Trop Anim Health Prod DOI 10.1007/s11250-014-0614-5
REGULAR ARTICLES
Ureaplasma diversum as a cause of pustular vulvovaginitis in bovine females in Vale Guapore, Mato Grosso State, Brazil João Guilherme L. N. Gaeti & Marconni V. C. Lana & Gustavo S. Silva & Letycia Lerner & Camila G. de Campos & Fernanda Haruni & Edson M. Colodel & Eduardo F. Costa & Luis G. Corbellini & Luciano Nakazato & Caroline A. Pescador
Accepted: 1 May 2014 # Springer Science+Business Media Dordrecht 2014
Abstract Ureaplasma diversum has been associated with various reproductive problems in cattle that include granular vulvovaginitis, weak calves, and abortion. This study was conducted in a beef herd situated in the Middle-West region of Brazil, and the objectives were to verify the presence of U. diversum and to elucidate its possible relationships with independent variables in this bovine herd population. A total of 134 vaginal mucous swabs were taken for polymerase chain reaction (PCR). Of these, 51 (38 %) were PCR positive for U. diversum. Of the 58 heifers with vulvovaginal lesions characterized by hyperemia, granulated lesions, and edema distributed throughout the vulvar mucosa, 37 (64 %) were U. diversum positive; of the 76 heifers without reproductive lesions, 14 (18 %) were U. diversum positive. All tested samples were negative for bovine herpesvirus 1 (BoHV-1). Multivariate logistic regression revealed that the following two variables were significantly associated with the presence of U. diversum: the presence of vulvar lesions (p=0.001) and the presence of a progesterone (P4) device (p=0.001). These findings indicate that U. diversum should be considered a pathogen that is associated with pustular vulvovaginitis in heifers from the Mato Grosso state and that additional studies J. G. L. N. Gaeti : M. V. C. Lana : G. S. Silva : L. Lerner : C. G. de Campos : E. M. Colodel : C. A. Pescador (*) Laboratory of Veterinary Pathology, Federal University of Mato Grosso (UFMT), Av. Fernando Corrêa da Costa 2367, Cuiabá, Mato Grosso CEP 78069-900, Brazil e-mail: [email protected] F. Haruni : L. Nakazato Epidemiology Laboratory, Federal University of Rio Grande do Sul (UFRGS), Av. Bento Gonçalves 9090, Bairro Agronomia, Porto Alegre CEP 51900-000, Brazil E. F. Costa : L. G. Corbellini Laboratory of Veterinary Molecular Biology, Federal University of Mato Grosso (UFMT), Av. Fernando Corrêa da Costa, 2367, Bairro Boa Esperança, Cuiabá, Mato Grosso CEP 78069-900, Brazil
of the risk factors associated with intravaginal P4 device transmission should be performed. Keywords Mato Grosso . PCR . Ureaplasma diversum . Vulvovaginitis
Introduction Mycoplasma and Ureaplasma species are included within the class Mollicutes, which comprises 5 orders, 6 families, 14 genera, and nearly 200 know species, but only a few of these are considered pathogenic for livestock (Brown 2010, pp. 253–404). In cattle, two species in the genus Mycoplasma, Mycoplasma bovis and Mycoplasma bovigenitalium, have been associated with disease characterized by pneumonia, arthritis, mastitis, and reproductive disorders (Petit et al. 2008, pp. 325–333). Ureaplasma diversum is a bacterium that belongs to the Mycoplasmataceae family and is associated with a variety of reproductive illnesses including granular vulvitis, infertility, abortion, and weak neonates (Miller et al. 1994, pp. 479–490). However, the relationship between the presence of U. diversum in vaginal cultures from cows and disease is controversial because some herds with high rates of Ureaplasma have maintained normal conception rates (Marques et al. 2013, pp. 670–674). The transmission of this microorganism has been reported to occur by natural and artificial insemination with contaminated semen (Miller et al. 1994, pp. 479–490; Buzinhani et al. 2011, pp. 455) and by embryo transfer, which is thought to be a potential mechanism of the dissemination of this organism (Miller et al. 1994, pp. 479–490). Although infections caused by U. diversum in Brazil have been reported to be more frequent in dairy cows with reproductive disorders or poor reproductive performance that range
Trop Anim Health Prod
from 23 to 41 % (Cardoso et al. 2000, pp. 137–143; dos Santos et al. 2013, pp. 315–318), there is limited information about the presence of this microorganism among beef cattle in Mato Grosso State (Gambarini et al. 2009, pp. 1,421–1,426) or its effect on bovine fertility. The aims of this study were to verify the presence of U. diversum and to elucidate its possible relationships with explanatory variables in a beef herd population located in the Middle-West region of Brazil.
