Determination of Genes vacA and cagA in Helicobacter pylori and their Relationship with the Induction of Coccoid Forms in Isolates from Patients in Antioquia, Colombia

Abstract

Helicobacter pylori is a spiral bacillus that can also adopt a coccoid form, a morphology associated with antibiotic resistance and the ability to withstand unfavorable environmental conditions. The relationship between coccoid forms, virulence, and the development of gastroduodenal diseases remains unclear. Therefore, this study aimed to determine the association between virulence and the induction of coccoid forms in H. pylori isolates from patients in Antioquia, Colombia. DNA from 30 clinical isolates was extracted, and the vacA and cagA genes were amplified to classify strains into high, intermediate, or low virulence groups. Four methodologies—each with specific modifications—were evaluated to induce coccoid forms. Bacterial morphology was assessed by optical microscopy, and coccoid cells were quantified using ImageJ. The association between virulence level and coccoid induction was analyzed using the Mann–Whitney U test. Before induction, cultures contained more than 90% spiral forms. The solid medium protocol under aerobic conditions was identified as the fastest in promoting coccoid conversion. After induction, high-virulence isolates showed a higher proportion of coccoid forms (72.6%) compared with low-virulence isolates (49.6%). A significant association was observed between virulence level and the induction of coccoid forms (p = 0.014). The findings suggest a relationship between the presence of vacA and cagA genes and the ability of H. pylori to adopt coccoid morphologies as an adaptive response. However, variability among induction protocols may introduce methodological biases and contribute to divergent interpretations of the biological role of coccoid forms. Further studies are needed to determine whether recurrences represent recrudescence or reinfection and to clarify the role of coccoid forms.

Introduction

Helicobacter pylori (H. pylori) is a globally distributed bacterium whose prevalence reaches 85%–95% in low- and middle-income countries, compared with less than 40% in high-income regions [1,2]. It colonizes the gastric mucosa and duodenum, and its persistence depends on infection duration, environmental factors, host characteristics, and bacterial traits. H. pylori is associated with acute and chronic gastritis, peptic ulcers, and gastric cancer [1,3]. According to the World Health Organization (WHO), H. pylori is classified as a group 1 carcinogen [4].

H. pylori is a gram-negative spiral bacillus with 2 to 6 peritrichous flagella under optimal growth conditions, which include a pH of 1.5–5, temperatures of 35–37°C, a microaerophilic atmosphere (5–10% O₂, 5–10% CO₂, and 80–90% N₂), and incubation for 4–5 days [5]. In recent years, the bacterium has been observed in intermediate and coccoid morphologies. The intermediate form represents the transition between the bacillary and coccoid states and is characterized by flagella wrapping around the cell and the accumulation of extracellular matrix, which widens the cell wall and results in short, curved, and thick bacilli [3,5,6]. The coccoid form is the final stage of transformation and is classified into two types: A, or non-viable, with a rough and irregular surface associated with cell death, and B, or VBNC (Viable But Non-Culturable), with a smoother surface and smaller cells linked to bacterial persistence in nutritionally stressful environments [3,6]. Both intermediate and coccoid forms emerge in response to physical or chemical stress, such as aerobic conditions, temperatures below 35°C, antibiotic or PPI exposure, pH >5, or prolonged cultures [3,5,6].

The coccoid forms of H. pylori lose their ability to be cultured because their metabolic activity decreases, preventing reproduction, which has led to their classification as dead forms. However, it was later identified that they retain viability [7]. Although non-culturable, these forms survive in suboptimal environments and are associated with increased infections because they can remain undetectable by conventional techniques and resist antibiotic therapy [6-9]. Similarly, Brenciaglia et al. found that cultures containing 85–95% coccoid forms treated with amoxicillin did not show a reduction in colony-forming units, whereas those exposed to erythromycin or gentamicin showed nearly complete reduction. They concluded that coccoid forms are resistant to antibiotics targeting bacterial division (amoxicillin) but more susceptible to those acting on protein (gentamicin) or DNA synthesis (erythromycin) [8]. Likewise, Faghri et al. reported that amoxicillin has limited activity against coccoid forms compared with metronidazole and clarithromycin, emphasizing the need to eliminate both spiral and coccoid forms [9].

