- Open Access
Bacteria detected in the genital tract, semen or pre-ejaculatory fluid of Swedish stallions from 2007 to 2017
Acta Veterinaria Scandinavicavolume 61, Article number: 25 (2019)
Although artificial insemination (AI) was developed as a means of controlling disease transmission, pathogens can still be transmitted to females in semen used for AI. In addition, bacteria can cause deterioration in sperm quality during storage. Semen becomes contaminated by the male’s normal bacterial flora as it passes out of the reproductive tract but potential pathogens may also contaminate the semen. Therefore, semen samples from stallions to be used for AI are tested before the breeding season to minimize transmission of pathogens to inseminated mares. In Sweden, semen samples are tested at the National Veterinary Institute, Uppsala (SVA). For the present study, a retrospective analysis was made of potentially pathogenic bacteria isolated from samples submitted to the SVA from 2007 to 2017.
In our study, Taylorella equigenitalis was found infrequently (53 out of 25,512 samples), representing 11 out of 2308 stallions. If T. equigenitalis was detected, the stallions were treated with antibiotics and re-tested later in the same year. Klebsiella pneumoniae and beta haemolytic streptococci were the most commonly found potential pathogens, whereas Pseudomonas aeruginosa was also isolated occasionally. There were considerable differences in the number of species isolated each year.
Potential pathogens were identified in relatively few of the samples submitted to SVA during this period, with T. equigenitalis not being identified since 2015. Of the other potential pathogens, K. pneumoniae and beta haemolytic streptococci were the most common. The information is relevant for determining guidelines on the testing and treatment of stallions before breeding.
Although artificial insemination (AI) was first developed as a means of controlling disease transmission, pathogens can still be transmitted in semen used for AI. These pathogens may infect inseminated females or contribute to a rapid deterioration in sperm quality.
Semen of healthy stallions is considered to be sterile when produced, but it becomes contaminated by microorganisms as it passes through the genital tract at ejaculation. Furthermore, semen collected for AI may become contaminated from the air, from equipment used in semen collection and during handling , from the semen collector and by microorganisms naturally present on the mucosal membranes or skin of the penis and prepuce .
In addition to the normal flora, pathogens such as Pseudomonas aeruginosa , Klebsiella pneumoniae [4, 5] and Streptococcus equi subsp. zooepidemicus  are occasionally found. Taylorella equigenitalis may also be present . Due to the importance of T. equigenitalis infection in breeding horses, it is compulsory for stallions at AI centers in Sweden to be tested before the start of the breeding season. The presence of T. equigenitalis is notifiable.
Antibiotics are commonly added to semen extenders to reduce the growth of bacteria that may have contaminated the ejaculate during semen collection. However, this practice represents a non-therapeutic use of antibiotics, which may contribute to the development of antibiotic resistance. Since the reproductive tract of the mare has a well-developed immune defense to eliminate bacteria deposited during natural mating, small numbers of bacteria in inseminated semen may not represent a source of infection. Therefore, increasing awareness of the type of bacteria likely to occur in semen may enable a more prudent use of antibiotics to be suggested, in line with the current need to reduce antibiotic usage.
The aim of this study was to analyse bacteriological findings of genital tract swabs, semen and pre-ejaculatory fluids submitted to the National Veterinary Institute in Uppsala, Sweden (SVA) over a decade, to determine the prevalence of bacterial genital tract pathogens present in Swedish stallions.
During the period 2007 to 2017, 25,512 samples originating from 2308 stallions of various breeds and ages in Sweden were taken by veterinarians and submitted to SVA for testing specifically for T. equigenitalis. The samples consisted of swabs of the penile shaft or prepuce, urethral fossa, urethral orifice and semen or pre-ejaculatory fluid, although not all sites were sampled from all animals. Veterinarian were instructed to pass a cotton swab over the skin folds of the prepuce, urethra, urethral fossa, semen and/or the pre-ejaculatory fluid. The swabs were submitted to the laboratory in Amies charcoal transport medium within 48 h after sampling. Of the 2308 stallions, 2099 were sampled on more than once occasion, since animals from which T. equigenitalis was detected in any of the samples were treated and re-tested; in some instances, samples were taken on multiple occasions in the same year.
