Retraction Note: Detection of Helicobacter spp. DNA in the colonic biopsies of stray dogs: molecular and histopathological investigations

Background 
In dogs, the gastric Helicobacter spp. have been well studied, but there is little information regarding the other parts of the alimentary system. The incidence of Helicobacter spp. infection in dogs is largely unknown and to our knowledge there are no data about their potential pathogenic role. In light of these considerations, the aims of this study were (i) to assess the prevalence of Helicobacter spp. in colonic biopsies of healthy and symptomatic stray dogs also (ii) we isolate and characterize helicobacters in canine colonic biopsies to compare the commonly used tests for the identification of Helicobacter spp. and to determine the occurrence of these species in dogs.


R E T R A C T E D A R T I C L E Background
Based on their preferential site of colonization and phylogenetic analysis of 16S and 23S rRNA genes, members of the genus Helicobacter are divided into gastric and enterohepatic Helicobacter spp. [1,2]. Helicobacter spp. can be either part of the normal flora or associated with inflammatory and proliferative disorders of the alimentary tract in humans and animals [1]. A pathogenic role of the gastric and enterohepatic Helicobacter spp. in dogs has not yet been well elucidated. However, canine gastric and enterohepatic Helicobacter spp. have been identified in humans with gastrointestinal and systemic disorders [1], suggesting a potential zoonotic role, and warranting further investigations.
Since the first isolation of Helicobacter pylori from the stomach of humans by Warren and Marshall [3] several additional Helicobacter species have been identified in the gastrointestinal tract of various mammalian species and birds [1,4]. Similarlyto humans, gastric infection with Helicobacter spp. has been associated with variable degrees of gastritis in various animals, including dogs, cats, ferrets, pigs, monkeys and cheetahs [5], gastric adenocarcinoma in ferrets naturally infected by H. mustelae [5] and in Mongolian gerbils experimentally infected by H. pylori [6].
The discovery of H. pylori in humans [7] and its relationship to gastritis, peptic ulcer, and gastric neoplasia has increased interest in gastric bacteria. Research has focused on four specific areas: (i) prevalence of gastric bacteria in dogs and cats; (ii) investigation of these domestic animals as a source of zoonotic Helicobacter infection of humans [8,9]; (iii) clinical significance of gastric Helicobacters in dogs and cats; and (iv) development of animal models of human disease.
Recent reports indicate that in the canine stomachs three different culturable helicobacters can be present and they are regarded as different species: H. felis, H.bizzozeronii and H.salomonis [10,11].
Enterohepatic Helicobacter spp., including H. bilis, H. canis, H. cinaedi and other unclassified Helicobacter spp. have been recovered from the faeces of healthy and diarrhoeic dogs [1,[12][13][14]. In healthy dogs, enterohepatic Helicobacter spp. were described to preferentially colonize the crypts of the cecum and colon [15]. However, despite the extensive literature on enterohepatic Helicobacter spp. infection associated with IBD in laboratory animals, the association between Helicobacter spp. infection and gastrointestinal inflammatory and proliferative disorders has not been established in dogs so far.
The goals of this study were to investigate the prevalence of Helicobacter spp. in colonic biopsies of pet dogs and its six different species in colonic biopsies of stray dogs, and to evaluate the effect of the Helicobacter spp. colonization status (heavily colonized, poorly colonized and uncolonized biopsies) on the presence of histological lesions and DNA sequencing.

Animal and samplings
Tissues from fifteen stray dogs (8 males and 7 females, age 6 month -10 years) were used in this study. Stray dogs were housed in the same research facility ( Table 1). None of the animals had been treated with antibiotics within 4 weeks before examination. Endoscopy and biopsy sampling were performed as part of clinical investigation with the owners' agreement. The study was approved by the Ethical Committee of of the Science and Research Branch, Islamic Azad University of Tehran. Tissue samples from the gastrointestinal tract of each dog were collected in a sterile manner, with care taken to avoid cross-contamination between different sites. From each stray dog multiple colonic biopsies were taken for PCR. After each sampling, the endoscope and the biopsy forceps were thoroughly cleaned, then sterilized using an activated aldehyde solution. Biopsies were immediately frozen and stored at −20°C until DNA extraction. Samples were fixed in 10% neutral buffered formalin and embedded in paraffin wax for histological examination.