U. diversum in the vulvovaginal mucus samples from the heifers. The vaginal mucus samples were collected via the introduction of a sterile cottons swab into the vestibule of the vagina followed by the application of friction to the lateral walls (Cardoso et al. 2000, pp. 137–143). The samples were placed in tubes containing 2 ml of A3XB transport medium according to the protocol of Cunha et al. (1987) and kept refrigerated at 4 °C for up to 48 h until analysis in the laboratory. DNA extraction and PCR
Materials and methods Study population The study was conducted in a beef heifer batch of 378 animals on a farm situated in the Vale Guapore region of the State of Mato Grosso, Brazil. Mato Grosso is one of the largest beef cattle production regions in the country. The study population was composed of the following breeds: Aberdeen Angus (n= 126) and Hereford (n=252). All heifers of the batch were in good nutritional condition and were maintained on improved pastures (Brachiaria brizantha) and received mineral mixture and water ad libitum. The heifers were vaccinated against only Brucella abortus and exclusively served an embryo transfer program in which intravaginal progesterone (P4) devices were commonly used in the reproductive routine. Diagnoses of pregnancy were based on ultrasound examinations that were performed 30 days after embryo transfer. Sample collection A total of 378 heifers were selected for the embryo transfer program. The vaginal mucus samples were collected by a simple random sampling of 134 heifers. After the draw, evaluations of the presence or absence of clinical signs were performed. The samples were randomly collected, and the sample size was calculated based on a prevalence of 20 % (Oliveira Filho et al. 2005, pp. 151–156), an absolute error of 5 %, and a population size of 378 heifers using the following formula by Thursfield (2013): z2 ½pð1−pÞN ; d 2 ðN −1Þ þ z2 ½pð1−pÞ where n is the sample size, N is the population size, Z95 is the value used for the 95 % confidence interval, p is the prevalence, and d is the absolute error. Additionally, data regarding age, breed, vulvar macroscopic lesions, number of embryo transfer, number of abortions, pregnancy rates, and progesterone device application (P4) were recorded in an MS Excel database. Polymerase chain reactions (PCRs) were used to determine the presence of
To extract U. diversum DNA, 1 ml of swab transport medium was centrifuged (20,000×g, 4 °C, 5 min). Next, the pellets were resuspended in 500 μl of lysis buffer (1 mM EDTA; 10 mM Tris–HCl, pH 8.0; 0.1 % Triton X-100, 200 μg/ml Proteinase K; Sambrook and Russel 2001). Extractions with phenol and phenol-chloroform were performed, and the nucleic acids were precipitated from the aqueous phase via the addition of sodium acetate and 2 vol of cold 95 % ethanol. After centrifugation, the obtained pellets were washed once with 70 % ethanol, and the nucleic acids were solubilized in 100 μl of TE buffer. The assay was performed with a reaction mix containing 40 μM of each primer (UD1 and UD2) as described by Cardoso et al. (2000) in 50 mM KCl, 10 mM Tris–HCl, 1.5 mM MgCl2, 0.4 μl mM dNTP, and 1 unit of Taq polymerase; pH 8.3. Two microliters of extracted DNA was added to the reaction mix for a final volume of 25 μl per tube. After 5 min of incubation at 94 °C, thermal cycling involving 35 rounds of 94 °C for 1 min, 55 °C for 1 min, and 72 °C for 30 s was performed and followed by a final extension at 72 °C for 10 min. Bovine herpesvirus 1 (BoHV-1) DNA extractions were obtained from frozen vulvovaginal swab mucus using proteinase K followed by phenol and chloroform treatment (Sambrook and Russel 2001). BoHV-1 was detected by PCR as described by Bricker and Halling (1995). Negative and positive controls for BoHV-1 were run for each PCR. We used DNA from SV-56/90 isolates of BoHV-1 as a