In addition, results were available from 730 semen samples or pre-ejaculatory fluid from 319 stallions in Sweden submitted to SVA for general bacteriological analysis during the same period. Data were compiled from routine analysis for detection of presumptive pathogenic bacteria such as P. aeruginosa, K. pneumoniae and beta haemolytic streptococci. These semen samples were separate from those used for testing for T. equigenitalis.
Bacteriological culture and identification
Examination for Taylorella equigenitalis
Upon arrival at the laboratory swabs were plated on three different haematin “chocolate” blood agar, one with no inhibitors, one containing streptomycin sulphate (200 µg/mL) and the third one with 1 µg/mL trimethoprim, 5 µg/mL clindamycin and 5 µg/mL amphotericin B. All culture media were prepared by SVA. The plates were incubated in CO2 at 37 °C for 7 days. Any 0.8–1.0 mm wide white and peaked colonies on blood agar containing Gram-negative rods, and producing catalase and oxidase, were suspected of being T. equigenitalis. Further identification were performed by API zym kits (BioMerieux, USA) , Mono-Tayl-agglutinations test (BioNor, Norway) and ALA-test. From 2012, T. equigenitalis was identified by matrix assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS).
General bacteriological analysis
Samples of semen and pre-ejaculatory fluid were cultured on horse blood agar, bromocresole lactose purpur agar and Pseudomonas isolation agar (SVA, produced in-house) and incubated at 37 °C for 48 h. The agar plates were examined after both 24 and 48 h’ incubation. Presumptive pathogenic bacterial colonies were re-cultured on horse blood agar to obtain monocultures for further identification. Only the presence of potentially pathogenic bacteria, such as Klebsiella spp. and Pseudomonas spp., were recorded in the data set; however, the presence of Enterobacter spp. was also reported because of its resemblance to Klebsiella spp. In the case that potential pathogens were detected, the organism was identified, together with a susceptibility test for the isolated pathogen. If no pathogenic bacteria were identified, the type of growth was recorded as either (i) no bacterial growth, (ii) growth of unspecific mixed flora, or (iii) sporadic, sparse, moderate or rich growth of unspecified mixed flora.
The methods used for bacterial identification at SVA have changed over the years. Before 2012, bacteriological identification was based on traditional methods, relying on morphological/macroscopic features on different agars, phenotypic identification using Gram-staining, culturing and simple biochemical tests for instance, catalase, oxidase, indole, urease, coagulase and API 20E (BioMérieux, Marcy-l’Étoile, France). After 2012 all identification to species level was performed using MALDI-TOF MS (Bruker Daltonics, Bremen, Germany) .
Examination for Taylorella equigenitalis
The samples originated from 71 studs around the country, mostly from the middle and south of Sweden. The distribution of material provided for testing is shown in Table 1.
Of the 2308 stallions tested, T. equigenitalis was detected from 53 swabs taken from 11 stallions (Table 2); 14 additional swabs taken at the same time from the same stallions were negative (Table 3). Infected animals were re-tested after treatment; therefore, some animals were tested repeatedly during any 1 year. At the post-treatment sampling in the same year, T. equigenitalis was isolated again from four stallions but was not isolated on subsequent samplings later in the year or in subsequent years. Two additional stallions (stallions 5 and 6, Table 3) appeared to have responded to treatment, since T. equigenitalis was not isolated from a set of swabs taken after treatment, but the organism was found again at subsequent testing later in the same year. For one of these two stallions (stallion 6), T. equigenitalis was not found subsequently after the 3rd and 4th testing, despite repeated samplings (10 sets of swabs) over 2 years. However, for stallion 5, T. equigenitalis was isolated from the first set of swabs, was not isolated from the 2nd, 3rd and 4th sets of swabs but was isolated from the 5th and 6th set of swabs. It is not known what became of this stallion afterwards as no further swabs were submitted to SVA for testing. For two of the stallions (9 and 11 in Table 3) T. equigenitalis was isolated in 2011 and 2010 respectively, despite not having been isolated in previous years.
Taylorella equigenitalis was identified from 11 out of 19 swabs from the penile shaft and prepuce, 18 out of 19 swabs from the urethral fossa, 15 out of 19 swabs from the urethral orifice, and 9 out of 10 samples from semen/pre-ejaculatory fluid. For six of the stallions, T. equigenitalis was found on two or more sampling occasions (Table 3). Nine of the positive animals that became negative after treatment were tested in subsequent years (between 2 and 10 years, depending on the stallion). The results remained negative for these animals.