Polymerase chain reaction (PCR) amplification
DNA was extracted from biopsies by using a DNA easy tissue KIT® (Qiagen, Hilden, Germany) according to the manufacturer's Instructions. The 16S rRNA gene of members of the family Helico-Bacteracea was amplified by primers H276f/H676r [16], by using a single PCR. Primers P17f/P17r [17] were used to amplify specifict protein of Helicobacter bilis. Primers Fe1F/Fe3R [18] were used to amplify urease gene of H. bizzozeronii. Furthermore, primers HSALF/ HSALR [19] were used to amplify HSP60 of H.salomonis and primers HP-for/HP-REV [20] were used to amplify the urease C gene of H.pylori. Primers canis3F/canis4R (EU233451) were used to amplify the HSp60 of H.canis. All PCR reactions were performed in an AIB Applied Biosystems Thermocycler (2720 thermal cycler, M/PCR/23), all the amplifications were performed in a final volume of 25 μl containing 5 μl of DNA extracted, 10.3 μl ultra-pure distilled water, 2.5 μl of 10×Tag polymerase buffer (Qiagen), 0.5 μl MgCl 2 (Qiagen), 0.2 μl of Tag polymerase enzyme and 0.5 μl of each primer (MWG), 0.5 μl DNTP (Qiagen), 0.5 μl DNA Nucleotide (Qiagen), 5 μl Q Solution (Qiagen). Samples were heated at 94°C for 15 min and 30 s, denaturation phase, followed by 30 s at primer annealing temperatures that were different for the species and the family Helicobacteraceae. Primer annealing temperature for the family Helicobacteraceae was 52°C, for H.bilis was 55°C, for H.felis was 52°C, for H. bizzozeronii was 54°C, for H. salomonis was 51°C, for H. pylori was 49°C and for H.canis was 52°C, followed by 30 s in 72°C (extension phase) this cycle repeated 35 times, followed by 7 min in 72°C, final extension step and repair ( Table 2). Ten microliters of each amplification product were analyzed by gel electrophoresis in 1.5% agarose gels in Tris-boric acid EDTA buffer (TBE).The gel was stained with ethidium bromide (0.5 mg/L) and examined under UV transilluminator for the presence of amplified DNA; the size of the expected fragments was compared to a 100 bp reference marker (Fermentas-e), (Table 3).

DNA purification and sequencing
PCR products (there were 53 PCR products) were purified by Sanger method and then sequenced on the ABI 3730XL DNA Analyzer which provides high quality of Inflammatory cell Infiltration e L P + Morphological changes Surface epithelial injury Mucosal fibrosis and atrophy

Histological evaluation
Formalin-fixed paraffin-embedded 4 μm sections of colonic biopsies were stained with hematoxylin and eosin (H&E) and Periodic Acid-Silver Methenamine (PEM). The biopsies after staining were scored by a single phathologist. These criteria included the evaluation of inflammatory (infiltrates of lymphocytes and plasma cells, eosinophils, neutrophils and macrophages in the lamina propria) and morphological changes (surface epithelial injury, crypt hyperplasia, crypt dilation, distortion, mucosal fibrosis and atrophy). The degree of severity of both inflammatory and morphological changes was scored as follows: normal (0), mild (+), moderate (++) and marked (+++).

PCR
Results of the overall PCR amplification of Helicobacter spp. DNA, and the group-specific canine colonic Helicobacter spp. are summarized in Table 3

DNA sequence analysis
Evolutionary relationships of taxa for H. canis, H. bilis and H.Salomonis were inferred using the UPGMA method [1]. The optimal tree with the sum of branch

Histological examination
Inflammatory changes in fiften cases there were various numbers of lymphocytes and plasma cells in the lamina propria of the colonic biopsies, occasionally admixed with lower numbers of neutrophils and rare eosinophils, whereas detectable aggregates of macrophages were found. Mild to severe lymphoplasmacytic and neutrophilic infiltration was found in colonic biopsies. (Figure 5A-D) Moreover, slides stained with Giemsa and Periodic Acid-Silver Methenamine ( Figure 5A and B), the samples coinfected with less than 3 different species of Helicobacter (according to the PCR results) that had mild and unclear infection with helicobacter spirochetes. Although the mean value of neutrophilic infiltration was higher in heavily colonized biopsies compared to uncolonized and poorly colonized ones, this difference was not statistically significant.

Morphological changes
Morphological changes of the colonic mucosa were present in 12/15 examined stray dogs with different degrees of severity. A statistically significant difference was found between heavily colonized colonic biopsies and Figure 1 Helicobacteraceae DNA was detected with single PCR in all the colonic biopsies: Evolutionary relationships of taxa for helicobacter canis the evolutionary history was inferred using the UPGMA, Kimura 2-parameter and MEGA5 methods. Figure 2 Helicobacteraceae DNA was detected with Single PCR in all the colonic biopsies: Evolutionary relationships of taxa for helicobacter bilis the evolutionary history was inferred using the UPGMA, Kimura 2-parameter and MEGA5 methods.