Understanding the virulence factors of coccoid forms is essential because several are associated with gastrointestinal diseases. Among the main factors is urease, encoded by the constitutive ureA gene, which is commonly used for the identification of H. pylori. Another important factor is the vacuolating cytotoxin A (VacA), encoded by vacA [10,11]. The toxicity of this cytotoxin depends on two variable regions: the signaling region (alleles s1 and s2) and the middle region (alleles m1 and m2). Additionally, the cytotoxin-associated gene A (CagA), encoded by cagA, is produced in acidic conditions [5,10,11]. The variable expression of cagA and vacA allows H. pylori strains to be classified into three groups: type I, the most virulent (cagA+/ vacAs1m1 or cagA+/ vacAs1m2); type II, the least virulent (cagA-/ vacAs2m2); and type III, with intermediate virulence (cagA+/ vacAs2m2 or cagA-/ vacAs1m1) [5].

Previous studies identified a relationship between coccoid induction and strain virulence, showing that type I isolates (high virulence) generate more coccoid forms than type II isolates (low virulence). Thus, both virulence types appear to participate in the morphological transformation, although the molecular mechanisms remain unknown [5,12,13]. Krzyżek et al. reported that after 1 hour of incubation in suboptimal conditions, type I isolates produced 83.0% coccoid forms compared with 9.92% in type II isolates, supporting this association [5]. Similarly, Poursina et al. found that amoxicillin-induced cocci expressed virulence genes at lower levels than bacillary forms, although the difference was not statistically significant [12]. Finally, Sisto et al. evaluated the expression of ureA, cagA, and vacA during the conversion to coccoid forms, showing that DNA and RNA levels decreased over time but retained integrity. Moreover, after 31 days of incubation, coccoid forms continued to express virulence genes [13].

The coccoid forms of H. pylori are not detected through culture because their reduced metabolic activity hinders growth, and their low replication generates insufficient DNA concentrations for molecular identification [7,14]. Some studies propose that the reduction in size may act as a pathogenicity marker, as it reflects adaptations that allow survival in hostile environments and potential damage to the host [3,5,6]. In H. pylori, coccoid forms express virulence genes associated with gastric diseases, which may contribute to recrudescence and dissemination; however, limited research has prevented confirmation of this premise [6,15,16]. Most available studies instead focus on clarifying whether these forms are viable or represent cellular death. In Colombia, it remains unknown whether circulating strains can adopt coccoid morphology, and this form is neither sought nor detected by current diagnostic methods, which could influence the high infection prevalence reported in the country (up to 69%) [17,18]. Therefore, this study aims to determine the relationship between the virulence genes vacA and cagA and the induction of coccoid forms in H. pylori isolates from Antioquian patients with gastroduodenal diseases.

Methods

Isolations, Strains, Growth Conditions, and Cryopreservation

The study used 30 H. pylori isolates from the Bacteria & Cancer Group biobank, derived from gastric biopsies collected during a previous study approved by the Ethics Committee of the Faculty of Medicine at the Universidad de Antioquia (Act 019 of 2019, Faculty of Medicine, Universidad de Antioquia). The methodology for culturing biopsies and cryopreserved isolates has been described in detail in a previous publication [19]. We provide a brief summary of this process.

Biopsies were obtained by endoscopy and transported in Brucella broth (BD^®) supplemented with 10% Fetal Bovine Serum (FBS) at 4°C within 12 hours. In the laboratory, samples were manually homogenized in 2 mL of 0.85% saline and cultured on Brucella agar (BD^®) supplemented with 0.4% IsoVitaleX (BD^®), 0.2% DENT (Oxoid^®), and 7% horse blood. Plates were incubated at 37°C under microaerophilic conditions for up to 15 days. Colonies with typical H. pylori morphology were evaluated for catalase, oxidase, and urease activity. Isolates positive for all three tests and showing Gram-negative spiral bacilli on modified Gram staining were identified as H. pylori. Colonies were subcultured onto a second plate under identical conditions to improve viability. Finally, harvested colonies were suspended in 1 mL of Brucella broth (BD^®) containing 20% glycerol and 10% FBS, transferred to cryovials, and stored at –80°C (Nuaire^®) [18,19].

For experimental use, 10 μL of each cryopreserved isolate was taken in Brucella broth with 20% glycerol and 10% FBS. Each isolate was subcultured on Brucella agar (BD^®) supplemented with 0.4% IsoVitaleX (BD^®) and 7% horse blood and incubated for up to 5 days at 37°C under microaerophilic conditions. The obtained colonies were re-seeded in a second Petri dish containing the same medium and identical incubation conditions to improve H. pylori viability. These subcultured H. pylori colonies (transparent, round, raised, and punctiform) were subjected to catalase, oxidase, and urease tests. A positive result for all three tests, along with modified Gram staining (fuchsin is used instead of safranin) showing typical Gram-negative spiral bacilli, confirmed the identification of H. pylori [18,19]. Experiments were conducted using colonies from this second subculture, verified as H. pylori. The reference strain H. pylori ATCC 43504, was used as a control in all experiments.