General bacteriological analysis
The 730 semen samples submitted for general bacteriological analyses originated from 319 stallions. The number of samples analysed per year (Table 4) was highest in 2008 (n = 145) and lowest in 2017 (n = 9). Three potentially pathogenic species were detected (Table 5), with K. pneumoniae, beta haemolytic streptococci and Pseudomonas spp. being isolated from 8, 8 and 3 animals, respectively.
In most of the cultures, growth of mixed flora was observed and the bacteria were considered to be non-pathogenic. Therefore in these cases, a report of “No specific infection” was made, together with an estimate of the amount of growth of mixed flora. The classification and distribution of bacteria in the samples where no specific infection was identified are shown in Table 6.
In this study, a retrospective data analysis was made of potentially pathogenic bacteria isolated from samples of the stallion reproductive tract submitted to SVA during the period 2007 to 2017. Most of the samples originated from stallions in the middle and southern parts of Sweden, reflecting the distribution of the human population.
There was a lower prevalence of T. equigenitalis in stallions in Sweden than in some other countries. In a study from the USA, T. equigenitalis was detected in 22 out of 222 stallions in the period 2008 to 2010 . In contrast, this organism was detected in four out of 120 imported stallions entering Darlington, Maryland, USA, between 1999 and 2001 . Furthermore, it was found in 17 out of 245 stallions in Slovenia, (6.9%)  while the rate of detection in the Republic of Korea has decreased from 13.5% in 2015 to 1.0% in 2017 .
Pseudomonas aeruginosa, K. pneumoniae and beta haemolytic streptococci were found in 5% of samples submitted for general bacteriological examination. K. pneumoniae were isolated from 2.5% (11/319) of the sampled animals, which is fewer than in a previous Swedish study  where Klebsiella spp. were isolated from 4% (5/115) of the examined horses. In other studies, Klebsiella spp. have been isolated more frequently, although in much smaller sample sizes. In a Spanish study, Ortega-Ferrusola et al.  detected Klebsiella spp. in 3 out of 5 semen samples, whereas Spergser et al.  detected Klebsiella spp. from 13 out of 39 stallions in a study from the northwestern region of Austria.
Beta haemolytic streptococci were isolated infrequently (10/319 animals), in concordance with other studies. Previous reports showed these bacteria in only 7 out 116 of sampled stallions , 1 out of 20 stallions  and 14 of 115 samples , respectively. In contrast, there are also studies that report beta haemolytic streptococci to be the most frequently isolated bacteria, for instance in 4 out of 5 stallions .
In a Swedish study performed 20 years ago , 18 species of bacteria were found, of which five were also detected in our study. The most frequently isolated species in the previous study were Pseudomonas spp. (56/115), Enterobacter spp. (41/115) and Bacillus spp. (31/115) . In comparison, in the present study, P. aeruginosa was detected in 4 of 319 animals. No screening for Bacillus spp. or Enterobacter spp. was done as they are not considered to be potentially pathogenic in horses.
Although these results could indicate a change in the microflora, there could also be other explanations, such as that methods for culture and identification have changed in the intervening period since the previous study in Sweden. However, it was reassuring to note that few potential pathogens were identified in the present study. Methodological differences and differences in sampling sites probably also account for the variations between our results and similar studies in other countries [2, 17, 18].
Identification of T. equigenitalis and treatment of the carrier stallions appears to be effective in preventing transmission of the organism, since the organism was found in only one stallion since 2012. Other potential pathogens were identified in relatively few of the samples submitted to SVA during this period, with K. pneumoniae and beta haemolytic streptococci being the most commonly isolated. There were differences in the prevalence of specific bacteria in various studies in other countries differences in the occurrence of various bacteria, although this may be due, in part, to differences in the methodologies used for identification of bacteria.
Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request, in relation to the Swedish data protection act.
Althouse GC. Sanitary procedures for the production of extended semen. Reprod Domest Anim. 2008;43:374–8.
Malmgren L, Olsson EE, Engvall A, Albihn A. Aerobic bacterial flora of semen and stallion reproductive tract and its relationship to fertility under field conditions. Acta Vet Scand. 1998;39:173–82.
