in the fundic mucosa of 6 of 44 (14%) monkeys. These bacteria were typically associated with antral and cardiac mucosal regions of inflammation and were located in gastric pits and the upper portions of gastric glands. In addition, these bacteria were often in intimate association with or appeared attached to the epithelial cell surface. These bacteria were not uniformly distributed throughout inflamed regions; bacteria were often difficult to find in substantial portions of inflamed mucosa. Positive correlations between the intensity scores of H. pylori colonization and the severity of mononuclear inflammation (r.754) and epithelial hyperplasia (r.7) were evident in the antrum, although the relationship in individual monkeys was variable. Some monkeys with very few bacteria in histologic section had pronounced inflammation, whereas other monkeys with large numbers of bacteria had more modest responses. Fig. 4. Superficial antral mucosa from an H. pylori-infected monkey. Notice infiltrates of neutrophils into gland lumen and surface mucus. Neutrophilic infiltrates were often present in only a few glands per mucosal section. HE stain. Bar 28 m. Fig. 5. Fundic mucosa at transitional zone between antrum and fundic body from a monkey with H. pylori infection. Notice infiltrates of small mononuclear cells in superficial lamina propria. Hyperplasia of fundic glands is evident as expansion of proliferative region and reductions in parietal cells in upper portion of the glands. The bases of glands appear relatively normal. HE stain. Bar 57 m.

8 8 Reindel, Fitzgerald, Breider, Gough, Yan, Mysore, and Dubois Vet Pathol 36:1, 1999 Fig. 6. Antral mucosal gland showing comma-shaped H. pylori in gland lumen and mononuclear infiltrates in lamina propria (Fig. 6A). Notice lymphocyte within the epithelium and close association of many bacteria with epithelial surface. Higher magnification of bacteria is shown in Fig. 6B. Genta stain. Bar in Fig. 6A 23 m. Bar in Fig. 6B 5 m. Fig. 7. Fundic mucosal gland showing large numbers of spiral HHLOs in gastric pits, fundic glands, and within parietal cells (Fig. 7A). Notice a paucity of inflammatory infiltrates in region. Higher magnification of bacteria is shown in Fig. 7B. Genta stain. Bar in Fig. 7A 23 m. Bar in Fig. 7B 5 m.

9 Vet Pathol 36:1, 1999 H. pylori Gastritis in Cynomolgus Monkeys 9 When H. pylori were evident in fundic mucosa, they were identified in the inflamed transitional zone of the antrum fundus where morphologically altered fundic glands had substantial reductions in parietal cells, increases in undifferentiated epithelial cells of neck region and upper gland, and corresponding prominent mononuclear infiltrates in the superficial lamina propria. H. pylori were not recognized in normal or nearnormal fundic glands in monkeys of this study. In antral mucosa, H. pylori were most commonly observed in antral glands bordering the transitional zone of the antrum fundus. These bacteria generally did not occur in great numbers near the gastroduodenal junction even when inflammation was pronounced. Organisms morphologically consistent with H. heilmannii (HHLOs) were observed in the stomach of all 63 monkeys (Fig. 7A, B). These tightly spiralled organisms measuring up to 8 m in length were present within the gland lumens, parietal cells, and surface mucus of the fundic mucosa. Six to eight spirals were evident along the length of many of these organisms. Numbers were diminished deep in gastric glands where parietal cells were less numerous. In heavy infections, organisms filled the cytoplasm of parietal cells, yet had no obvious detrimental effects on cell morphology. The proportion of organisms in parietal cells relative to gland lumens differed among animals. Small numbers of these organisms were identified in mucus and superficial glands of antral mucosa in 11 of 63 (17%) monkeys and of