Diarrhea is one the most common causes of mortality of neonatal foals, which results in economic losses worldwide [1, 2]. Until 6 months of life, up to 20% of foals have been reported to suffer from diarrhea caused by infectious agents. The etiology of diarrhea in neonatal foals is complex and involves infectious agents, management, and facilities, as well as nutritional, environmental, and physiological conditions [3–5].
Infectious causes of foal diarrhea predominantly include bacteria, viruses, and parasitic agents [5–7]. Some studies of diarrhea in neonatal foals have focused on only one pathogen [2, 8–11] although it is usually caused by diverse combinations of enteric infections [1, 5, 7, 12, 13]. The understanding of the prevalence and complexity of major pathogens associated with enteric infections in neonatal foals is limited. Comprehensive studies have highlighted the impacts of enteric coinfections in foals with diarrhea and have identified major virulence factors of some pathogens in these infections [1, 5, 7, 13]. However, few studies have evaluated the normal microbial population of the equine intestine, particularly in foals [15–17]. Nondiarrheic foals have been suggested to be potential reservoirs of enteric pathogens, including agents with zoonotic potential [13, 18]. The aim of this study was to identify enteric microorganisms involved in monoinfections and coinfections and the associated virulence factors in healthy and diarrheic foals up to three months of age and to identify dehydration, alterations in the complete blood count (CBC), and electrolyte abnormalities.
2.1. Study Design
Fecal samples were collected from 56 diarrheic and 60 nondiarrheic foals of different breeds up to 90 days of age and divided into three groups according to age for bacteriological, virological, and parasitological tests. Monoinfections and coinfections with different enteric pathogens were investigated, as well as selected virulence factors and toxin production in Escherichia coli, Clostridium sp., and Rhodococcus equi isolates. Clinical abnormalities and hematological and blood gas parameters were also assessed in the diarrheic foals. This study was carried out from July 2011 to October 2012 at horse farms in the central region of the State of São Paulo, Brazil.
2.2. Selection of Farms and Animals
A cross-sectional study was carried out at 15 horse farms (50–150 horses per farm) (Table 1). All foals with diarrhea identified at each farm were included in the study. Samples were collected from each foal only once. Brief clinical histories were also obtained from the diarrheic and nondiarrheic animals. The exclusion criteria were as follows: (1) foals undergoing antimicrobial therapy and (2) foals older than 90 days of age. An average of 7.5 foals per farm (1 to 28) was used. All of the farms were similar in terms of the general management of the horses, animal hygiene, and facilities, and they all had histories of diarrhea in foals. None of the equines had been vaccinated against rotavirus or other enteric pathogens.
All of the procedures followed in this study were approved by the Animal Use and Ethics Committee (CEUA), Brazil (CEUA–protocol 155/2011).
2.3. Clinical Examination
Prior to obtaining fecal samples, all foals underwent clinical examination [19–21]. The foals with diarrhea had abnormal fecal consistency and color, and clinical signs (anorexia, depression, and/or weakness) lasting for more than 24 hours. All foals with and without diarrhea were nursed and were thus considered to have ingested colostrum within the first 24 hours of life (nursing was observed). Dehydration rates were classified at clinical examination as discrete (6%), moderate (8 or 10%), and severe (≥12%). Dehydration was estimated based on several physical parameters (skin elasticity, mucous membrane color, sunken eyes/enophthalmia, capillary refill time, and heart rate) [4, 21].
The 116 foals were divided into diarrheic and nondiarrheic groups. They were then subdivided into the following three groups according to the age: GI (0–30 days of age); GII (31–60 days); and GIII (61–90 days).
2.5. Fecal Samples
Fresh fecal material was collected directly from the rectum of each animal and was then stored at 4–8°C and transported to the laboratory. In addition, aliquots of fecal samples were kept frozen (−80°C) in the laboratory until analyses with different methods.
2.6. Identification of Escherichia coli, Rhodococcus equi, Clostridium perfringens, Clostridium difficile, and Salmonella spp.
For the culturing of Escherichia coli, fecal material was plated on defibrinated sheep blood agar and MacConkey agar and evaluated at 24, 48, and 72 hours. In addition, microbiological culturing of the feces was performed using CAZ-NB selective media for Rhodococcus equi isolation.
