Materials and Methods

Animals—One hundred seventy Holstein dairy calves (68 heifer calves and 102 bull calves) born sequentially from cows of various parities with confirmed breeding dates were selected from the University of Missouri Foremost Teaching and Research Dairy for use in the study. This herd used artificial insemination exclusively and recorded all estrus and breeding events. Pregnancy diagnosis was performed on all cows within 45 days after breeding. Consequently, accurate calving dates were available for all cows. Adult cows were vaccinated annually with a multivalent vaccine (infectious bovine rhinotracheitis virus, BVDV types 1 and 2, parainfluenza-3 virus, bovine respiratory syncytial virus, Campylobacter fetus, and Leptospira spp [serovars Canicola, Grippotyphosa, Hardjo, Pomona, and Icterohemorrhagiae]).^a Additionally, cows were vaccinated with an Escherichia coli^b bacterin at the end of lactation, 1 month before parturition, and at parturition. Calves were vaccinated with a multivalent vaccine (infectious bovine rhinotracheitis virus, BVDV type 1 and 2, parainfluenza-3 virus, and bovine respiratory syncytial virus)^c at 2, 4, 6, and 12 months of age. The study was approved by the University of Missouri, Columbia, Animal Care and Use Committee.

Sample collection—Only calves whose births were observed were enrolled in the study. After parturition, each calf was immediately separated from its dam, weighed, and assigned a unique identification number. Heifer and bull calves were enrolled in the study. Serum samples were obtained from all calves within 1 hour after birth but before calves were provided colostrum. Samples were stored at −20°C until processed for serum immunoglobulin determinations and serologic testing to detect antibodies against BVDV and Neospora caninum.

Serum IgG determination—Total serum IgG concentrations were determined by adaptation of a radial immunodiffusion technique described elsewhere.^7,8 Radial immunodiffusion plates for measuring IgG were prepared by dissolving 1% agarose^d in a sodium barbital buffer^e containing 0.1% sodium azide.^f Rabbit–anti-bovine IgG^g (1%) was added to thawed agarose solution. Eleven milliliters of the agarose solution was added to 10-cm Petri dishes. After the agarose solidified, 3-mm wells were cut in the agar. Serum samples were diluted 1:2 with barbital buffer, and 5 μL was inoculated in each well. Plates were incubated for 72 hours at 23°C, and diameter of the zone of precipitation was then recorded. Sample IgG concentrations were determined by comparing the diameter of zones of precipitation with a standard curve generated by use of serial dilutions of a bovine IgG standard.^h The regression equation generated for this technique can accurately predict inoculum IgG concentration. Lower detection limit of the radial immunodiffussion assay was 16 mg/dL.

Serologic testing—A serum neutralization test for BVDV type 1 was performed. Serologic titers against BVDV were determined by adaptation of a technique reported elsewhere.⁹ Briefly, titers were determined for heat-inactivated (56°C) serum samples (in duplicate), starting at a dilution of 1:4. The serum-virus mixture was incubated for 1 hour at 37°C. Trypsinized MadinDarby bovine kidney cells (0.05 mL of cell suspension) in medium were added to microtiter plates containing growth media (2mM L-glutamine and Earle's buffered saline solution containing 1.5 g of sodium bicarbonate/L, 0.1mM nonessential amino acids, 1.0mM sodium pyruvate, and 10% horse serum). Microtiter plates were incubated for 48 to 72 hours at 37°C in a humid, 5% carbon dioxide environment. Test sample wells were assessed for cytopathic effect at 48 to 72 hours by use of an inverted ocular of a light microscope. Endpoint antibody titer was defined as the highest dilution of serum at which virus neutralization was detected.

A commercial competitive ELISAⁱ for detecting antibody against N caninum also was performed on all serum samples from calves, as described elsewhere.¹⁰ Diagnostic sensitivity and specificity of the competitive ELISA were 97.6% and 98.6%, respectively, by use of a cutoff value of 30% inhibition.¹⁰

