Occurrence, Serotypes and Virulence Characteristics of Shiga-Toxin-Producing Escherichia coli Isolates from Goats on Communal Rangeland in South Africa

Shiga-toxin-producing Escherichia coli is a foodborne pathogen commonly associated with human disease characterized by mild or bloody diarrhea hemorrhagic colitis and hemolytic uremic syndrome. This study investigated the occurrence of STEC in fecal samples of 289 goats in South Africa using microbiological culture and PCR. Furthermore, 628 goat STEC isolates were characterized by serotype (O:H) and major virulence factors by PCR. STEC was found in 80.2% (232/289) of goat fecal samples. Serotyping of 628 STEC isolates revealed 63 distinct serotypes including four of the major top seven STEC serogroups which were detected in 12.1% (35/289) of goats: O157:H7, 2.7% (8/289); O157:H8, 0.3%, (1/289); O157:H29, 0.3% (1/289); O103:H8, 7.6% (22/289); O103:H56, 0.3% (1/289); O26:H2, 0.3% (1/289); O111:H8, 0.3% (1/289) and 59 non-O157 STEC serotypes. Twenty-four of the sixty-three serotypes were previously associated with human disease. Virulence genes were distributed as follows: stx1, 60.6% (381/628); stx2, 72.7% (457/628); eaeA, 22.1% (139/628) and hlyA, 78.0% (490/628). Both stx1 and stx2 were found in 33.4% (210/628) of isolates. In conclusion, goats in South Africa are a reservoir and potential source of diverse STEC serotypes that are potentially virulent for humans. Further molecular characterization will be needed to fully assess the virulence potential of goat STEC isolates and their capacity to cause disease in humans.


STEC O:H Serotypes
Among the 63 O:H distinct serotypes, 55.5%, (35/63) were each represented by a single isolate while the remaining 44.4% (28/63) were represented by more than one isolate (Table 1). O:H serotype combinations can be found in Supplementary Table S1.