cardiac mucosa in 2 of 63 (3%) monkeys. The correlation coefficient for intensity of HHLO colonization and fundic inflammation was not significant (r.5). Discussion The present study demonstrates that cynomolgus monkeys can be naturally colonized by H. pylori. That the bacteria visualized in these animals were H. pylori is indicated by their morphology and location in tissue section as well as the in vitro behavior of these bacteria including microaerophilic growth characteristics, colony and bacterial morphology, and biochemical characteristics. In addition, we recently demonstrated that the sequence of the entire 16S rrna gene of one of these isolates was % homologous with published sequences for eight strains isolated from humans, whereas homology to H. acynonix was 97% and homology to other Helicobacter species was 9%. 9 Furthermore, the present study shows that infection with H. pylori is associated with antral gastritis and gastric erosions as in naturally infected rhesus monkeys and humans 2,18,4 and in experimentally infected cynomolgus monkeys. 52 Gastric disease of the type observed in these monkeys had not previously occurred at our facility, supporting the proposed epizootic nature of this outbreak. The animals were all relatively young ( 3.5 years old) at the time of sacrifice, indicating that prevalence of H. pylori infection can be high (7%) in colonies of young animals and is associated with pronounced gastric inflammation, epithelial hyperplasia, and gastric erosions. The age of our monkeys may have precluded development of certain chronic manifestations of H. pylori-related gastric disease that occur in humans. In humans, infection with H. pylori has been associated with gastric gland intestinal metaplasia 1 and neoplasia including gastric adenocarcinoma 44 and lymphoma. 63 Such manifestations presumably relate to chronicity of infection; infection in humans is often present for years or decades before clinical symptoms become manifest. 55 In the present study, we saw no evidence of intestinal metaplasia within the gastric mucosa or gastric neoplasia. Chronic active duodenal inflammation and ulceration are also associated with H. pylori infection in humans, 2 but these were not observed in these monkeys. Infection of cynomolgus monkeys with HHLOs was pervasive in our colony. A high incidence of infection with HHLOs has been reported in rhesus monkeys. 8,18,47,48 Infection in a small group of baboons has also been documented. 6 This organism, although often quite numerous in fundic glands and parietal cells, has not been associated with substantial fundic inflammation in nonhuman primates. 8,18,47,48 Subtle damage to infected parietal cells has been reported. 18 This organism has been associated with gastric disease in humans, but the inflammation is usually less, and no intestinal metaplasia or atrophy has been observed. 7 In addition, incidence of infection in the general population is quite low ( 2%) and capacity to induce gastric disease is also considered low. 42,45,5 Areviewof monkey gastric mucosal sections collected before this epizootic showed that infection with this organism was common and not associated with significant gastric inflammation. Whether the minor residual mononuclear cell infiltrates occasionally observed in the upper lamina propria of the fundus of monkeys are a consequence of infection with these organisms is uncertain. However, our findings appear to confirm the relatively nonpathogenic nature of HHLOs in nonhuman primates. Natural infections of rhesus monkey colonies with H. pylori have been reported and prevalence in such colonies can be high based on serologic surveillance. 15,16,18,22,32 Reports of H. pylori-related gastric disease in cynomolgus monkeys, however, have been limited to experimental studies in which small groups of cynomolgus monkeys were inoculated with monkeyderived or human-derived H. pylori strains. 41,52 Based on endoscopic biopsies, bacteria colonized the gastric mucosa and induced gastritis with erosions. 52 Our find-