positive control according to Silva et al. 2007. Statistical analyses To verify the association between detection of U. diversum and the explanatory variables, Poisson regressions were performed. The outcome was the presence/absence of U. diversum as detected by PCR, and the explanatory variables were age, breed, vulvar macroscopic lesions, number of embryo transfers, number of abortions, and progesterone device application (P4). First, the variables that explained less than 10 % of the variability and those for more than 10 % of the data were missing were excluded. Next, the variables with
Trop Anim Health Prod
p values ≤0.2 were selected for inclusion in the univariate model. Correlation analyses of these pre-selected variables were performed. When correlations >0.7 were found, only the variables with biological plausibility and/or the lowest p values were retained in the analysis. Age was included as a confounder, and variation of the estimate for this variable greater than 20 % implies a possible confounder effect. All analysis was performed with the MASS package in R software (R Core Team 2012). Multivariate Poisson regressions were performed with the GLM function. The model was built using the stepwise selection method based on AIC values. The fit of the final model was tested using the residual deviance and a significance level of 5 %. Adjustments for standard errors were performed with the sandwich function package.
Results Beginning in 2011, the female bovines displayed signs of vulvovaginal pustular lesions that were most prominent 7 days after estrous at 24–30 months of age. The conception rate of the entire batch (N=378) was 65 %; the conception rate 30 days after embryo transfer was 58.6 % (34/58) in the heifers with vulvar lesions, and this rate was 23.6 % (18/76) for the heifer without lesions. The average conception rate of all heifers was 39 % (52/134). A total of 134 swabs were collected. Fifty-one (38 %; 95 % CI 30–46 %) of the heifers were PCR positive for U. diversum (Table 1); of these, 37/51 (72.5 %) had vulvovaginal lesions that were characterized by hyperemia, granulated lesions (1– 3 mm in diameter, white or lightly red), and edema distributed throughout the vulvar mucosa (Fig. 1). All of the tested samples were negative for BoHV-1 as assessed with PCR. Of the 37 heifers that were U. diversum positive based on PCR, 24 (64.8 %) were pregnant. Of the 21 U. diversumnegative heifers, 10 (47.6 %) were pregnant. The univariate analyses resulted in the inclusion of vulvar lesion, breed, and the use of a progesterone device (P4) in the multivariate model. Age was not found to be significant in the univariate analysis. However, the variables of breed and progesterone (P4) device exhibited a correlation above 0.7; thus, Table 1 The presence of U. diversum in the vulvovaginal mucus of beef heifers as assessed by PCR Vulvar lesions
Total
U. diversum PCR result
PCR positive PCR negative Total
Presence
Absence
37 21 58
14 62 76
51 83 134
Fig. 1 Cattle, female. Moderate hyperemia and lightly red granulate lesion in the ventral area of the vulvar mucosa
only the P4 device variable was maintained to avoid collinearity problems. In the final model, the residual deviance statistic D had a p value of 0.1, which indicates that the model had a good fit. Two of the four variables (Table 2) included in the multivariate analysis were significantly associated with the presence of U. diversum; these variables were the presence of vulvar lesions (p=0.001) and the use of a progesterone (P4) device (p = 0.001). Increases in the prevalence ratio (PR) of U. diversum were positively associated with the presence of vulvar lesions (PR=3.4) and the use of P4 advice (PR=2.5). Furthermore, despite its high p value of 0.79, the variable age was included as a confounder of P4 device usage and forced into the model.