Determination of Virulence Genes

To determine the virulence of genes ureA, vacA (m1/m2 and s1/s2 regions), and cagA, H. pylori DNA was first extracted from each isolate using the DNAzol^® kit (Life Technologies LTD) following the manufacturer’s recommendations. The DNA concentration was quantified using the NanoDropOne-2000 Spectrophotometer (Thermo Fisher Scientific). The obtained DNA was stored at –20°C in TE buffer (10mM Tris and 1mM EDTA). The genes were then amplified by Polymerase Chain Reaction (PCR) using 12.5 μL of taq polymerase master mix (New England Biolabs, Inc), 2.5μL of primers (Table 1), 8μL of molecular grade water (Amresco^®), and 2 μL of extracted DNA (40 ng/μL) for a final volume of 25μL. PCR was performed in Labnet international^® thermal cyclers, following the specific conditions for each gene, which were previously standardized in the laboratory (Table 1). PCR results were visualized on a 3% agarose gel diluted in 5X TE buffer with 3 μL HydraGreen™ (ACT Gene, Inc). The gel was run for 2 hours (50 V and 80 mA), and the 50 bp molecular weight marker (New England Biolabs, Inc) was used. The bands were observed on the Bio-Rad Molecular Imager Gel Doc™XR+ imaging and documentation system with ImageLab™ software (Bio-Rad Laboratories, Inc).

Table 1. Primers Used to Amplify Genes and Standardized Laboratory Conditions for PCR.

Gene	Primers Forward (5’-3’)	Primers Reverse (3’-5’)	Size (bp)	Amplification conditions	Ref
ureA	GCCAATGGTAATATTAGTT	CTCCTTAATTGTTTTTAC	411	94°C for 30 s, then 40 cycles of 94°C for 30 s, 45°C for 1 min, 68°C for 5 min	[20]
cagA	TTGACCAACAACCACAAACCGAAAG	CTTCCCTTAATTGCGAGATTCC	183	94°C for 30 s, then 30 cycles of 94°C for 30 s, 55°C for 1 min, 68°C for 5 min	[21]
vacA s1	ATGGAAATACAACAAACACACAC	GTCAGCACACACACCACAC	259	94°C for 30 s, then 30 cycles of 94°C for 30 s,55°C for 30 s, 68°C for 5 min	[22]
vacA s2	ATGGAAATACAACAAACACACAC	GTCAGCACACACACCACAC	286
vacA m1	GGTCAGATTGCGAATGGG	CCATTGGTACCTGTAGAAAC	290
vacA m2	GGAGCCCCAGAGCAATATTG	CATAACTAGCCTGCCTGCAC	352

Temperature in Celsius; Time: seconds (s) and minutes (min).

Induction of Coccoid Forms

To induce the coccoid forms of H. pylori ATCC 43504, four methodologies were used to determine the most suitable for evaluating the 30 clinical isolates. Each methodology started from the second subculture. The experiments were performed in triplicate at different time points. From each experiment, a microscope slide with a bacterial smear was obtained, and the bacteria were stained using a modified Gram method. The ability of each methodology to induce coccoid forms was assessed by analyzing 300 bacteria per plate and comparing the morphological differences observed using the ImageJ program, version 1.8.0 (National Institutes of Health, USA). The selection of the best methodology was based on the following parameters: required time, risk of contamination, ease of observing the bacteria, and cost. The specific details of each methodology, including culture conditions and duration, are summarized in Table 2.

Table 2. Methodologies Evaluated for the Induction of Coccoid Forms in H. pylori.

Methodology
Variables	One: Tsugawa et al., with modifications [23]	Two: Costa et al., with modifications [24]	Three: Tsugawa et al., with modifications [23]	Four: Costa et al., with modifications [24]
Modifications	The broth used and the supplements	The broth used and the supplements	Instead of broth we used agar and the supplements	Instead of broth we used agar and the supplements
Objective	Culture under prolonged microaerophilic conditions	Evaluation under mixed conditions (microaerophilic and aerophilic)	Solid culture under prolonged time and microaerophilic conditions	Solid culture in aerobic conditions
Culture media	Brucella broth supplemented with 0.4% IsoVitaleX and 10% FBS	Brucella broth supplemented with 0.4% IsoVitaleX and 10% FBS	Brucella agar supplemented with 0.4% IsoVitaleX and 7% horse blood	Brucella agar supplemented with 0.4% IsoVitaleX and 7% horse blood
Materials used (container)	15 mL conical tubes transferred to 25 cm² cell culture flasks	15 mL conical tubes transferred to 25 cm² cell culture flasks	Petri dishes 94 x 16 mm	Petri dishes 94 x 16 mm
Inoculation	Three colonies in 5 mL	Three colonies in 5 mL	Subculture two	Subculture two
Initial conditions and duration	Microaerophilic, 37°C, and 100 rpm for 3-5 days	Microaerophilic, 37°C, and 100 rpm for 24 hours	Microaerophilic, 37°C, and 100 rpm for 5 days	Aerobics, 25°C, without shaking for 5 days
Final conditions and duration	Microaerophilic, 37°C, 100 rpm until 15 days	Aerobics, 25°C, 100 rpm until 3 days	Microaerophilic, 37°C, without shaking until 15 days	Aerobics, 25°C, without shaking until 3 days