10 1 Reindel, Fitzgerald, Breider, Gough, Yan, Mysore, and Dubois Vet Pathol 36:1, 1999 ings confirm that H. pylori can be associated with significant subclinical gastric disease in cynomolgus monkeys. This finding has relevance for toxicity testing in nonhuman primates because cynomolgus monkeys are now routinely used for safety evaluation of drug candidates. In such studies, subclinical disease induced by H. pylori can be problematic and confound diagnosis of potential compound-related toxicity to the gastric mucosa. Recognition of H. pylori infection and associated gastritis is important to prevent inadvertent ascription of gastric inflammation as a compound effect. Furthermore, once infection is identified, steps can be taken to eliminate infection and associated gastritis in the colony. 13 Endoscopy and biopsy sampling of the gastric mucosa have been successfully used to evaluate groups of rhesus and cynomolgus monkeys for H. pylori infection and to monitor presence and progress of the experimentally induced disease. 15,18,22,41,52 Although labor intensive, this procedure is an accurate and reliable procedure for diagnosing and monitoring colony infection in vivo. The patchy distribution of H. pylori in antral mucosal sections of infected animals indicates that organisms could be missed in limited biopsy sampling of individual animals and may even be missed when larger specimens are taken at necropsy. Thus, sensitivity of this procedure can be limited. In the present study, some animals categorized as histologically negative for H. pylori may have indeed had low-grade and patchy colonization of the gastric mucosa. This may explain why several H. pylori-negative animals had some, although minor, evidence of gland hyperplasia and accentuated mucosal infiltrates. Silver stains are useful and perhaps necessary for detection of bacteria in many circumstances. In our experience, the antral mucosa bordering the transition zone with the fundus was an area where organisms could most readily be detected microscopically. Mucosal inflammation, however, was most pronounced near the gastroduodenal junction. Bacteria could not always be identified in inflamed mucosa near the pylorus and were typically less numerous than in the antrum near the fundic transition zone. If endoscopic biopsy is used for surveillance, multiple biopsy sections should be collected throughout the gastric antrum for confirmation of infection. 1 If antral mucosal inflammation and hyperplasia comparable to that observed in this review are seen in biopsy specimens and bacteria are not evident in silver-stained sections, additional biopsies or procedures may be worthwhile to document infection. Our findings also indicate that fundic sampling at biopsy or necropsy is not generally useful for detection of infection. Procedures used clinically to identify H. pylori infections in humans may not be applicable to nonhuman primates because of the high infection rate of various monkey species with HHLOs. For example, a tissue urease test is used in humans with clinical symptoms of gastritis for detection of bacterial urease activity in gastric mucosal biopsy specimens. 33 Urease is produced by both H. pylori and H. heilmannii. 12 Lack of test specificity is less problematic in humans because prevalence and pathogenicity of H. heilmannii infection are so low that positive results are largely considered diagnostic for H. pylori. Because of the nearly ubiquitous presence of urease-producing HHLOs in stomachs of cynomolgus monkeys, the tissue urease test cannot be reliably used to selectively detect H. pylori infection. As recently demonstrated in rhesus monkeys, the urea breath test has similar limitations in nonhuman primates because this test detects exhaled labeled carbon dioxide produced by gastric bacterial urease activity following oral administration of 13 C- or 14 C-labeled urea. 17,39 Serologic methods to detect serum antibody titers to H. pylori have been used to detect infection or exposure in humans 11,53 and may be of value in monitoring infected monkey colonies. 