Table 2 The variables included in the final multivariate model Ureaplasma isolation
Total
PCR positive (%)
Vulvar lesion Yes 58 63.8 No 76 18.4 Progesterone device (P4) Yes 42 66.6 No 92 25 Age 30 24
20 114
40 37
PR (95 % CI)
p value
3.4 (1.7–5.3) 1
REGULAR ARTICLES
Ureaplasma diversum as a cause of pustular vulvovaginitis in bovine females in Vale Guapore, Mato Grosso State, Brazil João Guilherme L. N. Gaeti & Marconni V. C. Lana & Gustavo S. Silva & Letycia Lerner & Camila G. de Campos & Fernanda Haruni & Edson M. Colodel & Eduardo F. Costa & Luis G. Corbellini & Luciano Nakazato & Caroline A. Pescador
Accepted: 1 May 2014 # Springer Science+Business Media Dordrecht 2014
Abstract Ureaplasma diversum has been associated with various reproductive problems in cattle that include granular vulvovaginitis, weak calves, and abortion. This study was conducted in a beef herd situated in the Middle-West region of Brazil, and the objectives were to verify the presence of U. diversum and to elucidate its possible relationships with independent variables in this bovine herd population. A total of 134 vaginal mucous swabs were taken for polymerase chain reaction (PCR). Of these, 51 (38 %) were PCR positive for U. diversum. Of the 58 heifers with vulvovaginal lesions characterized by hyperemia, granulated lesions, and edema distributed throughout the vulvar mucosa, 37 (64 %) were U. diversum positive; of the 76 heifers without reproductive lesions, 14 (18 %) were U. diversum positive. All tested samples were negative for bovine herpesvirus 1 (BoHV-1). Multivariate logistic regression revealed that the following two variables were significantly associated with the presence of U. diversum: the presence of vulvar lesions (p=0.001) and the presence of a progesterone (P4) device (p=0.001). These findings indicate that U. diversum should be considered a pathogen that is associated with pustular vulvovaginitis in heifers from the Mato Grosso state and that additional studies J. G. L. N. Gaeti : M. V. C. Lana : G. S. Silva : L. Lerner : C. G. de Campos : E. M. Colodel : C. A. Pescador (*) Laboratory of Veterinary Pathology, Federal University of Mato Grosso (UFMT), Av. Fernando Corrêa da Costa 2367, Cuiabá, Mato Grosso CEP 78069-900, Brazil e-mail: [email protected] F. Haruni : L. Nakazato Epidemiology Laboratory, Federal University of Rio Grande do Sul (UFRGS), Av. Bento Gonçalves 9090, Bairro Agronomia, Porto Alegre CEP 51900-000, Brazil E. F. Costa : L. G. Corbellini Laboratory of Veterinary Molecular Biology, Federal University of Mato Grosso (UFMT), Av. Fernando Corrêa da Costa, 2367, Bairro Boa Esperança, Cuiabá, Mato Grosso CEP 78069-900, Brazil
of the risk factors associated with intravaginal P4 device transmission should be performed. Keywords Mato Grosso . PCR . Ureaplasma diversum . Vulvovaginitis
Introduction Mycoplasma and Ureaplasma species are included within the class Mollicutes, which comprises 5 orders, 6 families, 14 genera, and nearly 200 know species, but only a few of these are considered pathogenic for livestock (Brown 2010, pp. 253–404). In cattle, two species in the genus Mycoplasma, Mycoplasma bovis and Mycoplasma bovigenitalium, have been associated with disease characterized by pneumonia, arthritis, mastitis, and reproductive disorders (Petit et al. 2008, pp. 325–333). Ureaplasma diversum is a bacterium that belongs to the Mycoplasmataceae family and is associated with a variety of reproductive illnesses including granular vulvitis, infertility, abortion, and weak neonates (Miller et al. 1994, pp. 479–490). However, the relationship between the presence of U. diversum in vaginal cultures from cows and disease is controversial because some herds with high rates of Ureaplasma have maintained normal conception rates (Marques et al. 2013, pp. 670–674). The transmission of this microorganism has been reported to occur by natural and artificial insemination with contaminated semen (Miller et al. 1994, pp. 479–490; Buzinhani et al. 2011, pp. 455) and by embryo transfer, which is thought to be a potential mechanism of the dissemination of this organism (Miller et al. 1994, pp. 479–490). Although infections caused by U. diversum in Brazil have been reported to be more frequent in dairy cows with reproductive disorders or poor reproductive performance that range
Trop Anim Health Prod
from 23 to 41 % (Cardoso et al. 2000, pp. 137–143; dos Santos et al. 2013, pp. 315–318), there is limited information about the presence of this microorganism among beef cattle in Mato Grosso State (Gambarini et al. 2009, pp. 1,421–1,426) or its effect on bovine fertility. The aims of this study were to verify the presence of U. diversum and to elucidate its possible relationships with explanatory variables in a beef herd population located in the Middle-West region of Brazil.
