INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

In 1998, a program to eradicate bovine herpesvirus 1 (BHV1) will start in The Netherlands. Marker vaccines (1, 5-7) will be used in combination with a BHV1 glycoprotein E (gE) enzyme-linked immunosorbent assay (ELISA) to differentiate infected from vaccinated cattle. The companion diagnostic test detects antibodies against wild-type BHV1, whereas antibodies against the marker vaccine from which gE is deleted are not detected. Van Oirschot et al. (21) have described a gE ELISA for the detection of antibodies directed against the gE of BHV1 in serum. However, large-scale epidemiological screening programs involve the collection and testing of millions of serum samples, and these procedures are highly expensive and laborious. Therefore, several studies have investigated the use of milk samples, which can be collected more easily and inexpensively, for the detection of antibodies against several viruses such as bovine respiratory syncytial virus (3), bovine leukemia virus (2, 8, 12), and bovine viral diarrhea virus (13). These results have shown that milk samples can be used as alternatives to serum samples in large-scale screening programs. Also, several ELISAs for the detection of BHV1 in milk have been described (10, 18, 20), but these ELISAs are unable to differentiate infected from vaccinated animals.

The aim of this study was to examine whether individual milk samples can be used instead of serum samples for the detection of BHV1 gE antibodies. We used two different gE ELISA systems and the standard virus neutralization test (VNT) to examine both serum and milk samples from cattle in BHV1-free herds, unvaccinated herds, and vaccinated herds. The addition of a sodium azide mixture as a preservative as well as the influence of storage on the BHV1 gE antibody detection results were examined for milk.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Test samples. A serum sample and a milk sample were collected at the same time from cows in four Dutch herds certified to be free of BHV1 (n = 155), cows in four Dutch herds with a history of BHV1 infection (n = 203), and cows in four Dutch herds that had outbreaks of BHV1 prior to being vaccinated twice with the attenuated BHV1 gE-negative marker vaccine (n = 111). The serum and milk samples from the vaccinated herds were collected 6 months after the second vaccination. To all individual milk samples a preservative mixture with final concentrations of 0.02% sodium azide and 0.01% bronopol (preservatives), 0.001% Triton X-100 (detergent), and 4 µg of patent blue (color component) per ml was added. Within 1 day after collection, milk samples were defatted by placing the milk samples in a refrigerator for 12 to 18 h, followed by collecting the fraction below the lipid layer. The elimination of fatty compounds is necessary, because lipids can affect the test results. Serum samples were prepared from blood samples by centrifugation at 1,000 × g for 10 min. The defatted milk samples and serum samples were stored at 20°C.

gE-blocking ELISA. The gE-blocking ELISA was a commercially available product (Idexx, Westbrook, Maine) primarily developed for the detection of BHV1 gE antibodies in serum. The principle of this gE-blocking ELISA is based on a blocking method in which the reaction of an epitope on the gE of the BHV1 Lam strain with its corresponding monoclonal antibody (MAb; MAb 66) can be blocked by specific antibodies in the test sample. The analysis of test samples was performed according to the instructions of the manufacturer. Briefly, serum samples in a 1:2 dilution in sample diluent (100 µl) or undiluted defatted milk samples (100 µl) were incubated at 2 to 8°C or 18 to 25°C, respectively, for 15 to 18 h. After incubation, the plates were washed six times with washing solution containing 0.05% (vol/vol) Tween 80 in deionized water. A volume of 100 µl of horseradish peroxidase (HRPO)-labelled anti-BHV1 gE MAb (MAb 66) was added to the wells, and the plates were incubated at 18 to 25°C for 30 min. The plates were washed six times with the washing solution, and 100 µl of substrate-chromogen (H₂O₂-tetramethylbenzidine) solution was added. After an incubation period of 15 min at 18 to 25°C, 100 µl of hydrofluoric acid (0.125%) stop solution was added. The optical density (OD) value was measured at 650 nm with a Bio-Tek Microplate reader (model EL312; Bio-Tek Instruments Inc., Winooski, Vt.). The blocking percentage for each sample was calculated against the OD value for the negative control. The blocking percentages were calculated by the following formula: [(OD₆₅₀ of the negative control  OD₆₅₀ of the test sample)/OD₆₅₀ of the negative control] × 100.

gE ELISA. The specificities of the MAbs and the principle and the procedure of the gE ELISA for the detection of antibodies directed against BHV1 in serum have been described by Van Oirschot et al. (21). Briefly, in the gE ELISA two MAbs (MAb 67 and MAb 75) directed against different antigenic epitopes of the gE of BHV1 were used. MAb 67 reacts with the same antigenic domain on gE as MAb 66. To perform the test, ELISA microplates (catalog no. 655092; Greiner) were coated with 100-µl volumes of a 1:4,000 dilution of MAb 75 in phosphate-buffered saline (pH 7.3) for 18 h at 37°C. The plates were stored at 20°C until use.