15,16 Commercially available test kits for detection of H. pylori antibodies come with human-specific reagents. Modification of such tests with monkey-specific reagents is necessary to improve accuracy of serologic tests for nonhuman primates. 15,32 To our knowledge, little is known of serologic responses of monkeys to H. heilmannii infections. Antigenic cross-reactivity between H. pylori and H. heilmannii could limit application of conventional serology in monkeys, but was not observed in one study. 15 In our experience, immunocytochemical staining with anti-h. pylori antibodies did not discriminate between H. pylori and H. heilmannii because both organisms stained positive. Similar cross-reactivity was observed with human biopsy material by other investigators. 35 Furthermore, we observed large amounts of positively stained material in the lumen and epithelia of fundic glands that did not correspond with identifiable bacteria. This material could represent shed bacterial antigen. Bacterial morphology and mucosal location of colonization were most useful in discerning infection and discriminating between these organisms. Neither bacteria was readily apparent in HE-stained sections unless bacterial numbers were great and sections of tissue were thin. Warthin Starry, modified Steiner s, and Genta stains were equally useful for detection of these argyrophilic bacteria in tissue sections even when bacterial numbers were low. Genta stain was particularly appealing as a stain for accentuating gastric spiral organisms in tissue section because tissue architectural details were maintained in conjunction with silver-stained bacteria. Conversely, histologic details of tissues are not readily discerned with Warthin

11 Vet Pathol 36:1, 1999 H. pylori Gastritis in Cynomolgus Monkeys 11 Starry or modified Steiner s stains. Culture of H. pylori from inflamed mucosa was confirmatory of infection, but is not necessary for each colony member once H. pylori infection in the group has been established. Bacterial isolation and culture of H. pylori from gastric mucosa obtained at necropsy can be difficult because bacteria are microaerophilic and ph sensitive. In the present study, we improved the recovery rate by reducing the time to mucosal biopsy sampling and specimen freezing after animal euthanasia. The source of the infection of these monkeys with H. pylori was not established. We presume that some monkeys attained by our facility were infected in the country of origin. Humans or rhesus monkeys are considered possible sources of infection. Animal handling practices of nonhuman primates in the United States make human-to-monkey transmission of these bacteria highly unlikely in this country. The extent or potential in-house spread of infection among cynomolgus monkeys was also not determined. Transmission via fecal contamination at our facility was unlikely as animals were housed in individual cages upon arrival. Cages were cleaned daily; individual food dishes were not shared and were regularly sanitized. In-house transmission was possible during administration of dosing suspensions via the use of gastric gavage tubes and oral speculae. Common gavage tubes were used for dosing monkeys in each individual dose group and, therefore, transfer of saliva and gastric contents were possible within each treatment group. However, not all animals within specific dose groups had identifiable infections or gastric lesions. In conclusion, an epizootic of lymphoplasmacytic gastritis with mucosal