U. diversum in the vulvovaginal mucus samples from the heifers. The vaginal mucus samples were collected via the introduction of a sterile cottons swab into the vestibule of the vagina followed by the application of friction to the lateral walls (Cardoso et al. 2000, pp. 137–143). The samples were placed in tubes containing 2 ml of A3XB transport medium according to the protocol of Cunha et al. (1987) and kept refrigerated at 4 °C for up to 48 h until analysis in the laboratory. DNA extraction and PCR
Materials and methods Study population The study was conducted in a beef heifer batch of 378 animals on a farm situated in the Vale Guapore region of the State of Mato Grosso, Brazil. Mato Grosso is one of the largest beef cattle production regions in the country. The study population was composed of the following breeds: Aberdeen Angus (n= 126) and Hereford (n=252). All heifers of the batch were in good nutritional condition and were maintained on improved pastures (Brachiaria brizantha) and received mineral mixture and water ad libitum. The heifers were vaccinated against only Brucella abortus and exclusively served an embryo transfer program in which intravaginal progesterone (P4) devices were commonly used in the reproductive routine. Diagnoses of pregnancy were based on ultrasound examinations that were performed 30 days after embryo transfer. Sample collection A total of 378 heifers were selected for the embryo transfer program. The vaginal mucus samples were collected by a simple random sampling of 134 heifers. After the draw, evaluations of the presence or absence of clinical signs were performed. The samples were randomly collected, and the sample size was calculated based on a prevalence of 20 % (Oliveira Filho et al. 2005, pp. 151–156), an absolute error of 5 %, and a population size of 378 heifers using the following formula by Thursfield (2013): z2 ½pð1−pÞN ; d 2 ðN −1Þ þ z2 ½pð1−pÞ where n is the sample size, N is the population size, Z95 is the value used for the 95 % confidence interval, p is the prevalence, and d is the absolute error. Additionally, data regarding age, breed, vulvar macroscopic lesions, number of embryo transfer, number of abortions, pregnancy rates, and progesterone device application (P4) were recorded in an MS Excel database. Polymerase chain reactions (PCRs) were used to determine the presence of
To extract U. diversum DNA, 1 ml of swab transport medium was centrifuged (20,000×g, 4 °C, 5 min). Next, the pellets were resuspended in 500 μl of lysis buffer (1 mM EDTA; 10 mM Tris–HCl, pH 8.0; 0.1 % Triton X-100, 200 μg/ml Proteinase K; Sambrook and Russel 2001). Extractions with phenol and phenol-chloroform were performed, and the nucleic acids were precipitated from the aqueous phase via the addition of sodium acetate and 2 vol of cold 95 % ethanol. After centrifugation, the obtained pellets were washed once with 70 % ethanol, and the nucleic acids were solubilized in 100 μl of TE buffer. The assay was performed with a reaction mix containing 40 μM of each primer (UD1 and UD2) as described by Cardoso et al. (2000) in 50 mM KCl, 10 mM Tris–HCl, 1.5 mM MgCl2, 0.4 μl mM dNTP, and 1 unit of Taq polymerase; pH 8.3. Two microliters of extracted DNA was added to the reaction mix for a final volume of 25 μl per tube. After 5 min of incubation at 94 °C, thermal cycling involving 35 rounds of 94 °C for 1 min, 55 °C for 1 min, and 72 °C for 30 s was performed and followed by a final extension at 72 °C for 10 min. Bovine herpesvirus 1 (BoHV-1) DNA extractions were obtained from frozen vulvovaginal swab mucus using proteinase K followed by phenol and chloroform treatment (Sambrook and Russel 2001). BoHV-1 was detected by PCR as described by Bricker and Halling (1995). Negative and positive controls for BoHV-1 were run for each PCR. We used DNA from SV-56/90 isolates of BoHV-1 as a positive control according to Silva et al. 2007. Statistical analyses To verify the association between detection of U. diversum and the explanatory variables, Poisson regressions were performed. The outcome was the presence/absence of U. diversum as detected by PCR, and the explanatory variables were age, breed, vulvar macroscopic lesions, number of embryo transfers, number of abortions, and progesterone device application (P4). First, the variables that explained less than 10 % of the variability and those for more than 10 % of the data were missing were excluded. Next, the variables with