VNT. A 24-h VNT was performed as described by Kramps et al. (9). In the VNT, serum samples were analyzed in duplicate in serial twofold dilutions in cell culture medium (starting with a 1:2 dilution). The titer of the test serum was taken as the reciprocal of the highest dilution giving complete inhibition of the cytopathic effect. If both 1:2 serum dilutions did not inhibit the cytopathic effect, the serum was considered negative. In case one of the duplicate samples had a titer of 2 and the other had a titer of 4, the VNT titer was reported as 3. In this study the VNT was used as a standard for the selection of the BHV1-seropositive animals within the unvaccinated herds and to check the response after vaccination, and the VNT results for serum were also used for the evaluation of the value of the BHV1 gE results for milk and serum. The VNT was not suitable for testing milk samples for BHV1 antibodies because of toxic reactions in the 1:2 and 1:4 dilutions.

Detection limit. For the determination of the detection limits of both gE ELISA systems and the VNT, serum and milk samples were collected from the same animals at the same time (n = 10). Only unvaccinated animals that were BHV1 seropositive in the VNT were used, and these animals were randomly chosen. Serum and milk dilutions were analyzed in serial twofold dilutions by using sample diluent and ELISA buffer (gE-blocking ELISA and gE ELISA, respectively) or cell culture medium (VNT). The titer of the test sample was taken as the reciprocal of the highest dilution giving a positive reaction.

Batch control. Prior to the analyses of the test samples, the quality of the BHV1 gE ELISA batches was checked. Each new batch was checked for specificity, sensitivity, and detection limit by using a reference panel of 36 defined positive and negative serum samples, including serum obtained from hypervaccinated cattle, sequential serum samples, and a serum sample comparable to the EU2 serum sample. This quality check was performed to ensure the standardization of the test results and to ensure that the new batches were of high quality. Only qualified BHV1 gE ELISA batches were accepted for use in this study.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Influence of preservative and storage at 20°C. In the gE-blocking ELISA (Fig. 1A) and the modified gE ELISA, the results for the milk samples, with or without preservative, were the same. In the gE-blocking ELISA and the gE ELISA, the correlation coefficients were 0.99 and 0.96, respectively (gE ELISA, y = 0.94 · x + 7.4).

View larger version (20K): [in this window] [in a new window]   FIG. 1.   Blocking percentages in the gE-blocking ELISA of defatted milk samples with or without sodium azide mix (A) and defatted milk samples after storage for 1 day or 32 days at 20°C (B).

Herds certified to be free of BHV1. (i) Serum. All serum samples from herds certified to be free of BHV1 were negative in the gE-blocking ELISA and the VNT. In the gE ELISA one serum sample reacted positively in two different test runs, with blocking percentages of 83 and 81%, respectively.

(ii) Milk. All milk samples from herds certified to be free of BHV1 were negative in the gE-blocking ELISA. To assess the cutoff level for milk samples in the modified gE ELISA, the mean blocking percentage for all negative milk samples (26.8%) and the standard deviation (10.2%) were determined. The calculated cutoff value for milk samples was 47.1% (mean value plus 2 times the standard deviation). On the basis of these results, milk samples with a blocking percentage of 50% were considered to be positive for the presence of antibodies against BHV1 gE. In the modified gE ELISA one milk sample had a blocking percentage higher than 50%. This milk sample was not from the animal that was seropositive by the gE ELISA.

Unvaccinated herds. (i) Serum. The results of both gE-ELISAs for unvaccinated herds were compared with the VNT results (Table 1). All 45 serum samples that reacted positively in the gE-blocking ELISA were positive in the VNT. One sample with a doubtful reaction in the gE-blocking ELISA had a VNT titer of 6. This serum sample reacted positively in the gE ELISA. Seven serum samples with negative results in the gE-blocking ELISA had VNT titers of 3 and 6. Three of these serum samples were positive by the gE ELISA.