Ferris RA, McCue PM, Borlee GI, Glapa KE, Martin KH, Mangalea MR, et al. Model of chronic equine endometritis involving a Pseudomonas aeruginosa biofilm. Infect Immun. 2017;85:e00332-17. https://doi.org/10.1128/IAI.00332-17.
Ferrer MS, Palomares R. Aerobic uterine isolates and antimicrobial susceptibility in mares with post-partum metritis. Equine Vet J. 2018;50:202–7.
Pickett BW, Voss JL, Jones RL. Control of bacteria in stallions and their semen. JEVS. 1990;19:424–69.
Petersen MR, Nielsen JM, Lehn-Jensen H, Bojesen AM. Streptococcus equi subspecies zooepidemicus resides deep in the chronically infected endometrium of mares. Theriogenology. 2009;1:161–7.
Timoney PJ, Powell DG. Contagious equine metritis—epidemiology and control. JEVS. 1988;8:42–6.
Engvall A. Survival of contagious equine metritis organisms (CEMO) in different transport media as influenced by storage time, temperature and contaminating flora. Zentralbl Veterinarmed B. 1985;1985(32):454–9.
Singhal N, Kumar M, Kanaujia KP, Virdi SJ. MALDI-TOF mass spectrometry: an emerging technology for microbial identification and diagnosis. Front Microbiol. 2015. https://doi.org/10.3389/fmicb.2015.00791.
Erdman MM, Creekmore LH, Fox PE, Pelzel AM, Porter-Spalding BA, et al. Diagnostic and epidemiologic analysis of the 2008–2010 investigation of a multi-year outbreak of contagious equine metritis in the United States. Prev Vet Med. 2011;101:219–28.
Kristula MA, Smith BI. Diagnosis and treatment of four stallions, carriers of the contagious metritis organism—case report. Theriogenology. 2004;61:595–601.
Zdovc I, Ocepek M, Gruntar I, Pate M, Klobucar IB, Krt B. Prevalence of Taylorella equigenitalis infection in stallions in Slovenia: bacteriology compared with PCR examination. Equine Vet J. 2005;37:217–21.
Jeoung HY, Lee SK, Park JY, Kim HJ, Yang SJ, Lee SK, et al. Status of Taylorella equigenitalis infection in thoroughbred horses in the Republic of Korea and the molecular characterization of the Korean Taylorella equigenitalis isolates. JEVS. 2018;69:102–7.
Ortega-Ferrusola C, González-Fernández L, Muriel A, Macías-García B, Rodríguez-Martínez H, Tapia JA, et al. Does the microbial flora in the ejaculate affect the freezeability of stallion sperm? Reprod Domest Anim. 2009;44:518–22.
Spergser J, Aurich C, Aurich JE, Rosengarten R. High prevalence of mycoplasma in the genital tract of asymptomatic stallions in Austria. Vet Microbiol. 2002;87:119–29.
Guimaraes T, Lopes G, Pinto M, Silva E, Miranda C, Correia MJ, et al. Colloid centrifugation of fresh stallion semen before cryopreservation decreased microorganism load of frozen-thawed semen without affecting seminal kinetics. Theriogenology. 2015;83:186–91.
Stephanie S, Aurich C, von Lewinski M, Wulf M, Krüger M, Aurich JE. Development of the genital microflora in stallions used for artificial insemination throughout the breeding season. Anim Reprod Sci. 2013;139:53–61.
Neto RC, Silva Y, Resende HL, Guasti PN, Monteiro GA, Papa PM, et al. Control methods and evaluation of bacterial growth on fresh and cooled stallion semen. JEVS. 2015;35:277–82.
Special thanks to the people working in National Veterinary Institute, Uppsala, Sweden for their help in this study.
Data have not been published previously.
We thank the Ministry of Higher Education and Scientific Research, Iraq for student scholarship (Z.A.-K.). This study was co-funded by the Faculty of Veterinary Medicine and Animal Health, Swedish University of Agricultural Sciences (JMM and MW), and by the National Veterinary Institute, Uppsala, Sweden (EE and EB).
Ethics approval and consent to participate
This study did not require official or institutional ethical approval.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.