Identification of Coccoid Forms

Two subcultures were seeded from the first subculture of each isolate to induce coccoid forms in duplicate simultaneously. After 5 days of incubation under optimal conditions, a slide was prepared with three fixed colonies to be stained with the modified Gram stain. The cultures were then subjected to methodology four for the induction of coccoid forms, and finally, three plates were made with three fixed colonies of prepared H. pylori with the modified Gram stain.

To evaluate the morphology, a count of 50 bacteria was made on each Gram-stained slide on a microscope (Olympus, Japan) with a 100X objective and immersion oil. Therefore, a total of 100 cells were counted per isolation before inducing coccoid forms and a total of 300 cells per isolation after inducing coccoid forms. The size of each bacterium was determined using ImageJ software. In the present study, we used the sizes defined in the protocol of Krzyżek et al. to classify the bacterium as coccoid (0.5–1μm) and bacilli (>1μm) [5]. Two researchers performed the recount, each analyzing 15 isolates. The researchers independently counted the strain of ATCC 43504 to avoid bias in the counting of isolates.

Statistical Analysis

For the classification of H. pylori morphologies before and after being induced to coccoid forms, the mean (x̅) ± standard deviation (SD) of the three Gram smears of each of the methodologies applied to the ATCC 43504 was calculated, and the Microsoft 365^® Excel program, version 2404, was used to calculate the average percentage of the triplicate of cocci and bacilli.

When applying methodology four to all clinical isolates in duplicate, the percentage of cocci and bacilli before and after induction was calculated for each isolate. The average of the duplicate was calculated using the Microsoft 365^® Excel program version 2404. The normal distribution of the data was analyzed with the Shapiro-Wilk test in the IBM SPSS Statistics^® software, version 28.0 (SPSS), where a p-value <0.05 was considered a non-normal distribution. The Student’s t-test was used, where a p-value <0.05 was considered significant to identify differences between the x̅ of the size of each isolate before and after inducing coccoid forms. Finally, to relate the induction of coccoid forms with the virulence profile, the x̅ of the sizes after inducing coccoid forms was implemented for all isolates with high virulence (type I) and low virulence (type II) [5], and the differences of both x̅ were evaluated using the Mann-Whitney U test with a p-value <0.05.

Results

Characteristics of the Population

The 30 H. pylori isolates were obtained from 20 patients with chronic gastritis, of which 90% (18/20) were non-atrophic and 10% (2/20) atrophic, with one diffuse and the other multifocal, both with intestinal metaplasia (Table 3).

Table 3. Sociodemographic and Histopathological Data of the Patients in the Study.

Patient characteristics and gastric histopathology findings
Patient Code	Age	Gender	Type of gastritis	Atrophic type	Neutrophilic Activity	Metaplasia
439	80	Female	Chronic	Diffuse atrophic	Non-activity	Yes
451	34	Female	Chronic	Non-atrophic	Moderate	No
452	44	Male	Chronic	Non-atrophic	Moderate	No
461	63	Female	Chronic	Non-atrophic	Non-activity	No
504	72	Female	Chronic	Multifocal atrophic	Moderate	Yes
505	37	Female	Chronic	Non-atrophic	Moderate	No
508	57	Female	Chronic	Non-atrophic	Active	No
511	29	Male	Chronic	Non-atrophic	Non-activity	No
512	38	Male	Chronic	Non-atrophic	Moderate	No
517	48	Male	Chronic	Non-atrophic	Moderate	No
519	36	Female	Chronic	Non-atrophic	Moderate	No
540	42	Female	Chronic	Non-atrophic	Non-activity	No
545	75	Female	Chronic	Non-atrophic	Moderate	No
547	26	Female	Chronic	Non-atrophic	Moderate	No
548	52	Male	Chronic	Non-atrophic	Moderate	No
550	74	Female	Chronic	Non-atrophic	Non-activity	No
573	42	Female	Chronic	Non-atrophic	Moderate	No
576	43	Female	Chronic	Non-atrophic	Moderate	No
577	49	Female	Chronic	Non-atrophic	Non-activity	No
583	61	Female	Chronic	Non-atrophic	Moderate	No