Data analysis—The proportion of calves with a detectable serum IgG concentration and the mean ± SEM serum IgG concentration in calves with a detectable IgG concentration were calculated. The RR, AR, PAR, and PAF for calves that had a detectable serum IgG concentration attributable to a positive result for serologic testing to detect N caninum or BVDV also were calculated.^11,12 Logistic regression was used to determine whether dam parity, calf sex, calf weight, and season of calving were associated with detection of precolostral IgG concentrations (dependent variable).^j Detection or nondetection of an IgG concentration was treated as a binomial variable (0, serum IgG was not detectable; 1, detectable serum IgG concentration). Dam parity was treated as a categoric independent variable (first lactation, second lactation, and ≥ 3 lactations). Calf sex was treated as an independent binomial variable (0, male; 1, female). Calf weight was treated as a continuous independent variable. Season of calving was treated as an independent categoric variable and was divided into 4 seasons (winter [December, January, and February], spring [March, April, and May], summer [June, July, and August], and autumn [September, October, and November]). Variables were considered for inclusion in a stepwise logistic regression model when the value to enter was P < 0.05 by use of the Wald-C statistic.^j The variable with the smallest P value to enter was added to the model at each step. The goodness-of-fit of the final model was estimated through the Hosmer-Lemeshow χ²-test statistic.^j

Results

The regression equation generated by comparing the diameter of zones of precipitation with a standard curve generated by use of serial dilutions of a bovine IgG standard accurately predicted IgG concentrations (R² = 0.98; Figure 1). Ninety (52.9%) calves had a detectable serum IgG concentration (IgG ≥ 16 mg/dL). Mean ± SEM serum IgG concentration for calves that had a detectable IgG concentration was 63.9 ± 4.9 mg/dL. The range of serum IgG concentrations in calves with a detectable serum IgG concentration prior to ingestion of colostrum was 16 to 234 mg/dL. The distribution of serum IgG concentrations in calves was determined (Figure 2).

View Full Size

Figure 1—Plot of serial dilutions of a standard bovine IgG and the resulting diameter of the zone of precipitation. Linear regression analysis revealed a high correlation (R² = 0.98). Equation for the linear regression was as follows: IgG concentration = 0.0209 + (0.0039 × [zone diameter]²).

Citation: American Journal of Veterinary Research 69, 6; 10.2460/ajvr.69.6.791

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Histogram of the relative frequency of various IgG concentrations in serum samples obtained from 170 Holstein dairy calves prior to ingestion of colostrum.

Citation: American Journal of Veterinary Research 69, 6; 10.2460/ajvr.69.6.791

• Download Figure
• Download figure as PowerPoint slide

One of 170 (0.6%) calves had positive results for a serum neutralization test (serum neutralization titer of 1:4) to detect BVDV type 1. Seven of 170 (4.1%) calves had positive results when serologically tested for N caninum. All serum samples that had positive results for BVDV type 1 or N caninum, except for 1 (positive result for N caninum), had a detectable serum IgG concentration. The RR, AR, PAR, and PAF for calves that had a detectable serum IgG concentration attributable to a positive result for N caninum were 1.66, 0.34, 0.014, and 0.03, respectively. The RR, AR, PAR, and PAF for calves that had a detectable serum IgG concentration attributable to a positive result for BVDV were not reported because only 1 calf had a positive result for BVDV for the serum neutralization test.

A logistic regression equation for the probability that a precolostral calf would have a detectable IgG concentration was generated: P_IgG = 1/(1 + exp[0.1457 − {0.7996 × second parity}]), where P_IgG is the probability of a detectable IgG concentration and exp is the exponential function. The probability that a precolostral calf from a second-lactation cow would have a detectable IgG concentration was 0.66, whereas the probability that a precolostral calf from a first-lactation cow or a cow in her third or greater lactation would have a detectable IgG concentration was 0.46. Thus, precolostral calves from second-lactation cows were 1.4 times as likely to have a detectable IgG concentration (RR = 0.66/0.46 = 1.4) as were precolostral calves from a firstlactation cow or a cow in her third or greater lactation. Calf sex, calf birth weight, and season of calving were not significant predictors of a detectable precolostral serum IgG concentration.