Discussion
Previous reports from different countries have shown that goats are a reservoir of STEC [4,[39][40][41][42]. Furthermore, contact with goats and food products of goat origin have been associated with STEC disease in humans [4,43]. However, published reports on the occurrence and characteristics of STEC in goats are few in comparison to cattle and sheep. Furthermore, reports on the occurrence of STEC in goats in South Africa are non-existent. This study investigated the occurrence of STEC and characterized STEC isolates in four separate goat herds in South Africa. The overall occurrence of STEC in the goat populations surveyed was 80.2% (232/289). The occurrence of STEC in this study was very high in comparison to similar studies in Germany [42,44], Brazil (57.5%) [40], Spain (47.7%) [39,45], Vietnam (31.5%) [41] and Bangladesh (11.8%) [46] which reported STEC detection rates ranging from 11.8% to 75.3% in goats. Other reports have found STEC occurrence rates ranging from 23.9% to 89.3% in different countries, but these studies were conducted on far smaller goat sample populations (≤46) to warrant a valid comparison with the present study [47][48][49][50].
The within-herd occurrence of STEC ranged from 75.3% (116/154) to 90.6% (39/43) which was significantly higher in comparison to similar studies in Brazil (46.7-73.3%) [40] and Vietnam (15-65%) [41]. Moreover, all the four goat herds were positive for STEC, in agreement with similar reports elsewhere [40][41][42]. However, the number of goat samples which were tested per herd in this study was "significantly" higher compared to the reports from Brazil (106), Vietnam (205) and Germany (93) which may explain why the within-herd STEC occurrence in this study was also higher. The higher occurrence of STEC in goats in this study may be ascribed to higher shedding of STEC in the goat population studied, variations in geographic locations, age (kids vs. adults), goat diet (grazing or browsing vs. concentrate), and management practices. Furthermore, the use of a suitable enrichment broth and two selective and sensitive STEC culture and isolation media may have increased STEC recovery [51][52][53][54][55][56].
In the present study, 99.0% of goat STEC isolates were serotypeable by PCR. A total of 63 serotypes (34 O and 17 H groups) were recovered from goats. The number of serotypes detected in this study was very high compared to previous studies [39,40,45]. The recovery of a very high number of serotypes may also be ascribed to the high shedding of STEC in the goat populations tested. Furthermore, the use of a sensitive, specific, accurate and reliable PCR protocol for O:H serotyping may have led to the identification of more serotypes than usually found with traditional serotyping [57][58][59]. Furthermore, PCR O:H serotyping has the advantage of detecting O-untypable (OUNT) and H-nontypeable and/or non-motile (HNT/NM) E. coli isolates that carry genes encoding O:H antigens but cannot be expressed. In this study, we were able to validate the Iguchi et al. [58,59] and Banjo et al. [57] E. coli PCR serotyping (O:H) protocols which were highly discriminatory and unambiguously serotyped the large number of goats STEC isolates tested in this study [57][58][59]. To our knowledge, this is the most extensive serotyping of goat STEC isolates, worldwide.
Among the 63 serotypes, only 4 serotypes belonged to the major 7 STEC serogroups. STEC O103:H8 (15.6%) was the most frequent Big seven STEC among goats, followed by STEC O157:H7 (4.7%), O111:H8 and STEC O26:H2. Overall, the major seven STEC serotypes accounted for 21.3% of all isolates which were serotyped, in contrast to most similar studies which never recovered major seven STEC from goats [39,40,[43][44][45]60,61]. However, Schilling et al. [48] found a higher proportion of top seven STEC, although the recovered serotypes were those which have never been reported in human disease, in contrast to our results which showed that most of the top seven serotypes we recovered were previously incriminated in human disease outbreaks except for STEC O157:H29, O103:H8 and O103:H56.
Previously, STEC O157:H7 has been incriminated in foodborne disease after consumption of raw goat milk and home-made cheese made from raw milk [43,62,63]. Furthermore, STEC O157 and STEC O103 have been incriminated in human disease after contact with goats in the USA [8] while sources other than goats have frequently associated STEC O157:H7, O111:H8 and O26:H2 to human disease worldwide including South Africa [16]. According to the STEC seropathotype classification, STEC O157:H7 is considered a seropathotype A strain, frequently incriminated in outbreaks and severe human disease while O111:H8 and O26:H2 are moderately implicated in outbreaks and less frequent in severe human disease, in comparison to STEC O157:H7 [32]. However, in this study most major seven STEC isolates were classified as STEC O103:H8. Previously, STEC O103:H8 was isolated from healthy goats and calves in China and Argentina, respectively [64,65]. In addition, only one study has reported the recovery of STEC O103:H8 from patients and asymptomatic food handlers in Japan [66]. However, this study never specified whether the STEC O103:H8 isolate was from patients or asymptomatic food handlers [66]. Therefore, although O103:H8 is classified as a major STEC (serogroup), its importance as a human pathogen remains unclear as there are no reports until now which have unequivocally associated this STEC serotype with human disease.
The remaining 59 serotypes were non-O157, of which 24 have been previously incriminated in mild to severe human disease worldwide including South Africa, Europe, North America and Asia [10][11][12]16,67]. The recovery of STEC serotypes which have been associated with mild to severe human disease is evidence that goats are a reservoir and a potential source of these highly pathogenic STEC strains in South Africa.
Highly diverse and farm specific STEC serotypes were observed in individual goat herds except for STEC O76:H19 which was the serotype shared among the four goat herds surveyed while STEC O146:H21 and OgX18:H2 were recorded in three herds. Overall, the highly diverse and farm specific serotypes are most likely a reflection of the fact that the four herds were situated in geographically separate and distant areas from each other to allow isolate interchange between herds.
Regarding the virulence characteristics of the STEC isolates under study, stx2 was more frequent that stx1 among goat STEC in contrast to similar studies which have shown that stx1 is predominant among goat STEC isolates [39,41,42,45,50,[68][69][70][71]. However, our findings agree with a study by Oliveira et al. [40] which reported that stx2 was more prevalent in goat STEC isolates. Reports on clinical STEC have suggested that stx2-positive isolates are more virulent and frequently incriminated in severe human disease including hemorrhagic colitis and hemolytic uremic syndrome in comparison to STEC isolates carrying stx1 or both stx1 and stx2 [21][22][23][24][25].
The hlyA gene was present in 78.0% (490/628) of goat STEC isolates, consistent with previous reports which have shown similar rates in goat STEC elsewhere [39,42,44]. However, lower rates of hlyA ranging from 35-60.9% have also been reported [40,41,45]. The hlyA gene encodes a pore-forming hemolysin which lyses human erythrocytes with subsequent release of iron from heme, a chemical needed for STEC growth and survival in the intestine. Previously, the presence and expression of hlyA has been associated with severe STEC disease in humans including HC and HUS [35]. However, STEC that were hlyA-negative have also been incriminated in severe disease including bloody diarrhea, HC and HUS, thereby suggesting that the pathogenic role of hlyA in STEC remains uncertain [23].
Most of the goat STEC were eaeA-negative except for the top seven STECs (22.1%) including O157:H7, O26:H2, O111:H8 and O103:H8 and a few (0.7%) non-O157/non-top seven isolates: O71:H14, O108:H25 and O163:H8, in agreement with previous studies which have shown that eaeA is not common among goat STEC [40,42,45]. The presence of eaeA in goat top seven STEC isolates is of clinical significance as eaeA is considered an important STEC adhesin and marker of high virulence and potential to cause severe disease (HC and HUS) in humans [72], especially when accompanied with stx2 [23]. However, in some cases, eaeA-negative serotypes (O91:H21 and O113:H21) STEC have also been associated with severe disease thereby suggesting that other virulence or unknown host factors may influence disease severity [72][73][74]. The absence of eaeA may indicate that goat STEC are less virulent and may also explain why goat STEC are rarely incriminated in human disease worldwide. Of particular interest were eaeA-positive goat isolates which belonged to serotypes O103:H8, O71:H14, O108:H25 and O163:H8 but have never been associated with human disease or outbreaks. These isolates will be worth monitoring closely as possession of eaeA may be indicative of higher virulence potential and likelihood to cause severe disease in humans.