Determination of Virulence Genes

The ureA gene was amplified in all isolates, confirming the identification of H. pylori infection. The strain ATCC 43504 was verified to have the type I virulence profile with the cagA+/vacA s1m1 genes. On the other hand, out of the 30 isolates, 73.3% (22/30) were identified with a type I virulence profile (cagA+/vacA s1m1) and 26.7% (8/30) had type II, mainly because they were negative for cagA. No isolates had a type III virulence profile (Table 4).

Table 4. Virulence Profile Results of Each Isolate and the Morphology of H. pylori Before And After Inducing Coccoid Forms.

Isolates	Virulence a				Morphology b
	Genes				Before inducing coccoid forms		After inducing coccoid forms
	ureA	vacA	cagA	Virulence type	X̅ size ± SD (μm)	Shapiro-Wilk	X̅ size ± SD (μm)	Shapiro-Wilk
ATCC 43504	Positive	s1m1	Positive	I	1.90 ± 0.73	0.015	0.98 ± 0.58	<0.001
451A	Positive	s1m1	Positive	I	2.21 ± 1.18	< 0.001	0.85 ± 0.32	<0.001
452A	Positive	s1m1	Positive	I	2.93 ± 1.32	< 0.001	1.05 ± 0.46	<0.001
452C	Positive	s1m1	Positive	I	2.56 ± 0.86	0.87	1.04 ± 0.61	<0.001
461A	Positive	s1m1	Positive	I	2.63 ± 1.14	0.002	0.86 ± 0.33	<0.001
504A	Positive	s1m1	Positive	I	2.39 ± 0.91	0.106	0.84 ± 0.31	<0.001
505A	Positive	s1m1	Positive	I	2.00 ± 0.91	< 0.001	0.86 ± 0.29	<0.001
505C	Positive	s1m1	Positive	I	2.27 ± 1.05	0.003	0.79 ± 0.25	<0.001
508A	Positive	s1m1	Positive	I	2.00 ± 0.88	< 0.001	1.12 ± 0.41	<0.001
511A	Positive	s1m1	Positive	I	3.01 ± 1.14	0.393	1.10 ± 0.74	<0.001
511C	Positive	s1m1	Positive	I	2.19 ± 0.60	0.006	0.89 ± 0.36	<0.001
519A	Positive	s1m1	Positive	I	3.08 ± 1.22	0.458	0.87 ± 0.37	<0.001
519C	Positive	s1m1	Positive	I	3.72 ± 1.11	0.302	0.88 ± 0.34	<0.001
545C	Positive	s1m1	Positive	I	2.18 ± 0.79	0.380	0.93 ± 0.53	<0.001
547C	Positive	s1m1	Positive	I	2.17 ± 0.46	0.470	0.81 ± 0.31	<0.001
548A	Positive	s1m1	Positive	I	2.04 ± 0.75	< 0.001	0.93 ± 0.49	<0.001
548C	Positive	s1m1	Positive	I	1.87 ± 0.65	< 0.001	1.09 ± 0.61	<0.001
550A	Positive	s1m1	Positive	I	2.13 ± 0.83	0.281	1.00 ± 0.52	<0.001
550C	Positive	s1m1	Positive	I	2.15 ± 1,07	< 0.001	1.03 ± 0.45	<0.001
573A	Positive	s1m1	Positive	I	3.08 ± 0.80	0.03	0.91 ± 0.21	0.011
577A	Positive	s1m1	Positive	I	2.51 ± 0.92	0.005	1.14 ± 0.59	<0.001
577C	Positive	s1m1	Positive	I	1.95 ± 0.84	<0.001	0.81 ± 0.23	<0.001
583C	Positive	s1m1	Positive	I	2.24 ± 1.17	0.014	1.01 ± 0.45	<0.001
439A	Positive	s2m2	Negative	II	2.53 ± 0.85	0.091	1.36 ± 0.87	<0.001
512A	Positive	s2m2	Negative	II	2.07 ± 0.87	0.004	0.97 ± 0.31	<0.001
512C	Positive	S2m2	Negative	II	2.79 ± 0.73	0.012	1.26 ± 0.50	<0.001
517A	Positive	s2m2	Negative	II	2.09 ± 0.62	0.003	1.39 ± 0.74	<0.001
517C	Positive	s2m2	Negative	II	2.20 ± 0.90	< 0.001	0.99 ± 0.31	<0.001
540C	Positive	s2m2	Negative	II	2.03 ± 0.60	< 0.001	1,01 ± 0.38	<0.001
576C	Positive	s2m2	Negative	II	1.95 ± 0.77	< 0.001	0.96 ± 0.27	<0.001
576A	Positive	s2m2	Negative	II	2.73 ± 0.78	0.127	1.13 ± 0.41	<0.001