erosions was associated with the presence of H. pylori infection in a colony of cynomolgus monkeys. HHLO infection was ubiquitous in these cynomolgus monkeys, but was not associated with significant gastric disease. Infection with H. pylori was widespread and typified by histopathologic changes of H. pylori-related gastric disease as reported for rhesus monkeys and humans. This may represent a relatively new and naturally evolving disease entity in captive cynomolgus monkeys. Because infections are typically subclinical, disease prevalence could become widespread before colony infection is recognized, and colony infection can potentially complicate toxicologic testing programs. Acknowledgements We thank C. Schray, K. Toy, and N. Diepert for histochemical and immunohistochemical technique development and slide staining. We also thank M. Anstett for preparation of photomicrographs, and D. Altrogge, D. Paterson, and L. Dillon for veterinary and technical support during the studies. References 1 Bayerdorffer E, Oertel H, Lehn N, Kasper G, Mannes GA, Sauerbruch T, Stolte M: Topographic association between active gastritis and Campylobacter pylori colonization. J Clin Pathol 42: , Blaser MJ: Helicobacter pylori and the pathogenesis of gastroduodenal inflammation. J Infect Dis 161: , Bronsdon MA, Goodwin CS, Sly LI, Chilvers T, Schoenknecht FD: Helicobacter nemestrinae sp. nov., a spiral bacterium found in the stomach of a pigtailed macaque (Macaca nemestrina). Int J Syst Bacteriol 41: , Bronsdon MA, Schoenknecht FD: Campylobacter pylori isolated from the stomach of the monkey Macaca nemestrina. J Clin Microbiol 26: , Cover IT, Blaser MJ: Purification and characterization of the vacuolating toxin from Helicobacter pylori. JBiol Chem 267: , Curry A, Jones DM, Eldridge J: Spiral organisms in the baboon stomach. Lancet ii: , Debongnie JC: Helicobacter heilmannii alias Gastrospirillum hominis : un autre pathogene. Lettre de l Infectiologue 11(Suppl to No. 5):3 33, Doenges JL: Spirochetes in the gastric glands of Macacus rhesus and humans without definite history of related disease. Arch Pathol 27: , Doi SQ, Mysore JV, Yang M, Reindel J, Dubois A: Natural colonization of cynomolgus monkeys (M. fascicularis) by H. pylori: confirmation by sequence of 16S rrna. Gut 43(Suppl 2):A4 (5/132), Drazek ES, Dubois A, Holmes RK: Characterization and presumptive identification of Helicobacter pylori isolates from rhesus monkeys. J Clin Microbiol 32: , Drumm B, Perez-Perez GI, Blaser MJ, Sherman PM: Intrafamilial clustering of Helicobacter pylori infection. N Engl J Med 322: , Dubois A: Spiral bacteria in the human stomach: the gastric Helicobacters. Emerg Infect Dis 1:79 85, Dubois A, Berg DE, Fiala N, Heman-Ackah LM, Perez- Perez GI, Blaser MJ: Cure of Helicobacter pylori infection by omeprazole clarithromycin-based therapy in non-human primates. J Gastroenterol 33:18 22, Dubois A, Berg DE, Incecik ET, Fiala N, Heman-Ackah LM, Perez-Perez GI, Blaser MJ: Transient and persistent experimental infection of nonhuman primates with Helicobacter pylori: implications for human disease. Infect Immun 64: , Dubois A, Fiala N, Heman-Ackah LM, Drazek ES, Tarnawski A, Fishbein WN, Perez-Perez GI, Blaser MJ: Natural gastric infection with Helicobacter pylori in monkeys; a model for spiral bacteria infection in humans. Gastroenterology 16: , Dubois A, Fiala N, Weichbrod RH, Ward GS, Nix M, Mehlman PT, Taub DM, Perez-Perez GI, Blaser MJ: Seroepizootiology of H. pylori gastric infection in sociallyhoused rhesus monkeys. J Clin Microbiol 33: , 1995