For the isolation of C. perfringens and C. difficile, fecal material was diluted. Aliquots were plated in sulfite polymyxin sulfadiazine agar and in taurocholate cycloserine cefoxitin fructose agar and incubated anaerobically. Suspected Clostridium colonies were classified using conventional phenotypic methods and were then analyzed by PCR for detection of the major C. perfringens toxin genes (alpha, epsilon, beta, and iota) and beta-2, NetB-, and enterotoxin-encoding genes [25, 26]. For C. difficile, a multiplex PCR for a housekeeping gene (tpi), toxins A (tcdA) and B (tcdB), and a binary toxin gene (cdtB) was performed on all suspected colonies.
For Salmonella detection, samples were inoculated into specifics Rappaport-Vassiliadis and Tetrathionate broth. Each broth culture was plated in xylose-lysine-deoxycholate agar and in bismuth sulfite agar and was then incubated. Suspected Salmonella sp. colonies were inoculated in triple sugar iron and lysine iron agar and classified using conventional biochemical tests. A commercial agglutination test using polyvalent anti-Salmonella serum (Probac™) and serotype identification were also carried out.
2.7. Molecular Detection of E. coli and Virulence Factors
E. coli strains were cultured overnight in BHI (Brain Heart Infusion) and were then inoculated in trypticase soy agar for confluent growth overnight, followed by DNA extraction. The sequences of the primers used, predicted sizes of the amplified products, and specific annealing temperatures have been previously described. The following groups of genes were analyzed by PCR: papC and papG alleles (P fimbria), sfaC/D (S fimbria), afaB/C, saa, iucD, cnf-1, cnf-2, hly, vt1, vt2, sta, stb, ipaH, and eae, eaf, and bfp.
2.8. Virulence of R. equi
Plasmid DNA was isolated using the alkaline lysis method with several modifications. The target DNA for PCR amplification was based on reported vapA (15- to 17-kDa antigen) and vapB (20-kDa antigen) gene sequences (GenBank accession numbers D212361 and D44469). Plasmid DNA was digested with the restriction endonucleases EcoRI, EcoT22I, and HindIII [30, 31]. Then, the plasmid samples were separated by electrophoresis and examined under UV light. PCR amplification was performed as previously described.
2.9. Cytotoxicity Assay (CTA) of C. difficile
C. difficile A/B toxins were investigated using Vero cells (Vero ATCC CCL 81). Fecal samples were diluted in phosphate-buffered saline (PBS) and centrifuged. The supernatants were filtered and diluted twofold to 1 : 1,024, and then serial dilutions and parallel samples containing Clostridium sordellii antitoxin were added to Vero cell monolayers. The cells were examined after incubation. A sample was considered positive in CTA if at least 90% of cells were round and if the effect was neutralized by the antitoxin at the same dilution in the parallel sample.
2.10. Rotavirus and Coronavirus Detection
RNA was extracted from fecal samples using TRIzol™ reagent (Invitrogen™), following the manufacturer's instructions, and the phenol/chloroform method. Samples were tested for the presence of coronavirus using a Betacoronavirus-1-specific RT-PCR assay targeting the RNA-dependent RNA-polymerase gene (RdRp). The BCoV Kakegawa strain and PBS were used as positive and negative controls, respectively. The samples were also analyzed for the presence of rotavirus by polyacrylamide gel electrophoresis (PAGE). The NCDV group A rotavirus strain was used as a positive control.
2.11. Lawsonia intracellularis, Cryptosporidium parvum, and Giardia duodenalis DNA Detection
DNA was extracted from fecal samples using a Qiagen DNA Stool Mini kit (Qiagen™). Nested PCR amplification was performed using the set of primers previously described for detection of Cryptosporidium spp. [36, 37], G. duodenalis [38, 39], and L. intracellularis DNA [8, 40], with some modifications. Nested PCR products from the genera Cryptosporidium and Giardia with the expected sizes were analyzed by automated direct sequencing using a 3500 Genetic Analyzer (Applied Biosystems, Foster City, CA) and BigDye™ Terminator v3.1 Cycle Sequencing Kit.
2.12. Parasitological Examination
To identify intestinal parasites (helminthes), fresh fecal samples were assessed by flotation in zinc sulfate, followed by microscopic examination in a McMaster counting chamber.
2.13. Analysis of Blood from Diarrheic Foals
Blood counts were performed using a hematological counter (Poch 100iV Diff, Roche®). Hematocrit (Htc) was measured using the microhematocrit method. Leukocyte counts were performed using 100 cells, along with evaluations of erythrocyte, leukocyte, and platelet morphologies, in blood smears stained using quick Panotic dye (LaborClin®). Analysis of venous blood [blood pH, Htc, hemoglobin (Hb), partial pressure of CO2 (PCO2), partial pressure of O2 (PO2), total CO2 concentration (tCO2), oxygen saturation (SO2), base excess, and HCO3, Na, K, and ionized calcium (iCa) concentrations] was performed with I-STAT EG7+ Cartridges (Abbott™, East Windsor, NJ, USA).