^aIt indicates the genes present or absent in each isolate. The polymorphism is presented for vacA. From these results, the isolates are defined as having virulence type I, II, or III. ^bX̅ of the sizes of each isolate before and after the induction of coccoid forms, where the normal distribution of all the data of each isolate was analyzed with the Shapiro-Wilk test, where a p-value < 0.05 did not follow a normal distribution and a p-value > 0.05 had a normal distribution.

Out of the 20 patients evaluated by histopathology, 15 were infected with H. pylori with type I virulence and five with type II. Out of the 18 patients with chronic non-atrophic gastritis, 66.7% (12/18) were infected with H. pylori type I and 22.2% (4/18) with type II. On the other hand, out of the two patients who had chronic atrophic gastritis with intestinal metaplasia, the diffuse one was infected with H. pylori type I and the multifocal one with type II.

Induction of Coccoid Forms

Before using each coccoid form induction protocol, strain ATCC 43504 was prepared with modified Gram stain, where Gram-negative bacilli characteristic of H. pylori was observed in 86% of bacteria (Figure 1). In methodologies one and two, it was difficult to quantify and analyze the morphology of H. pylori because the bacteria were grouped together, which prevented visualization (Figure 2a and 2b). On the other hand, in methodologies three and four, bacteria with defined morphology were observed, allowing the identification and measurement of the size of H. pylori (Figure 2c and 1d). Methodology three induced 46% of coccoid forms, but the cultures were more prone to contamination due to having a longer incubation time. Finally, methodology four reduced contamination of cultures by having a shorter incubation period, and it induced coccoid forms in 67.3% of the bacteria. For this reason, methodology four was implemented in the clinical isolates.

Figure 1.

Modified Gram. Modified Gram stain of ATCC 43504 at 100× magnification before coccoid-form induction. Circles labeled 1 show the bacillary form of H. pylori, 2 the intermediate form, and 3 the coccoid form. Overall, 86% Gram-negative bacilli and 14% cocci were observed.

Figure 2.

Results of the protocols for inducing coccoid forms in the ATCC 43504 strain. Circles labeled 1 indicate the bacillary form of H. pylori, 2 the intermediate form, 3 the coccoid form, and 4 a clump of bacteria. a) Methodology one: poorly defined Gram-negative cocci and bacilli of H. pylori are observed, which do not allow their morphological identification due to the presence of bacterial groups. b) Methodology two: poorly defined Gram-negative cocci and bacilli of H. pylori are observed, which do not allow their morphological identification due to the presence of bacterial groups. c) Methodology three: Gram-negative cocci of H. pylori are observed in 46% and bacilli in 54%. d) Methodology four: Gram-negative cocci of H. pylori are observed in 67.3% and bacilli in 32.7%.

Morphologies of Clinical Isolates

Before inducing coccoid forms, the mean size was >1.0 μm (Table 3), with a percentage of bacilli > 90% for both isolates with type I and II virulence (Figure 3). After inducing coccoid forms, isolates with types I and II virulence presented coccoid forms ≥50% (Figure 3). The Student’s t-test showed that there was a significant difference between the mean sizes of all isolates before and after inducing coccoid forms (p≤0.001). In contrast, the mean size of types I and II isolates differed, with a result of 0.95μm and 1.15μm, respectively.

Figure 3.

Virulence types of H. pylori isolates and the induction of coccoid forms. Percentage of cocci and spiral-shaped bacilli according to the types of virulence obtained. It is observed that before inducing coccoid forms, the isolates mostly have bacilli (> 90%). After the induction of coccoid forms, a higher number of these forms were obtained in isolates with type I virulence (72.6%) than in those with type II (49.6%).

Relationship of the Virulence Profile with the Induction of Coccoid Forms

In the isolates with virulence types I (including the control strain) and II, the average number of cocci and spiral-shaped bacilli obtained before and after inducing the coccoid forms was determined. It was found that the isolates with virulence type I induced more coccoid forms (72.6%), while type II presented fewer coccoid forms (49.6%) (Figure 3). Since the data did not follow a normal distribution (Table 3), the Mann-Whitney U test was applied to relate the type of virulence with the induction of coccoid forms, showing a statistically significant relationship (p=0.014).