12 12 Reindel, Fitzgerald, Breider, Gough, Yan, Mysore, and Dubois Vet Pathol 36:1, Dubois A, Mysore JV, Goldberg G, Cantet F, Megraud F: Detection of H. pylori colonization in Rhesus monkeys using the urea breath test. Gut 43(Suppl 2):A57 (6/ 197), Dubois A, Tarnawski A, Newell DG, Fiala N, Dabros W, Stachura J, Krivan H, Heman-Ackah LM: Gastric injury and invasion of parietal cells by spiral bacteria in rhesus monkeys. Are gastritis and hyperchlorhydria infectious diseases? Gastroenterology 1: , Dye KR, Machall GJ, Frierson HF, Guerrant RL, Mc- Callum RW: Ultrastructure of another spiral organism associated with human gastritis. Dig Dis Sci 34: , Eaton KA, Radin MJ, Krakowka S: Animal models of bacterial gastritis: the role of host, bacterial species and duration of infection on severity of gastritis. Int J Med Microbiol Parasitol Infect Dis 28:28 37, Ernst PB, Pecquet S: Interactions between Helicobacter pylori and the local mucosal immune system. Scand J Gastroenterol 26(Suppl 187):56 64, Euler AR, Zurenko GE, Moe JB, Ulrich RG, Yagi Y: Evaluation of two monkey species (Macaca mulatta and Macaca fascicularis) as possible models for human Helicobacter pylori disease. J Clin Microbiol 28: , Ferguson DA, Li C, Patel NR, Mayberry WR, Chi DS, Thomas E: Isolation of Helicobacter pylori from saliva. J Clin Microbiol 31: , Figura N, Guglielmetti P, Rossolini A, Barberi A, Cusi G, Musmanno RA, Russi M, Quaranta S: Cytotoxin production by Campylobacter pylori strains isolated from patients with chronic gastritis only. J Clin Microbiol 27: , Flejou JF, Diomande I, Molas G, Goldfain D, Rotenberg A, Florent M, Potet F: Human chronic gastritis with non- Helicobacter pylori spiral organisms (Gastrospirillum hominis): four cases and review of the literature. Gastroenterol Clin Biol 14:86 81, Fox JG, Blanco MC, Yan L, Shames B, Polidoro D, Dewhirst FE, Paster BJ: Role of gastric ph in isolation of Helicobacter mustelae from feces of ferrets. Gastroenterology 14:86 92, Fox JG, Lee A: Campylobacter-like organisms: their role in gastric disease of laboratory animals. Lab Anim Sci 39: , Fox JG, Lee A: The role of Helicobacter species in newly recognized gastrointestinal tract diseases of animals. Lab Anim Sci 47: , Freeburg AS, Barron LE: The presence of spirochaetes in human gastric mucosa. Am J Dig Dis 7: , Genta RM, Robason GO, Graham DY: Simultaneous visualization of Helicobacter pylori and gastric morphology: a new stain. Hum Pathol 25: , Handt LK, Fox JG, Dewhirst FE, Fraser GJ, Paster BJ, Yan LL, Rozmiarek H, Rufo R, Stalis IH: Helicobacter pylori isolated from the domestic cat: public health implications. Infect Immun 62: , Handt LK, Fox JG, Yan LL, Shen Z, Pouch WJ, Ngai D, Motzel SL, Nolan TE, Klein HJ: Diagnosis of Helicobacter pylori infection in a colony of rhesus monkeys (Macaca mulatta). J Clin Microbiol 35: , Hazell SL, Borody TJ, Gal A, Lee A: Campylobacter pyloridis gastritis I; detection of urease as a marker of bacterial colonization and gastritis. Am J Gastroenterol 82: , Heilman KL, Borchard F: Gastritis due to spiral shaped bacteria other than Helicobacter pylori: clinical, histological and ultrastructural findings. Gut 32:137 14, Jhala D, Jhala N, Lechago J, Haber MH: Comparative analysis of histopathological features of Helicobacter heilmannii and Helicobacter pylori gastritis. Lab Invest 76:57A, Kasai K, Kobayashi R: The stomach spirochete occurring in mammals. J Parasitol 6:1 11, Krakowka S, Morgan DR, Kraft WG, Leunk RD: Establishment of gastric Campylobacter pylori infection in the neonatal gnotobiotic piglet. Infect Immun 55: , Lee A, O Rourke J: Gastric bacteria other than Helicobacter pylori. Gastroenterol Clin North Am 22:21 42, Marshall BJ, Surveyor I: Carbon-14 urea breath test for the diagnosis of Campylobacter pylori associated gastritis. J Nucl Med 29:11 16, Marshall BJ, Warren JR: Unidentified curved bacilli in the stomach of patients with gastritis and peptic ulceration. Lancet i: , Masubuchi N, Takahashi S, Nakano M, Igarashi H, Saito S, Aoyagi T, Itoh T, Hirata I: Experimental infection of the cynomolgus monkey with Helicobacter pylori. Nippon Rinsho 51: , Mazzuchelli L, Wilder-Smith CH, Ruchti C, Meyer-Wyss B, Merki HS: