In situ detection of urease-positive Helicobacter pylori-like organisms on swine gastric mucosa

Abstract

The objective of this study was to improve the visual localization of urease activity of Helicobacter pylori-like organisms (HPLO) on swine gastric mucosa by in vitro optimization of the urea concentration and pH indicator of a urease test reagent. Five 21-day-old conventional pigs were infected orally with HPLO (3 pigs) or Brucella broth alone (2 pigs). At 17 d after infection the pigs were euthanized and their stomachs excised and tested for HPLO by a modified urease test formulation sprayed onto the gastric mucosa, as well as confirmatory culture and isolation of HPLO from urease-positive sites. This study showed improved detection of HPLO in porcine gastric mucosa with the use of a modified urease test formulation containing 5% urea and the pH indicator bromocresol purple compared with the use of a conventional formulation of 2% urea and phenol red. This test can readily be applied to achieve a presumptive diagnosis of HPLO in cases of gastritis or gastric esophageal ulceration in pigs.

Helicobacter pylori, the prototype species of a group of bacteria that colonize the gastric mucosa of humans, was identified in 1984 (1). On the basis of 16S sequence analysis, more than 30 validated and provisional Helicobacter spp. have been described and isolated from a variety of mammalian species, including humans (2), dogs (3,4), cats (5), pigs (6), ferrets (7), cattle (8), and the cheetah (9). In humans, H. pylori is known to cause type B gastritis (10) and gastric and duodenal mucosal ulcers (10,11), which can progress to gastric carcinoma (10,12). Recently, an H. pylori-like organism (HPLO) has been associated with gastric ulcers in swine (13,14).

In histologic sections, H. pylori is a curved, gram-negative, motile rod 0.5 × 5 ×m with a characteristic “gull-wing” appearance. The fastidious microaerophilic growth requirements of these bacteria often make their detection by routine bacteriologic culture methods difficult. Because several Helicobacter spp. are urease-positive (15), a urease test is often used to distinguish HPLO from other bacteria in diagnostic samples (16). A related commensal gastric bacterium, H. heilmannii, is typically larger (0.5 to 1.0 × 10 to 12 μm) than HPLO and “corkscrew” in appearance, so can be morphologically distinguished from HPLO (17); in addition, H. heilmannii is located primarily in the fundic region of the gastric mucosa (18). Recently H. heilmannii has been cultured in vitro (19).

Bacterial urease hydrolyzes urea, resulting in ammonia production and a subsequent rise in the microenvironmental pH. This increase in pH is indirectly detected by a urease test reagent that contains urea and a pH indicator. The customary reagent formulation for the detection of cultured urease-producing organisms is 2% urea (w/v) and the indicator phenol red, which changes from yellow-orange to pink-red at a pH of 6.8 to 8.2. However, this indicator is problematic for use on Helicobacter-infected gastric mucosa because modest color changes cannot, in confidence, be distinguished from the normal color of the gastric mucosa.

Here we describe an improved detection and presumptive diagnostic method that allows for HPLO localization in situ on porcine gastric mucosa through the use of a modified urease test reagent formulation of 5% urea and the pH indicator bromocresol purple.

The urease test reagents were prepared fresh the day they were needed by adding various concentrations of urea (2, 3, 4, 5, 6, 7, 10, 12.5, 15, 17.5, or 20% w/v) to the pH indicators phenol red, bromothymol blue, bromocresol purple, and methyl red (0.1 or 0.05% w/v) in 0.01 M phosphate-buffered saline (PBS). The pH of each formulation was adjusted to the lower end of the known pH range for each indicator (20) with the use of 0.1 N or 1.0 N HCl (Table I). In duplicate tests, 100 μL of pH-adjusted HPLO culture was added to 100 μL of urease test reagent in 96-well microtiter plates. The plates were developed at room temperature. Absorbance readings at wavelengths of 595 and 655 nm were measured with the Bio-Rad plate reader (model 3550; Bio-Rad Laboratories, Mississauga, Ontario) at 2, 5, 10, 15, 20, 30, 45, and 60 min, and the plates were observed for color-change development.

Table I.

Specifications for the pH indicators used in the study

Indicator	pH range	Adjusted pH	Acid (color)	Base (color)
Phenol red	6.6 to 8.0	6.6	Yellow	Red
Bromothymol blue	6.0 to 7.6	6.0	Yellow	Blue
Bromocresol purple	5.2 to 6.8	5.2	Yellow	Purple
Methyl red	4.8 to 6.2	4.8	Red	Yellow

Cultures of HPLO (isolate 2662) were subcultured from a pig-derived HPLO (14) and grown for 24 to 48 h at 37°C with 10% CO₂ and constant agitation in sterile 95-mm petri plates containing 10 mL of sterilized Brucella broth [28.0 g of Brucella broth (Oxoid; Basingstoke, England) + 1000 mL of distilled water] and fetal calf serum (FCS) (Cansera International, Etobicoke, Ontario) to a final concentration of 10% (v/v). Cultures were confirmed to be devoid of contaminating organisms by Gram’s staining, microscopic examination, and growth of a 100-μL aliquot on Columbia III blood agar supplemented with 5% sheep blood (Becton Dickinson, Oakville, for 5 d. Serial 10-fold dilutions were Ontario) at 37°C with 5% CO₂ prepared in Brucella broth by the average of 3 hemocytometer counts of 1.0 × 10⁷ to 1.0 × 10⁵ HPLO/mL. Final concentrations of 1.0 × 10⁶, 1.0 × 10⁵, and 1.0 × 10⁴ HPLO per well were prepared by adding 100 μL of diluted sample to 100 μL of urease test reagent. Before testing, cultures (HPLO in Brucella broth) and negative controls (Brucella broth with 10% FCS) were adjusted to a pH of 4 by the addition of 0.1 N and 1.0 N HCl. Adjusted culture stocks were examined microscopically after 1 h to ensure culture viability.