Discussion

The results from methodologies one and two did not allow differentiation of bacterial morphologies because Gram staining showed clusters that hindered assessment, which contrasts with the findings of Costa et al. [24] and Tsugawa et al. [23]. Their studies induced coccoid forms of H. pylori in liquid medium through prolonged incubation and aerobic conditions, respectively, without reporting difficulties in the morphological analysis [23,24]. In contrast, methodologies three and four provided clearer morphological evaluation, allowing accurate measurement of bacterial size. Methodology four induced coccoid forms more rapidly, achieving morphological change after 3 days of aerobic incubation, whereas methodology three required approximately 15 days, as bacteria first needed to exhaust nutrients before transforming. Previous studies using solid media applied prolonged incubation under microaerophilic conditions for up to 7 days and confirmed that coccoid forms maintained viable cellular structures and preserved DNA integrity, although in reduced quantities [25,26].

H. pylori possesses virulence genes essential contributing to disease development. Type I isolates (high virulence) are associated with gastric diseases such as ulcers and gastric cancer, mainly due to the presence of the cagA gene and the vacA s1m1 polymorphism [1]. In countries with a high prevalence of H. pylori, type I isolates circulate more frequently [27-29], which agrees with our findings. This may be explained by the fact that type I isolates are associated with more severe clinical outcomes, leading symptomatic patients to seek medical care and increasing the likelihood of their detection. In contrast, type II isolates are often linked to asymptomatic infections, making their isolation less common in a population composed mainly of symptomatic individuals [30]. Previous studies also report that cagA+ isolates typically express vacA s1, whereas cagA- isolates express vacA s2. This pattern may account for the absence of type III isolates in our study [28,29].

Previous research has shown that the virulence of H. pylori is linked to the severity of the diseases it causes [3,6]. However, few studies have evaluated whether the virulence of isolates is associated with their ability to generate coccoid forms. In this study, a statistically significant relationship (p = 0.014) was found between the induction of coccoid forms and the virulence profile, showing that type I isolates induced more coccoid forms (72.6%) compared to type II isolates (49.6%), which is consistent with the findings of Krzyżek et al., where type I isolates induced 83.33% of coccoid forms compared to 9.92% in type II [5]. Additionally, Sisto et al. induced coccoid forms through prolonged incubation and identified that although their metabolic activity decreased, their DNA and RNA were not degraded, allowing these cocci to continue expressing virulence genes [13].

Similar studies observed that coccoid forms retain their complete cellular structure, that the cagA and vacA genes remain intact in the DNA, and that cagA continues to be expressed [12,31]. Benaissa et al. concluded that coccoid forms maintain their metabolism due to the preservation of cellular structures compatible with viability [25]. Furthermore, She et al. demonstrated that coccoid forms induced in antibiotic-containing media show decreased protein synthesis while retaining the ureA, vacA, and cagA genes. They also found that coccoid forms are capable of adhering to Hep-2 cells; therefore, they concluded that this morphology may have pathogenic potential, as it can remain viable in the host [16].

Coccoid forms can resist antibiotics used in H. pylori eradication, including amoxicillin, clarithromycin, and metronidazole. [7-9,32]. Faghri et al. reported greater resistance to amoxicillin due to cell wall changes and loss of division. In contrast, clarithromycin and metronidazole, which target protein and DNA synthesis, show higher eradication rates in coccoid forms [9]. Similarly, Brenciaglia et al. found that cultures containing 85–95% coccoid forms exposed to amoxicillin, erythromycin, gentamicin, and metronidazole showed varied eradication rates: amoxicillin had little effect on VBNC cells, while the others significantly reduced counts [8]. In addition to resisting treatment, certain antibiotics can induce coccoid forms. Faghri et al. also observed that amoxicillin induced coccoid morphology in 99.9% of the bacterial population after 72 hours, while clarithromycin and metronidazole induced coccoid forms in 42.5% and 59.6% of cells, respectively. These findings indicate that although amoxicillin is ineffective in eliminating coccoid forms, it promotes their formation [9].

Moreover, recurrence of H. pylori infection may result from recrudescence or reinfection. Recrudescence occurs when the same strain reappears after temporary suppression by treatment, while reinfection refers to acquiring the bacterium again from external sources [6,15,16]. At the global level, recurrence varies by prevalence: 10.9% in countries with high prevalence (>70%), 7.1% in intermediate-prevalence regions (30–70%), and 1.6% in low-prevalence areas (<30%) [33]. In Colombia, reported rates include 18.2% in Túquerres (2013) and 5.8% in Bogotá (2015), although these values may be underestimated because coccoid forms are not routinely investigated in gastric biopsies [15,34]. These forms may contribute to recrudescence, as they can persist after therapy and reemerge once antibiotic pressure ceases [33]. However, most studies on coccoid morphology have been conducted in vitro, limiting understanding of their clinical relevance and role in infection persistence [7,35].