Trop Anim Health Prod
p values ≤0.2 were selected for inclusion in the univariate model. Correlation analyses of these pre-selected variables were performed. When correlations >0.7 were found, only the variables with biological plausibility and/or the lowest p values were retained in the analysis. Age was included as a confounder, and variation of the estimate for this variable greater than 20 % implies a possible confounder effect. All analysis was performed with the MASS package in R software (R Core Team 2012). Multivariate Poisson regressions were performed with the GLM function. The model was built using the stepwise selection method based on AIC values. The fit of the final model was tested using the residual deviance and a significance level of 5 %. Adjustments for standard errors were performed with the sandwich function package.
Results Beginning in 2011, the female bovines displayed signs of vulvovaginal pustular lesions that were most prominent 7 days after estrous at 24–30 months of age. The conception rate of the entire batch (N=378) was 65 %; the conception rate 30 days after embryo transfer was 58.6 % (34/58) in the heifers with vulvar lesions, and this rate was 23.6 % (18/76) for the heifer without lesions. The average conception rate of all heifers was 39 % (52/134). A total of 134 swabs were collected. Fifty-one (38 %; 95 % CI 30–46 %) of the heifers were PCR positive for U. diversum (Table 1); of these, 37/51 (72.5 %) had vulvovaginal lesions that were characterized by hyperemia, granulated lesions (1– 3 mm in diameter, white or lightly red), and edema distributed throughout the vulvar mucosa (Fig. 1). All of the tested samples were negative for BoHV-1 as assessed with PCR. Of the 37 heifers that were U. diversum positive based on PCR, 24 (64.8 %) were pregnant. Of the 21 U. diversumnegative heifers, 10 (47.6 %) were pregnant. The univariate analyses resulted in the inclusion of vulvar lesion, breed, and the use of a progesterone device (P4) in the multivariate model. Age was not found to be significant in the univariate analysis. However, the variables of breed and progesterone (P4) device exhibited a correlation above 0.7; thus, Table 1 The presence of U. diversum in the vulvovaginal mucus of beef heifers as assessed by PCR Vulvar lesions
Total
U. diversum PCR result
PCR positive PCR negative Total
Presence
Absence
37 21 58
14 62 76
51 83 134
Fig. 1 Cattle, female. Moderate hyperemia and lightly red granulate lesion in the ventral area of the vulvar mucosa
only the P4 device variable was maintained to avoid collinearity problems. In the final model, the residual deviance statistic D had a p value of 0.1, which indicates that the model had a good fit. Two of the four variables (Table 2) included in the multivariate analysis were significantly associated with the presence of U. diversum; these variables were the presence of vulvar lesions (p=0.001) and the use of a progesterone (P4) device (p = 0.001). Increases in the prevalence ratio (PR) of U. diversum were positively associated with the presence of vulvar lesions (PR=3.4) and the use of P4 advice (PR=2.5). Furthermore, despite its high p value of 0.79, the variable age was included as a confounder of P4 device usage and forced into the model.
Table 2 The variables included in the final multivariate model Ureaplasma isolation
Total
PCR positive (%)
Vulvar lesion Yes 58 63.8 No 76 18.4 Progesterone device (P4) Yes 42 66.6 No 92 25 Age 30 24
20 114
40 37
PR (95 % CI)
p value
3.4 (1.7–5.3) 1