Conclusions
Historically, studies on the presence of STEC in goats are very few compared to cattle which are considered the main STEC reservoir. This study is the first report on the presence of STEC in goats in South Africa. The findings of this study show that goats carry a diverse range of STEC serotypes, some of which have been previously incriminated in mild to severe enteric disease in humans. Collectively, these findings suggest that goats grazing on communal rangeland in South Africa are a reservoir and potential source of STEC for humans in South Africa. Further molecular characterization of goat STEC isolates will be needed in the future to fully assess the virulence potential of goat STEC and capacity to cause disease in humans. In addition, studies that compare STEC isolates from goats and humans will be necessary to fully understand the role played by goats as a source of STEC human disease in South Africa. Data from this study will be useful for understanding the epidemiology of STEC in animals and formulating policies aimed at preventing and controlling zoonotic or foodborne diseases along the food chain.

Study Population and Sample Collection
Goat fecal samples (N = 289) were obtained from four goat herds. The goat herds were located on different communal rangelands in Gauteng province, South Africa. The herds were designated using alphabetical letters: herd A (n = 154), herd B (n = 43), herd C (n = 52) and herd D (n = 40). Each herd was visited once. Refer to Figure 4 for a map of the Gauteng province, South Africa showing the locations of the different herds (A, B, C and D) from which goat fecal samples were obtained. Fresh fecal samples were collected by rectal palpation, using a new nitrile examination glove per animal. Samples were placed in sterile specimen containers and transported in a cooler box on ice to the laboratory where they were stored at 4 • C until further processing. Ethical clearance for conducting this research was obtained from the Research Ethics and Animal Ethics Committees of the Faculty of Veterinary Science, University of Pretoria, under approval number REC110-21.
lates will be needed in the future to fully assess the virulence potential of goat STEC and capacity to cause disease in humans. In addition, studies that compare STEC isolates from goats and humans will be necessary to fully understand the role played by goats as a source of STEC human disease in South Africa. Data from this study will be useful for understanding the epidemiology of STEC in animals and formulating policies aimed at preventing and controlling zoonotic or foodborne diseases along the food chain.

