Serum and milk levels of antibodies to bovine viral diarrhoea virus, bovine herpesvirus-1 and -4, and circulation of different bovine herpesvirus-4 genotypes in dairy cattle with clinical mastitis

Bovine viral diarrhoea virus (BVDV) and bovine herpesvirus-1 (BoHV-1) infections are common in cattle populations worldwide. The cytopathogenic and non-cytopathogenic biotypes of BVDV, which is in the Pestivirus genus of the Flaviviridae family, cause both transient and persistent infections (25), and BoHV-1 is an alphaherpesvirus that establishes latency in the neuronal ganglia following the primary acute infection. The virus shed after reactivation under stress factors can infect naive animals (8, 26, 27).

Like BVDV and BoHV-1, bovine herpesvirus-4 (BoHV-4) has been reported to be among the causative agents of reproductive or respiratory system disorders characterised by infertility, abortion, repeat breeding, respiratory distress, and neonatal calf mortality (5, 10). As a gammaherpesvirus, it exhibits latency in monocyte/ macrophage lineage cells. This virus also plays a cofactor role in the escalation of clinical episodes and the extension of recovery periods in cases of postpartum metritis and mastitis caused by the primary bacterial agents such as Streptococcus uberis, Escherichia coli, Staphylococcus aureus, Streptococcus dysgalactiae, Streptococcus agalactiae, Trueperella pyogenes and viral agents such as BVDV and BoHV-1 (2, 3, 5, 10, 19).

A microorganism-related inflammation of the mammary glands or quarters, mastitis is a major problem that negatively affects dairy cattle worldwide and has serious economic repercussions. The primary causes of mastitis are a wide range of bacterial species (2); however, it has been reported that foot and mouth disease virus, parainfluenza virus-3, BVDV, BoHV-1 and BoHV-4 have also been isolated from cases of natural clinical mastitis (7, 19, 20, 28). If the suffering animal’s immunological activity is reduced or the virulence of the pathogenic microorganisms is high, a partially eliminated infection may persist for a longer time, often presenting as subclinical or chronic mastitis (21). Bovine immunodeficiency virus, bovine leukaemia virus, BVDV and BoHV-1 infections in particular may play indirect roles in bovine mastitis through immunosuppression (28). Although BoHV-4 is not a direct mammary pathogen, it has been shown to reach the mammary epithelium after an inflammatory process in the mammary tissues and replicate both in the mammary epithelium and in the immune cells in the udder (2, 3, 19).

Humoral immune components such as immunoglobulins A and G are critical to the immunological defence response of the cow udder (21). Some studies reported that udder inflammation affects the concentration of immunoglobulins in serum and milk, and there are many factors that regulate and affect the mammary–blood barrier, which determines the migration of immunoglobulins from the bloodstream to the mammary gland tissue (9, 21, 23).

In this study, we aimed not only to correlate anti-BVDV, anti-BoHV-1 and anti-BoHV-4 antibody levels in serum and milk from clinically mastitic dairy cattle but also to determine the presence of and characterise genetically BoHV-4 in clinical mastitis cases in these animals.

Study design and population. The veterinarian of a private herd farmed intensively stated that individual milk samples had been sent to a private laboratory several times because new mastitis cases had been emerging daily, and that only Strep. agalactiae and Strep. dysgalactiae had been detected using diagnostic techniques, rather than Mycoplasma spp., Staph. aureus, E. coli, or other bacterial agents. He stated that despite using antibiotics according to the antibiogram results, there had been no significant positive effect, and the incidence of mastitis had continued to increase (personal communication with Fevzi Akkan, DVM). The veterinarian sent whole blood, serum and milk samples from 59 mastitic cows in this herd to the laboratory of the Department of Virology, Faculty of Veterinary Medicine at the Hatay Mustafa Kemal University to be evaluated for viral agents, this quantity comprising samples from all mastitic cows. Blood and milk samples were also taken from 16 healthy cows in a commercial organic dairy herd certified as free from major diseases (e.g. brucellosis, tuberculosis or enzootic bovine leucosis) by the Ministry of Agriculture and Forestry of the Republic of Türkiye after the herd had been screened for subclinical mastitis with a California mastitis test. This sampling was carried out in order to establish a controlled research standard while evaluating the blood-to-milk transition levels in the mastitic group under field conditions. Both groups had routine vaccination schedules with a commercial vaccine containing BoHV-1 and BVDV (Hiprabovis 4; Laboratorios HIPRA, Amer, Spain). When all samples reached the laboratory, the mastitic and healthy groups had been vaccinated approximately 14 weeks and 16 weeks previously, respectively. The mean age and the number of lactations of the group with clinical mastitis were 50.40 ± 17.09 months and 2.19 ± 1.28, and the mean age and the number of lactations of the healthy group were 57.70 ± 25.42 months and 3.00 ± 1.90, respectively.