Discussion

A detectable serum IgG concentration in calves before ingestion of colostrum indicates transplacental transfer, de novo IgG production by the fetus after transplacental exposure or transcervical exposure, or an invalid assay for measurement of serum IgG concentrations. The linear regression equation for predicting IgG concentration in serum samples was generated by use of a broad range of standard concentrations of bovine IgG. Results clearly indicated a linear dose-response relationship and lack of systematic error at IgG concentrations as low as 16 mg/dL (Figure 1). Hence, transplacental transfer of IgG or production of IgG by the fetus appears to be more likely. Possible reasons for the wide range of IgG concentrations include the type of antigen, duration of exposure, and age of the fetus at the time of exposure to the antigen. An increase in IgG and IgM concentrations has been detected in serum samples obtained from precolostral calves naturally or experimentally infected with BVDV, Coxiella burnetii, Campylobacter fetus, Chlamydia organisms, and blue tongue virus, compared with immunoglobulin concentrations for control bovine fetuses and uninfected precolostral calves.⁵ It should be mentioned that only serum IgG was determined in the study reported here, whereas serum IgM and IgA concentrations were determined in the aforementioned study.⁵ Additionally, IgG subclassification was not performed. To further evaluate potential causes of precolostral serum IgG concentrations, infectious agents that can be transmitted transplacentally were considered. Common infectious agents that can be transmitted in utero in cattle include N caninum,¹³ MAP,¹⁴ BLV,¹⁵ and BVDV.¹⁶

The study herd consisted of 200 lactating dairy cows. The seroprevalence of MAP for the study herd was 6%, as determined on the basis of results for an ELISA.^k Reported rates of transplacental MAP transmission range from 21% to 37% in cows with clinical signs of paratuberculosis.^17,18,19 However, a low prevalence (8.6%) of fetal infections has been detected in calves born to cows without clinical signs of paratuberculosis.²⁰ Consequently, MAP transmission and de novo serologic responses are unlikely causes of common detectable serum IgG concentrations in precolostral calves. Reported rates of transplacental BLV transmission range from 3% to 20%,¹⁵ with lower estimates reported more often.^21,22,23 Accepting that there was an in utero transmission rate of 10%, as well as a reported BLV prevalence of 80% in this herd,²⁴ it would appear that BLV infection was an unlikely cause of serum IgG concentrations in precolostral calves because only 8% of calves would have been infected at birth. In addition, another study²⁵ that involved this herd revealed a transplacental transmission rate of 0% in 32 cows. It is important to mention that serologic tests used in the study reported here were qualitative. Thus, it is possible that the serologic tests may have required a higher IgG concentration than was detected in our study to be considered serologically positive. Consequently, precolostral IgG concentrations against the agents tested may have existed but not been detected by the serologic assays used in the study; thus, the number of calves with IgG attributable to each of the agents may have been underestimated.

Cattle in the study herd were annually vaccinated against BVDV by use of a multivalent modifiedlive vaccine that contained both type 1 and type 2.^a All cattle, regardless of pregnancy status, were vaccinated. Only 1 (0.6%) calf had positive results when tested to detect antibodies against BVDV type 1. If a positive serologic result was attributable to transfer of immunoglobulins from dam to calf in utero as a result of vaccination, all calves should have had positive results for BVDV type 1. Vaccination with BVDV type 1 protects a fetus from infection with heterologous virulent strains.¹⁶ Results for that single calf suggest that in utero BVDV infection and subsequent response to the infection in utero by fetuses is an uncommon event in this herd.

Seroprevalence for N caninum varies with country, region, and type of serologic test used.^26,27 Historical prevalence of N caninum on the farm for the study reported here was 50%, as determined on the basis of whole-herd serologic testing performed before study samples were acquired. Hence, 85 precolostral calves typically would have been expected to be seropositive. However, detected rates of vertical transmission vary,²⁸ and infection can be transmitted intermittently.²⁹ For any 100 calves in the herd, only 2 calves had detectable serum IgG concentrations attributable to positive results for N caninum serologic testing (PAR = 0.014). Consequently, N caninum was considered an unlikely cause of the high prevalence of detectable precolostral serum IgG concentrations.

The time at which a bovine fetus becomes immunocompetent ranges from 145 to 200 days of gestation.^16,30 Results of the study reported here suggested that a significant proportion of calves are gammaglobulinemic at birth prior to ingestion of colostrum. Calves from second-lactation cows were 1.4 times as likely to have a detectable precolostral IgG concentration as were calves from first-lactation cows or cows in their third or greater lactation. The cause for this result remains unknown. It should be mentioned that another study⁶ as well as the study reported here were performed in single herds. Variation among herds with regard to prevalence of precolostral IgG concentrations may exist. On the basis of our results, it appeared that in utero production of detectable concentrations of IgG was more common than has been reported. Serologic testing was unable to identify correlations with any agents that are believed to be commonly transmitted transplacentally in cattle. Consequently, fetal production of IgG was associated with transplacental infection of a specific undetermined pathogen or, alternatively, should be considered a common event associated with in utero exposure and response to a broad range of antigens.