Five 21-day-old conventional swine of unknown HPLO status were fasted for 18 h and then infected orally with 20 mL of Brucella broth containing 1.0 × 10⁶ HPLO/mL (3 pigs) or 20 mL of Brucella broth alone (2 pigs). Seventeen days after infection, the swine were sedated by intramuscular injection of ketamine HCl (40 mg/kg) and euthanized by intravenous injection of sodium pentobarbital (100 mg/kg). All were housed and euthanized according to the guidelines of the Ohio State University Institutional Laboratory Animal Care and Use Committee, and the experimental protocol was approved by that committee.

One liter of urease test reagent containing 0.1% bromocresol purple (w/v) and 5% urea (w/v) in PBS had been prepared 1 h before euthanasia and decanted into a spray bottle. The stomach of each pig was removed intact, the liquid contents were removed by compression, and the exteriors were rinsed with water to remove blood and debris, care being taken not to contaminate the stomach openings or allow water contamination of the inner mucosal surface, as water will react with the urease reagent and cause a false-positive result. The stomach was then opened around the greater curvature, starting at the pylorus and following the attachment of the gastrosplenic ligament. The stomach was placed open in a dissection tray with the mucosal side up, and debris was removed with forceps. Urease test reagent was liberally sprayed over the mucosal surface of the stomach, which then sat at room temperature for 5 to 7 min to allow for color development. Urease-positive colonization sites in the gastric mucosa were identified and then cultured for HPLO to confirm that the sites of enzyme activity contained HPLO.

The most rapid color development upon the addition of HPLO to the 96-well microtiter plate occurred with urease test reagents containing more than 5% urea (w/v) (Figure 1). Uniform color development was observed across all test reagents with more than 5% urea. At an indicator concentration of 0.1% (w/v) the color change was faster with bromocresol purple than with phenol red, bromothymol blue, or methyl red (Figure 1, row F: columns 7 to 9). This concentration resulted in a darker, more pronounced, or better recognized color change in most of the indicators when compared with 0.05%. Detection was better with bromocresol purple than with bromothymol blue and phenol red because of the more rapid color change in the lower pH range provided by bromocresol purple (Table I). Methyl red was observed to change color rapidly, at a rate similar to that of bromocresol purple; however, the color change of methyl red from red to yellow was difficult to distinguish compared with the change with bromocresol purple from yellow to dark purple. All the negative controls (Brucella broth with 10% FCS) had no observed color change (data not shown).

Figure 1.

Results of adding 100 μL of cultured Helicobacter pylori-like organisms (HPLO) in Brucella broth (final concentration expressed as the number of HPLO per well) or Brucella broth alone, both adjusted to a pH of 4.0, to urease test reagents prepared with various urea concentrations, pH indicators, and indicator concentrations, in a 96-well microtiter plate. Rows A, B: Brucella broth alone; rows C, D: 100 000 HPLO; rows E, F: 1 million HPLO; rows G, H: 10 million HPLO.

The urease test preparation containing 0.1% bromocresol purple and 5% urea was observed to have a rapid and pronounced color reaction on the swine gastric mucosa. This pronounced reaction was evident in the in vitro comparison between the conventional (0.1% phenol red and 2% urea) and modified (0.1% bromocresol purple and 5% urea) urease test preparations (Figure 1, row F: columns 1 and 8). All urease-positive sites in the gastric mucosa were confirmed by culture and isolation to contain HPLO (data not shown). The stomachs of the 2 uninfected pigs did not produce urease-positive sites in the gastric mucosa (data not shown) and were therefore not cultured for HPLO.

For the detection of urease-producing bacteria, most published diagnostic protocols require testing of cultured bacteria with a reagent of 2% urea and the pH indicator phenol red. Our results suggest that a formulation of 5% urea allows the detection of urease-positive colonization sites in the gastric mucosa and a rapid presumptive diagnosis. Not only did our urease formulation result in a more rapid color change when compared with a 2% urea formulation, but the distinct color change from yellow to dark purple facilitated identification of colonization sites within the stomachs of the infected piglets. Uniform color development was observed for all test reagents containing more than 5% urea, indicating that the urea concentration is not the limiting component in the assay: increased concentrations above 5% did not reduce the reaction time or lower the detection limit of the assay.

The color change from yellow to dark purple produced by bromocresol purple was distinct on the mucosal surface of the swine stomachs (Figure 2) and provided a more definitive presumptive diagnosis than the color change produced by the previously used indicator, phenol red. However, it is imperative to avoid gross contamination with water or blood when using the formulation containing bromocresol purple, as this indicator can react with these fluids and result in a false-positive diagnosis. If necessary, a volume of water (pH-adjusted to less than 5.2) can be used to rinse the mucosal surface of the stomach to prepare for the urease reagent if blood or other contaminating material is present.

Figure 2.

Swine mucosa 7 min after being flooded with a urease test reagent containing 0.1% bromocresol purple and 5% urea. Sites of colonization by HPLO are easily identifiable as purple foci (arrows).

This rapid test can be applied to obtain a presumptive diagnosis of HPLO infection in cases of gastritis and gastroesophageal ulceration in swine. When applying the test it is important to consider that it will also detect other urease-producing bacteria, including H. heilmannii, a commensal bacterium that is typically found in the fundic region of the gastric mucosa, as distinct from the cardiac region, where HPLO is isolated (18). Histologic examination of the gastric mucosa will allow differentiation of H. heilmannii and HPLO; H. heilmannii was not found in the gastric cardia of the experimentally infected animals in this study.