2.14. Statistical Analysis
The chi-square or Fisher's exact test was used to compare the frequencies of the different enteric pathogens between the diarrheic and nondiarrheic foals and to compare the presence of different virulence factors and production of toxins by isolates from these two groups of foals. Statistical analyses were conducted using SAS/STAT, and the statistical significance level was set at 0.05.
3.1. Clinical Signs of Dehydration and Diarrhea
The diarrheic and nondiarrheic sampled foals were distributed into the following three age groups: GI (0–30 days), with 19 diarrheic and 10 nondiarrheic foals; GII (31–60 days), with 17 diarrheic and 13 nondiarrheic foals; and GIII (61–90 days), with 20 diarrheic and 37 nondiarrheic foals. Signs of dehydration were observed exclusively in the diarrheic foals, among which 48.2% (27/56), 16% (9/56), and 3.6% (2/56) showed discrete, moderate, and severe dehydration, respectively.
Gastrointestinal tract auscultation resulted in the identification of increased intestinal sounds in the diarrheic foals, as well as a high frequency of defecation in 95% (n = 53/56) of the animals.
3.2. Hematological Findings and Blood Gas Parameters
Hematological evaluation of the diarrheic foals revealed that 46.9% (15/32) had an increased plasmatic fibrinogen level. In addition, leukocytosis was observed in 15.5% (7/45) of the diarrheic foals, and neutrophilia was detected in 22% (10/45). Lymphocytosis was observed in 37% (21/56) of the animals.
Blood gas examination revealed that 16% (9/56) of the foals had low pH (range 7.3 ± 0.1) and blood bicarbonate levels, characteristic of metabolic acidosis. Electrolyte imbalances were evident in the diarrheic foals, with 5.3% (3/56) exhibiting hypocalcemia, 66% (37/56) displaying hyponatremia (133.4 mmol/L ± 3.3), and 39% (22/56) exhibiting hypokalemia (3.1 mmol/L ± 0.7) (Table 2).
No associations were observed between the hematological or blood gas parameters and the enteric agents assessed or the mortality rate among the foals. In contrast, a tendency toward abnormal values of the hematological and blood gas parameters was observed in the foals with severe dehydration and signs of diarrhea.
3.3. Enteric Pathogens
The frequencies of enteric pathogens of bacterial, parasitic, and viral origins, as well as the selected virulence factors detected in the diarrheic and nondiarrheic foals sampled, are summarized in Table 3.
The most frequent enteric pathogens identified in the diarrheic foals were as follows: 30.3% (17/56) E. coli fimH, 25% (14/56) Salmonella spp., 25% (14/56) Strongyloides, 21.4% (12/56) Clostridium perfringens type A, 19.6% (11/56) E. coli ag43, 10% (6/56) Strongylus, and 5% (3/56) R. equi vapA. In contrast, the most frequent enteric pathogens found in the nondiarrheic foals were as follows: 35% (21/60) E. coli ag43, 25% (15/60) E. coli fimH, 25% (15/60) Strongyloides westeri, 18% (11/60) Strongylus, 10% (6/60) C. perfringens type A, and 2% (1/60) R. equi vapA.
C. perfringens type A was isolated from 21% and 10% of the diarrheic and nondiarrheic foals, respectively. Four of these strains were positive for the beta-2 toxin-encoding gene (cpb2), including three strains from apparently healthy animals and one from a diarrheic foal. None of the isolates were positive for NetB- or enterotoxin-encoding genes. The isolation of C. perfringens was significantly associated with the presence of diarrhea (p = 0.033). Among the 12 diarrheic foals that tested positive for C. perfringens type A, three (25%) also tested positive for another enteropathogen. C. perfringens type C was not detected in this study.
C. difficile infection was confirmed in only one diarrheic foal in this study. This animal, which was 19 days old, tested positive for the A/B toxins, and a toxigenic C. difficile strain (A+B+CDT−) was isolated. Three other strains of C. difficile were isolated in this study, including two nontoxigenic strains (A−B−CDT−) that were both isolated from nondiarrheic animals (GI), and a toxigenic strain (A+B+CDT−) that was isolated from a one-day-old foal that tested negative for the A/B toxins.