A R T I C L E
both uncolonized and poorly colonized colonic biopsies for the parameter of mucosal fibrosis/atrophy (P < 0.05; Tables 1 and 2) as well as lieber kuhn's glands secretion, crypt dilation, crypt hyperplasia and distortion were higher in samples with objective number of spirochetes than what there were in slides with poorly and non objective spirochetes, even though these differences were not statistically significant.

Discussion
Helicobacter spp. are usually found in large numbers in the stomach of dogs but few data are available on the presence of these bacteria in other parts of the alimentary system. To date, enterohepatic Helicobacter spp. have been identified by culture in the feces of dogs and there are no data about their spatial distribution within the canine intestinal tract [13]. The fastidious nature of most

A R T I C L E
Helicobacter spp. and the coexisting intestinal flora hinder their routine isolation from the intestine [23], therefore, we opted for a DNA sequence analysis approach to detect Helicobacter spp. in the canine alimentary tract. A combination of PCR, 16S rRNA gene cloning, sequencing and histopathology allowed us to identify specific regions of the alimentary tract that were colonized with Helicobacter spp. and to visualize morphologically intact microorganisms and their spatial distribution within the gastrointestinal tract of dogs, as previously reported in mice and children [24][25][26].
In present study a high prevalence of Helicobacteraceae (100%) with H. canis and H. bizzozeronii (100%) DNA was detected by PCR in examined colonic biopsies from healthy and symptomatic stray dogs. The incidence of other species including H. salomonis (80%), H. bilis (47%), H.felis (40%) and H. pylori (7%) were comparatively high to low, respectively. These results were in line with previous studies in which enterohepatic Helicobacter spp. were detected by PCR in 100% of the large intestines of a small group of healthy dogs [13]. Our study revealed higher outbreak of helicobacter spp. than some previous studies. Based on direct sequencing of the enterohepatic Helicobacter spp. PCR products, H. salomonis, H. bilis, H. felis and H. pylori were the most prevalent Helicobacter species present in the canine colonic biopsies that is in accordance with previous works [13,15].
The high incidence of enteric H. canis and H. bizzozeronii might be explained by anal-oral transmission due to typical canine social behaviour. The presence of H. canis and H. bizzozeronii in both cecum and colon is interesting from a zoonotic perspective since H. canis and H. bizzozeronii has caused infections in humans where contact with dogs was suspected [27]. H. bizzozeronii (100%) and H.salomonis (80%) were found in colon, supporting the idea that gastric Helicobacter spp. might transmit via oral contact [10,28]. The absence of gastric Helicobacter spp. in faecal samples is in agreement with previous studies [15].
Futhermore, Helicobacter spp. DNA was detected in the cecum and colon of all dogs. The degree of colonization assessed by histopathology showed variable amounts of positive bacteria in all the cecum and colon sections depending on the dog. In some of these samples, only a small amount of Helicobacter spp. DNA was detected by single PCR. The low level of colonization was reflected in Helicobacter spp. detected by PCR also histopathology had a comparable distribution and were localized either in the intestinal lumen at the mucosal surface admixed to other bacteria or within the intestinal crypts extending from the superficial to the deep portions of the crypts. This finding together with the occasional amplification of gastric Helicobacter spp. DNA and the absence of gastric Helicobacter spp. sequences in clone libraries likely reflects their relative scarcity in the large intestine compared to colonic Helicobacter spp. that suggests passive intestinal carriage of gastric Helicobacter spp. DNA.
In the present study, a significantly higher level of mucosal fibrosis/atrophy was found in heavily colonized colonic biopsies compared to uncolonized or poorly colonized ones. Fibrosis is considered a common consequence of IBD that results from the reaction of intestinal tissue to the damage induced by chronic inflammation but its pathogenesis has not been fully elucidated [29]. Since Helicobacter spp. have been considered as likely candidates for involvement in IBD pathogenesis [30][31][32][33] that fibrosis is a common IBD-associated lesion. The association between heavy Helicobacter spp. colonization and mucosal fibrosis/ atrophy found in this study support the role of enterohepatic Helicobacter spp. in the development of canine IBD [1,16,30,31,34]. Therefore, our data may well explain the discrepancy previous studies.

Conclusion
We have indicated that the crypts of the cecum and colon of healthy and symptomatic dogs are heavily colonized by Helicobacter spp.. Combined molecular and histological approaches demonstrated that Helicobacter spp. infection is rather common in colonic biopsies of healthy and symptomatic stray dogs, Helicobacter spp. Specialy H. canis, H. bizzozeroni, H.billis, H.felis and H. salomonis identified as the most common species. Additional studies are necessary to investigate if Helicobacter spp. are involved in cecal and colonic inflammatory or proliferative disorders in dogs, as demonstrated in other species.