Gastrospirillum hominis in asymptomatic, healthy individuals. Dig Dis Sci 38: , Newell DG, Hudson MF, Baskerville A: Naturally occurring gastritis associated with Campylobacter pylori infection in the rhesus monkey. Lancet ii:1338, Nomura A, Stemmerman GN, Chyou PH, Dato I, Perez- Perez GI, Blaser MJ: Helicobacter pylori infection and gastric carcinoma in a population of Japanese-Americans in Hawaii. N Engl J Med 325: , Queiroz DMM, Cabral MMDA, Nogueira AMMF, Barbosa AJA, Rocha GA, Mendes EN: Mixed gastric infection by Gastrospirillum hominis and Helicobacter pylori. Lancet 336:57 58, Radin JM, Eaton KA, Krokowka S, Morgan DR, Lee A, Otto G, Fox JG: Helicobacter pylori infection in gnotobiotic Beagle dogs. Infect Immun 58: , Reed KD, Berridge BR: Campylobacter-like organisms in the gastric mucosa of rhesus monkeys. Lab Anim Sci 38: , Sato T, Takeuchi A: Infection by spirilla in the stomach of the rhesus monkey. Vet Pathol 19(Suppl 7):17 25, Smoot DT, Mobley HLT, Chippendaele GR, Lewison JF, Resau JH: Helicobacter pylori urease activity is toxic to human gastric epithelial cells. Infect Immun 59: , 1991

13 Vet Pathol 36:1, 1999 H. pylori Gastritis in Cynomolgus Monkeys 13 5 Solnick JV, O Rourke J, Lee A, Paster BJ, Dewhirst FE, Tompkins LS: An uncultured gastric spiral organism is a newly identified Helicobacter in humans. J Infect Dis 168: , Sugiyama T, Awakawa T, Hayashi S, Hisano K, Yabana T, Kurokawa I, Yachi A: The effect of the immune response to Helicobacter pylori in the development of intestinal metaplasia. Eur J Gastroenterol Hepatol 1(Suppl):S89 92, Takahashi S, Igarashi H, Ishiyama N, Nakano M, Ozaki M, Ito M, Masubuchi N, Saito S, Aoyagi T, Yamagishi I, Itoh T, Hirata I, Shingaki M: Serial change of gastric mucosa after challenging with Helicobacter pylori in the cynomolgus monkey. Int J Med Microbiol Virol Parasitol Infect Dis 28:51 57, Talley NJ, Kost J, Haddad A, Zinsmeister AR: Comparison of commercial serological tests for detection of Helicobacter pylori antibodies. J Clin Microbiol 3: , Tarkkanen J, Kosunen TU, Saksela E: Contact of lymphocytes with Helicobacter pylori augments natural killer cell activity and induces production of interferon. Infect Immun 61: , Taylor DN, Blaser MJ: The epidemiology of Helicobacter pylori infection. Epidemiol Rev 13:42 59, Telford JL, Ghiara P, Dell Orco M, Comanducci M, Burroni D, Bugnoli M, Tecce MF, Censini S, Covacci A, Xiang Z, Papini E, Montecucco C, Parente L, Rappuoli R: Gene structure of the Helicobacter pylori cytotoxin and evidence of its key role in gastric disease. J Exp Med 179: , Thomas JE, Gibson GR, Darboe MK, Dale A, Weaver LT: Isolation of H. pylori from human faeces. Lancet 34: , Thomson MA, Storey P, Greer R, Cleghorn GJ: Canine human transmission of Gastrospirillum hominus. Lancet 343: , Uibo R, Salupere V, Krohn K: Autoimmune reactions to gastric mucosa in chronic gastritis: a review. Scand J Gastroenterol 26(Suppl 186):11 15, Vargus M, Lee A, Fox JG, Cave DR: Inhibition of acid secretion from parietal cells by non-human-infecting Helicobacter species: a factor in colonization of gastric mucosa? Infect Immun 59: , Weber AF, Sasa O, Sautter JH: Some observations concerning the presence of spirilla in the fundic glands of dogs and cats. Am J Vet Res 19:677 68, Weber AF, Schmittdiel EF: Electron microscopic and bacteriologic studies of spirilla isolated from the fundic stomachs of cats and dogs. Am J Vet Res 23: , Wotherspoon AC, Doglioni C, Diss TC, Langxing P, Moschini A, deboni M, Isaacson PG: Regression of primary low-grade B-cell gastric lymphoma of mucosa-associated lymphoid tissue type after eradication of Helicobacter pylori. Lancet 342: , 1993 Request reprints from Dr. J. F. Reindel, Department of Pathology and Experimental Toxicology, Parke-Davis Pharmaceutical Research Division of Warner-Lambert Co., 28 Plymouth Road, Ann Arbor, MI 4815 (USA).