Seven different serotypes of the Salmonella genus were identified. Among them, fourteen were isolated from the diarrheic foals, with a predominance of Salmonella Infantis, Salmonella Typhimurium, and Salmonella Saintpaul. Among the nondiarrheic animals, four serotypes were identified, namely, Salmonella Infantis, Salmonella Saintpaul, Salmonella Typhimurium, and Salmonella Newport. Salmonella serotypes were identified in single and combined infections in 3 and 11 diarrheic foals, respectively. In contrast, only four nondiarrheic foals tested positive for Salmonella, which was present together with other enteric pathogens in all of the animals.
Salmonella Saintpaul (67% versus 33%), S. Panama (100% versus 0%), S. Infantis (83% versus 17%), S. Typhimurium (75% versus 25%), S. enterica subsp. enterica (100% versus 0%), S. Muenchen (100% versus 0%), C. perfringens toxin A (67% versus 33%), and R. equi vapA (75% versus 25%) exhibited higher prevalence rates in the diarrheic foals than in the nondiarrheic foals.
Seventeen and 15 strains of Escherichia coli were isolated from the feces of the diarrheic and nondiarrheic foals, respectively. The virulence factors fimH and ag43 were detected in 32 isolates, of which 11 were obtained from diarrheic animals. The papC gene was detected in 4 diarrheic and 2 nondiarrheic foals. Coinfections of E. coli with specific virulent factors and Salmonella serotypes in the diarrheic foals were observed as follows: Salmonella Infantis, E. coli fimH and papC; Salmonella Newport, E. coli fimH and ag43 (Table 6); Salmonella Infantis, E. coli fimH and ag43; and Salmonella Infantis, E. coli fimH. Among the nondiarrheic foals, the following coinfections were observed: Salmonella typhimurium, E. coli ag43 and fimH; Salmonella saintpaul and E. coli ag43. Only one foal that tested positive for E. coli showed clinical signs of sepsis.
Nine samples from all foals tested positive for R. equi; three strains were identified in the diarrheic foals, and six were detected in the nondiarrheic foals. Virulence plasmid analysis of the nine strains revealed that one VapA 85-kb type I isolate and one VapA 87-kb type I isolate were present in two diarrheic animals in group I. One diarrheic animal in group II was positive for VapA 85-kb type I, whereas a nondiarrheic animal in the same group was positive for VapA 87-kb type I. The remaining five R. equi isolates identified in the nondiarrheic animals were considered avirulent (absence of the vapA and vapB genes).
Nonconventional primary agents of diarrhea in foals, such as Hafnia alvei, Proteus mirabilis, Proteus vulgaris, Enterobacter agglomerans, Providencia rettgeri, and Citrobacter spp., were identified in single isolates (without coinfection with other major enteric pathogens), despite their low frequencies (Table 4). Serratia rubidaea, Edwardsiella agglomerans, Edwardsiella tarda, Citrobacter spp., and Hafnia alvei were recovered in pure cultures at low frequencies in the nondiarrheic animals (without coinfection with enteric pathogens) (Table 4).
None of the fecal samples tested positive for Lawsonia intracellularis. Five (4%) animals up to 90 days of age tested positive for Cryptosporidium parvum (3 foals with diarrhea and 2 nondiarrheic foals). One diarrheic foal and one nondiarrheic foal presented C. parvum as the only detected pathogen.
Sequence analysis of the genetic material resulted in detection of Giardia duodenalis in only two (2%) foals in group II, including one diarrheic and one nondiarrheic animal.
Helminths were identified at relatively similar frequencies between the diarrheic and nondiarrheic foals up to 90 days of age, with the highest frequency observed in group III. No helminth was detected as a single pathogen in the diarrheic foals; rather, these enteric parasites were identified in combined infections among the diarrheic samples.
Only two foals tested positive for coronavirus (one diarrheic foal in group I and one nondiarrheic foal in group III). Rotavirus was not detected in the studied samples.
3.4. Co-Infections and Mortality of Foals
The distribution of enteric pathogens identified alone or in combination in the diarrheic and nondiarrheic foals up to 90 days of age is shown in Table 4. A total of 87% (49/56; p < 0.27) of the fecal samples from the diarrheic foals tested positive for enteric pathogens, in addition to 80% (48/60; p < 0.27) of those from the nondiarrheic animals. Among the diarrheic foals, 46% had coinfections, and 41% had monoinfections. In contrast, 47% of the nondiarrheic foals had monoinfections, and 33% had coinfections (p < 0.29) (Table 4). The most common coinfections involved Strongyloides westeri, Strongylus, and Salmonella spp.; combined infections with VapA-positive R. equi and C. perfringens toxin A were also common. However, no significant difference was observed between the most common enteric pathogens detected in the mono- versus coinfections (Table 5).