Detection of BVDV antigen. The whole blood samples were tested for the BVDV antigen using a commercial antigen ELISA kit (BVDV Ag/ Serum Plus Capture ELISA; IDEXX, Liebefeld-Bern, Switzerland) according to the manufacturer’s instructions.

Detection of BVDV, BoHV-1, and BoHV-4 antibodies. A commercial BVDV antibody ELISA kit was used to screen for BVDV (BVDV Total Antibody Test; IDEXX, Liebefeld-Bern, Switzerland) with specificity of 99.4% and sensitivity of 98.5%, a BoHV-1 antibody ELISA kit was selected for BoHV-1 (CIVTEST Bovis IBR gB blocking ELISA; Laboratorios HIPRA, Amer, Spain) with specificity of 97% and sensitivity of 93.8%, and a BoHV-4 antibody ELISA kit was suitable for BoHV-4 detection (Monoscreen Ab ELISA; Bio-X Diagnostics, Rochefort, Belgium) with specificity of 92.93% and sensitivity of 92.98%. Serum and milk samples were analysed with these kits according to the manufacturer’s instructions.

The sera were diluted 1 : 5 for BVDV, 1 : 2 for BoHV-1, and 1 : 100 for BoHV-4 antibodies, while the milk samples were tested without dilution. At the end of all three tests, optical density values were read using an ELISA reader (BioTek μQuant; BioTek Instruments, Winooski, VT, USA) at the 450 nm wavelength.

Polymerase chain reaction and sequencing. All milk samples were tested for BoHV-1 and BoHV-4 DNA by PCR using the primer sets described by Esteves et al. (15) and Wellenberg et al. (29), respectively (Table 1). A 575-base-pair (bp) conserved region of the glycoprotein C (gC) gene of BoHV-1 and a 615-bp conserved region of the gB gene of BoHV-4 were amplified by PCR after total DNA extraction from 200 μL volumes of the milk samples using a High Pure Viral Nucleic Acid kit (Roche Diagnostics, Mannheim, Germany).

A commercial master mix solution (5× HOT FIREPol Blend PCR Master Mix; Solis BioDyne, Tartu, Estonia) including Hot FIREPol DNA polymerase as a proofreading enzyme, 5× blend master mix buffer, 15 mM MgCl₂, 2 mM dNTPs, bovine serum albumin, blue dye and yellow dye was used in the amplification stage of both viruses. Amplification was performed in a 20 μL reaction mixture by adding 2 μL of template DNA to 18 μL of PCR master mix containing a final concentration of 0.5 μM of each primer. The amplification programme for BoHV-1 and BoHV-4 was: an initial denaturation at 95°C for 15 min followed by 35 cycles consisting of denaturation at 95°C for 20 s, annealing at 56°C for 1 min and extension at 72°C for 1 min. The reaction was terminated after a final hold at 72°C for 7 min. All PCR products were visualised in a transilluminator after electrophoresis in 1% agarose gel containing Safe-Red DNA stain (SafeView cat no. G108-R; Applied Biological Materials, Richmond, BC, Canada).

The amplified BoHV-4 DNA products were sequenced using an automatic sequence genetic analyser (CEQ 8000; Beckman Coulter, Brea, CA, USA) by a commercial company (BM Lab, Ankara, Türkiye). The gB sequence assembly and editing were performed using Bioedit (Version 7.0.5.3) (17). A phylogenetic tree was created with different BoHV-4 strains using the neighbour-joining method as implemented in Clustal W in MEGA 7.0 software (22). The bootstrap values were calculated from 1,000 replicates.