Study Population and Sample Collection
Goat fecal samples (N = 289) were obtained from four goat herds. The goat herds were located on different communal rangelands in Gauteng province, South Africa. The herds were designated using alphabetical letters: herd A (n = 154), herd B (n = 43), herd C (n = 52) and herd D (n = 40). Each herd was visited once. Refer to Figure 4 for a map of the Gauteng province, South Africa showing the locations of the different herds (A, B, C and D) from which goat fecal samples were obtained. Fresh fecal samples were collected by rectal palpation, using a new nitrile examination glove per animal. Samples were placed in sterile specimen containers and transported in a cooler box on ice to the laboratory where they were stored at 4 °C until further processing. Ethical clearance for conducting this research was obtained from the Research Ethics and Animal Ethics Committees of the Faculty of Veterinary Science, University of Pretoria, under approval number REC110-21.

DNA Extraction and STEC Screening
All Drigalski Lactose and CHROMagar STEC agar Petri dishes showing bacterial growth were screened for STEC by PCR [75]. Briefly, a loopful of bacterial colony sweep was collected from each Drigalski Lactose agar and CHROMagar STEC plate showing growth and suspended in 1 mL of FA Buffer (223143, Becton Dickinson and Company, Sparks, MD, USA) [38]. The suspension was homogenised and washed by vortexing, then centrifuged for 5 min. After centrifugation, the supernatant was discarded, and the pellet was re-suspended in FA buffer. After the second wash and centrifugation rounds, the pellet was re-suspended in 500 µL of sterile water, mixed and boiled at 100 • C for 25 min. The boiled preparation was thawed on ice and stored at −20 • C for further processing [38]. A multiplex PCR (mPCR) protocol was used to screen the DNA template for stx1, stx2, eaeA and hlyA using previously described cycling parameters and primers [75]. Briefly, each 25 µL PCR reaction mixture contained 2.5 µL of 10X Thermopol reaction buffer, 2.0 µL of 2.5 mM dNTPs (deoxynucleotide triphosphates), 0.25 µL of 100 mM MgCl 2 , 0.6 µL of each primer (10 µM final concentration), 1 U of Taq DNA Polymerase and 5 µL of DNA template. The DNA from Escherichia coli O157:H7 strain EDL933 (ATCC 43895) and sterile water were used as positive and negative PCR controls, respectively. All PCR reagents were purchased from New England BioLabs (NEB, Ipswich, MA, USA) except for the primers which were supplied by Inqaba Biotec (Pretoria, South Africa).

STEC Isolation and Identification
For STEC isolation and identification, colony sweeps were collected from Drigalski Lactose agar and CHROMagar plates which were positive for stx1 and/or stx2 on PCR and streaked onto Drigalski Lactose agar and CHROMagar STEC to obtain single colonies. Five single colonies were purified from each plate and multiplied individually on Luria Bertani agar (REF244520, Becton and Dickinson & Company, Sparks, MD, USA). Once again, DNA was extracted from purified colonies by the boiling method [38]. DNA from each purified colony was screened for stx1, stx2, eaeA and hlyA by PCR [75] to verify and confirm the STEC status of each pure colony. Colonies which were positive for stx1 and/or stx2 were preserved at −80 • C in a bacterial freezing mixture [38] for further O:H serotyping.

STEC Serotyping
All confirmed STEC pure single colonies were serotyped (O:H) by PCR using previously described primers and cycling conditions [57][58][59]. STEC strains which were previously serotyped by traditional serotyping at the National Microbiology Laboratory, Public Health Agency of Canada, Guelph, Ontario, Canada, and the Laboratorio de Referencia de Escherichia coli (LREC), Facultad de Veterinaria, Universidad de Santiago de Compostela, Lugo, Spain and a number of E. coli O:H types in our collection (unpublished) were also used as positive controls in PCR serotyping assays. Furthermore, the following STEC isolates which were provided by the European Union Reference Laboratory for Escherichia coli, Istituto Superiore di Sanità, Rome Italy, were used as positive controls for serotyping the major seven STEC serogroups: STEC-C210-03 (O157), STEC-ED476 (STEC O111), STEC-C1178-04 (STEC O145), STEC-C125-06 (STEC O103) and STEC-ED745 (O26).

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/toxins14050353/s1, Table S1: Association between O group and H-type(s) among goat STEC Isolates; Table S2: Goat STEC major virulence factors and gene combinations.