Eight foals died during the study (six within diarrhea group). Salmonella was detected in 7 out of 8 of the dead animals. The following coinfections were detected in the dead animals: (1) E. coli, Clostridium perfringens type A, Salmonella Panama, Strongyloides westeri, and Strongylus; (2) E. coli fimH, E. coli papC, Salmonella Infantis, and Strongyloides; (3) Salmonella Typhimurium and Strongyloides westeri; (4) Salmonella enterica subsp. enterica and Strongyloides westeri; and (5) Salmonella Muenchen and Strongyloides westeri (Table 6). Only one dead foal tested negative for all enteric pathogens. With regard to the two nondiarrheic foals that died, one was positive for E. coli and Salmonella Infantis, while the other was positive for E. coli fimH, E. coli ag43, Salmonella Newport, and Strongyloides westeri (Table 6). The relationships between the mortality rate and age in the diarrheic and nondiarrheic foals are shown in Table 6.
4.1. Coinfections and Mortality Rates
Considerable variations in the frequencies of the major enteric pathogens recovered from diarrheic and nondiarrheic foals worldwide have been reported in different studies [1, 6, 7, 13]. Moreover, many studies have focused on only one enteric pathogen [8, 9, 9, 10, 12, 40, 63, 64], despite the recognized etiological complexity and the potential combinations of pathogens in equine enteric infections [13, 52]. Notably, a large number of enteric pathogens of bacterial, viral, and parasitic origins, as well as diverse coinfections, have been identified in diarrheic and nondiarrheic foals. In fact, substantial etiological complexity of enteric organisms has been observed among foals, reinforcing the necessity to include major enteropathogens in the differential diagnosis of neonatal enteric infections in these animals [1, 7, 17]. The failure to identify significant differences in the frequencies of pathogens between diarrheic and nondiarrheic foals in the current study, with the exception of Clostridium perfringens toxin A+, has also been reported previously [6, 66], indicating that other factors, such as age, the nutrition and immune statuses, environmental conditions, general management practices, and characteristics of the farm facilities, may influence the occurrence of clinical cases of diarrhea.
In addition, no significant associations of specific coinfections with the presence (or not) of diarrhea were detected, and a high mortality rate was observed among the Salmonella-positive diarrheic foals, regardless of the serotype or other pathogens present. High mortality of foals with enteric salmonellosis has also been described elsewhere [1, 66]. These results provide important information for practitioners who detect Salmonella spp. in the feces of foals because proper antimicrobial therapy, rigorous hydration, and other critical care measures are required to prevent serious complications and to improve prognosis.
5. Study Limitations
No clinical examinations or blood counts were carried out for the nondiarrheic foals; thus, the reference values for this species according the age were used.
Blood culturing and measurement of the serum IgG concentration were not performed, and necropsy and microbiological culturing of the enteric samples from the dead foals were not carried out due to the large number of included farms and because not all farms had veterinary residents.
Although the results related to virulence factors of E. coli were not statistically significant, their identification is important by increase of the understanding of the involvement of this pathogen in development of sepsis in foals.
Although we attempted to collect samples from diarrheic and nondiarrheic animals at all farms, it was not possible to collect samples from control animals (without diarrhea) at farms 4, 5, and 13.
The low prevalence related to the studied protozoa (C. parvum and G. duodenalis) may be related to intermittency of its release in the environment, as only one sampling time was used in this study for each foal.
A wide variety of enteric pathogens in foals were identified in this study. The frequency of C. perfringens toxin A+ significantly differed between the diarrheic and nondiarrheic animals, confirming that this toxin is a major virulence factor for enteric infections in foals. Salmonella spp. was the most prevalent pathogen associated with mortality among the diarrheic and nondiarrheic foals sampled, indicating the necessity of proper therapy and critical care of foals showing fecal isolation of this pathogen.
The detection of different serotypes of Salmonella, toxigenic Clostridium perfringens and Clostridium difficile strains, virulent Rhodococcus equi (VapA-positive), and Cryptosporidium parvum in feces from the diarrheic and nondiarrheic foals highlights a public health concern and indicates that practitioners should be careful when handling fecal material from foals because these organisms are recognized as human pathogens through fecal contamination of food and water.