Statistical analysis. Descriptive statistics are shown as mean ± standard deviation. The normality assumption of the data was checked with the Shapiro– Wilk Test. Accordingly, the BoHV-1 percent positivity (PP) value, BVDV sample/positive (S/P) ratio, and BoHV-4 PP in both serum and milk samples did not show normal distribution. We assumed that the serum and milk samples were mutually dependent, since each sample of serum and milk was collected from the same animal. Since the two streams of data were considered to be dependent and showed non-normal distribution, we used generalised estimating equations (GEE) with robust standard errors to model the difference in BVDV and BoHV-1 levels between serum and milk samples. We also used the health status of the animals (healthy/clinical mastitis) as a factor in the GEE model. The Wilcoxon signed-rank test was used to compare BoHV-4 levels between serum and milk samples. Also, Bland–Altman plots showing the differences between serum and milk samples against the average PP and S/P index values were drawn to evaluate the variability between serum and milk samples in dairy cattle with clinical mastitis and healthy dairy cattle. We considered a P-value smaller than 0.05 as statistically significant. All statistical analyses were performed using IBM SPSS Statistics for Windows, version 23.0 (Armonk, NY, USA).

Serological results. Serum and milk antibodies of BVDV and BoHV-1 were detected in all cattle with clinical mastitis and all healthy cattle (100%). However, none of them were positive for BVDV antigen or BoHV-1 DNA. The antibody titres to BVDV and BoHV-1 were lower in milk than in sera from both groups.

For BVDV antibodies, there was a statistically significant difference in S/P indices between the blood and milk samples, regardless of health status (P < 0.001). Evaluating blood and milk samples together, a statistically significant difference in S/P indices existed between healthy and sick animals (P = 0.003). Nevertheless, there was no statistically significant difference between healthy and mastitic cattle in terms of blood-to-milk transition levels (P = 0.156).

For BoHV-1 antibodies, a statistically significant difference was observed in PP between blood and milk samples, regardless of health status (P < 0.001). Analysing blood and milk samples together, a statistically significant difference emerged between healthy and sick animals in terms of PP values (P < 0.001). Additionally, blood-to-milk transition levels were higher among sick animals than healthy ones (P < 0.001) (Table 2 and Fig. 1).

Fig. 1

The Bland–Altman analysis revealed a bias of 1.19 in anti-BVDV antibodies and 9.58 in anti-BoHV-1 antibodies between serum and milk samples from healthy cows. In cows with mastitis, there was a bias of 1.42 in anti-BVDV antibodies and 0.94 in anti-BoHV-1 antibodies between serum and milk. A statistical agreement between milk and serum samples was only found for BoHV-1 in cattle with clinical mastitis. The interpretation values of the milk and serum samples were markedly higher than the cut-off values for BoHV-1 in the healthy group and for BVDV in both groups. Therefore, it would be correct to consider the serum and milk samples to be clinically consistent (Figs 2 and 3).

Fig. 2

Fig. 3

Antibodies specific to BoHV-4 were only detected in the group with clinical mastitis. Also, anti-BoHV-4 antibody levels were higher in milk than in sera obtained from dairy cattle with clinical mastitis. For BoHV-4, the differences between serum and milk samples in PP and seropositivity degree were statistically significant (P < 0.001) (Table 3). The Bland–Altman plot showed a bias of 34.79 in BoHV-4 (Fig. 4).

Fig. 4

Molecular detection and phylogenetic analysis. Detection was achieved of BoHV-4 DNA in four milk samples (6.8%) in dairy cattle with clinical mastitis by PCR. Partial sequences of the gB gene of these four Turkish (TR) BoHV-4 strains were compared with the BoHV-4 strains registered in GenBank and with each other and deposited in GenBank (ON156604– ON156607) (Fig. 5). The 414-B4-TR22 strain (ON 156606) affiliated to genotype I. However, 380-B4-TR22 (ON 156607), 549-B4-TR22 (ON 156604) and 324-B4-TR22 (ON 156605) belonged to genotype II. The sequence analysis revealed that the TR genotype II strains in the present study shared 98.0–99.3% nucleotide (nt) sequence identities with each other. The strains of genotypes I and II were respectively 99.3% and 98.5–99.1% identical to GenBank BoHV-4 strain sequences, including previously recorded TR BoHV-4 strains.