This study is the first descriptive cross-sectional analysis in Colombia to associate the H. pylori virulence profile with its ability to form coccoid cells, showing that the standardized protocol generated over 50% coccoid morphology. Although once considered non-viable, recent evidence indicates that coccoid forms maintain cellular integrity and may express virulence factors [7,33]. Therefore, further research is needed to standardize induction methodologies and determine whether patient recurrences are driven by reinfection or the presence of coccoid forms [15]. Conducting this type of study regionally is crucial because the genetic diversity of H. pylori, its virulence patterns, and associated clinical outcomes vary across geographical areas [1,9,11]. Previous work has shown that factors such as local antibiotic use, host genetics, and environmental conditions shape the distribution of virulence genes and bacterial pathogenicity [11]. Moreover, non-invasive and most invasive diagnostic methods do not detect coccoid forms, underscoring the need to incorporate machine-learning tools capable of identifying the different H. pylori morphologies [34]. Finally, no studies in the continent have evaluated coccoid forms in histopathological samples, leaving their frequency and their association with gastroduodenal disease in the region unknown.

Limitations

The present study had some limitations. First, the virulence of the isolates was determined with the vacA and cagA genes only. It is important to highlight that other virulence genes, which were not evaluated in this study, may be related to the induction of coccoid forms. The virulence classification of H. pylori isolates did not include the vacA region i; therefore, it cannot be assured that this region is involved in the induction of cocci. Finally, a larger sample size is needed to obtain an equal proportion of type I and II isolates as well as identifying if type III isolates produce coccoid forms.

Conclusion

There are various methodologies for inducing coccoid forms, and the evaluation of the viability of the coccoid morphology could be biased between studies. This would lead to different hypotheses about the role of the coccoid forms of H. pylori in the infection. Therefore, it is necessary to evaluate the methodologies that have been used inducing coccoid forms to identify which method(s) can be universally implemented. In this study, methodology four proved to be more efficient and simpler compared to the others. It induced coccoid forms sooner and allowed for the bacteria to be clearly identified (ie, morphology, measurements, and quantification) through Gram staining. The results of this study demonstrated a significant association between the type I virulence profile and the induction of coccoid forms, which could be related to the recrudescence of infection and the increase in antibiotic resistance. It is necessary to carry out studies that describe the molecular mechanisms linking the presence of virulence genes with the morphological transformation of H. pylori under suboptimal conditions. Finally, similar research is needed in other regions of the country and the world to confirm the results of this study.

Glossary

H. pylori

Helicobacter pylori

ATCC

American Type Culture Collection

FBS

Fetal Bovine Serum

TE

Tris-EDTA (buffer used for DNA storage)

PCR

Polymerase Chain Reaction

VBNC

Viable but Non-Culturable

CODI

Comité para el Desarrollo de la Investigación

PPIs

Proton Pump Inhibitors

UdeA

Universidad de Antioquia

vacA

Vacuolating cytotoxin A

bp

Base pairs

cagA

Cytotoxin-associated gene A

Author Contributions

BSG: (https://orcid.org/0000-0003-1108-015X) provided the project’s original concept. BSG conceptualized the work and designed the study. BSG and TPC (https://orcid.org/0000-0001-5095-3289) conducted the project. JDA (https://orcid.org/0000-0001-8070-2093) collected and cultured the isolates. LCO (https://orcid.org/0009-0003-8873-4040) and JHB (https://orcid.org/0009-0008-2994-7384) conducted experiments and standardized procedures. LCO and JHB analyzed data, created databases, and crossed variables. LCO and JHB drafted and wrote the manuscript. TPC and BSG edited and polished the manuscript. All authors read and approved the final manuscript.

Ethics approval and consent to participate

This study was approved by the Institutional Review Board of the Faculty of Medicine of the University of Antioquia in Act 016 of 2019 and the participants signed the Informed Consent approved by the same committee. The research study was performed in accordance with the Declaration of Helsinki guidelines. All the materials and reagent sources used in this study are described in the methods section.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests

The authors declare no competing or financial interests.

Funding Statement

This study was funded by Convocatoria para Apoyar Trabajos de Grado en los Programas de Pregrado de la Escuela de Microbiología de la UdeA 2021 – Sede Central (CIEMBTG-04-2022) and by Programa Jóvenes Investigadores 2022 del CODI Universidad de Antioquia, grant number 2020 – 50698.