Fig. 5

In the present study, we detected BVDV-, BoHV-1-,and BoHV-4-specific antibodies in all serum and milk samples. Bovine viral diarrhoea virus- and BoHV-1-specific antibody titres were well above the cut-off positivity values in both blood and milk, and this extent of antibody production was thought to be vaccine induced, since both herds had been vaccinated regularly and recently against BVDV and BoHV-1. Moreover, the antibody titres to BVDV and BoHV-1 exhibited good correlation in both individual serum and milk as described elsewhere (4, 24), although the titres were relatively lower in milk than in serum from all cattle in our groups. Similarly, it was reported that the anti-BVDV antibody titre was lower in milk than in serum in relation to the lactation period (23). Data for the lactation period was not obtained in the present study, and detailed studies are needed to evaluate the effect of this factor on the correlation between antibody titres in milk and serum.

Anti-BoHV-4 antibodies were detected only in the clinical mastitis group, and virus nucleic acid was detected in the milk samples of four antibody-positive animals by PCR. Interestingly, the cases of clinical mastitis within the same herd were caused by different BoHV-4 genotypes, and virus was also detected in milk despite the presence of antibodies. It should be emphasised that the protection of virus-specific antibodies is controversial (12), and the status of infection as latent, acute or reinfection was difficult to determine in these cattle. However, anti-BoHV-4 antibody levels were higher in milk than in sera sampled from dairy cattle with clinical mastitis. This agrees with the results of a previous study comparing Schmallenberg virus antibody levels in serum and milk (11). We believe that the differences in serum-to-milk antibody transition levels were due to the passive transport of inactivated vaccine-derived BVDV- and BoHV-1-specific antibodies to the mammary tissue, whereas antibodies against BoHV-4 might be production triggered by natural infection. Several studies have reported that immunoglobulins in lacteal secretions are produced locally by plasma or epithelial cells in the mammary gland and are transferred passively from blood to the udder as a result of the blood–mammary barrier being compromised during mastitis (7, 13, 16, 18, 19, 28). However, there is an increase in BoHV-4 prevalence with increased lactation stage, lactation number, sex, age, management type, and post-partum metritis (1, 6, 14). In this study, the mean lactation stage for animals with clinical mastitis was 2.19 ± 1.28, the mean age 50.40 ± 17.09 months, and the management type intensive farming, all of which may have greatly increased the probability of exposure to BoHV-4 infection. Information on the lactation period could not be obtained for either herd. Although a previous study also reported the presence of BoHV-4 in healthy cattle (19), we made no such finding. The high biosafety standards may have been instrumental in preventing any detectable BoHV-4 presence, given that the herd was free of many important diseases such as brucellosis, tuberculosis, or enzootic bovine leucosis. Alternatively, this finding may be related to the small number of samples.

Many pathogens have been reported to infect the udder via the ductus papillaris (28). Pathogens in the udder are potentially threatening to calves because colostrum and milk are important in calf nutrition for a certain time after birth, colostrum being particularly so for its maternal antibody and immune cell content which protect newborn calves against infectious agents (21). Neutralising antibodies specific to BoHV-4 can facilitate the transmission of the virus to calves because of their low titre and avidity, but their role in virus transmission by infectious milk is unknown (12). In this study, we detected BoHV-4 in milk, but given that the latency region of the virus was monocyte/macrophage lineage cells, we can hypothesise that milk or colostrum may play a role in the early infection of susceptible calves, as it does in enzootic bovine leukosis. Further clinical and virological research might confirm this hypothesis.

The findings of the present study indicate that clinical mastitis cases in the same herd may have aetiology in different BoHV-4 genotypes. Moreover, since individual serum and milk antibody titres showed good correlation in this study, preventive vaccinations against microbial agents such as BVDV, BoHV-1, Staph. aureus, E. coli and Strep. uberis can help minimise clinical problems like mastitis due to the cofactor role of BoHV-4. Future studies on the epidemiological risk determinants of BoHV-4 will provide a better understanding of infection dynamics in dairy herds, as udder health is a key factor for the health of cows and their offspring and a constant benchmark for evaluating productivity in dairy herds.