(ii) Milk versus serum. For both gE-ELISA systems the results for the milk samples from the unvaccinated herds were compared with the results for the corresponding serum samples. In the gE-blocking ELISA (Fig. 2A) serum samples from 45 animals were positive, and 1 animal had a doubtful reaction, while milk samples from 46 animals reacted positively. Milk samples and the corresponding serum samples from 44 animals were both positive. By the gE ELISA (Fig. 2B) 51 animals were BHV1 gE seropositive, while 39 milk samples were positive by the modified gE ELISA. For 36 animals the results for both milk and serum were positive. In comparison to serum, the relative sensitivities of the gE-blocking ELISA and the modified gE ELISA for testing milk samples were 98 and 68%, respectively.

View larger version (18K): [in this window] [in a new window]   FIG. 2.   Blocking percentages of the milk samples versus blocking percentages of the corresponding serum samples analyzed in the gE-blocking ELISA (A) and the gE ELISA (B) (unvaccinated herds).

Vaccinated herds: milk versus serum. All 111 serum samples from vaccinated herds reacted positively in the BHV1 VNT, indicating that all animals responded to vaccination with the BHV1 (marker) vaccine or have been infected with BHV1. By the gE-blocking ELISA (Fig. 3A), 68 animals were seropositive, while 64 individual milk samples reacted positively. For all 64 animals the corresponding serum samples reacted positively as well. By the gE ELISA, 69 serum samples reacted positively against BHV1 gE antibodies, while 59 milk samples were positive by the modified gE ELISA (Fig. 3B). For all 59 individual milk samples that reacted positively in the gE ELISA, the gE-blocking ELISA test results were positive as well. In comparison to serum, the relative sensitivities of the gE-blocking ELISA and the modified gE ELISA for the testing of milk samples were 94 and 86%, respectively.

View larger version (16K): [in this window] [in a new window]   FIG. 3.   Blocking percentages of milk samples and blocking percentages of the corresponding serum samples analyzed in the gE-blocking ELISA (A) and the gE ELISA (B) (vaccinated herds).

Total relative sensitivity of both gE ELISA systems for testing of milk versus serum. To evaluate the sensitivity of testing of milk with both gE ELISA systems, the total relative sensitivity was calculated and was compared to the results for serum. The total relative sensitivity was calculated from the data that were obtained for milk and the corresponding serum samples collected from both unvaccinated herds and vaccinated herds. By the gE-blocking ELISA, 113 (45 + 68) animals were seropositive, and milk from 108 (44 + 64) animals were positive. By the gE ELISA, 120 (51 + 69) animals were seropositive, and milk from 95 (36 + 59) animals was positive. These data indicate that for the testing of milk samples, the total relative sensitivity of the gE-blocking ELISA was 96%, while for the modified gE ELISA the total relative sensitivity was 79% compared to the results for serum.

Detection limit. The detection limit, defined as the highest dilution that scored a positive reaction, of both gE ELISA systems was determined for serum and milk samples, and the detection limits were compared with the detection limit of VNT for serum. For serum samples the geometric mean titers (reciprocal of the mean logarithmic titers) of the VNT, the gE-blocking ELISA, and the gE ELISA were 269, 69, and 87, respectively. The geometric mean titers in milk samples were 9.8 for the gE-blocking ELISA and 4.9 for the gE ELISA. This means that the mean BHV1 gE antibody titer in milk was 7 times (gE-blocking ELISA) or 18 times (gE ELISA) lower than that in serum (Table 2).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Our study demonstrated that individual milk samples are suitable for use in the detection of antibodies directed against BHV1 gE. In the gE-blocking ELISA, milk samples can be used instead of serum samples for the detection of BHV1-positive animals in infected herds whether they have been vaccinated with the BHV1 marker vaccine or not.

The main problem with the use of milk in the antibody detection tests is the lower concentration of immunoglobulins in comparison with that in serum. According to Mach and Pahud (11), the total amount of immunoglobulin G1 in milk could be 30 times less than the amount in serum (in milligrams per milliliter). Also, the ELISA antibody titers against bovine leukemia virus (8) and bovine respiratory syncytial virus (3) in milk are lower than those in serum. Although the titers of immunoglobulins directed against the BHV1 gE epitopes in milk are lower than those in serum (Table 2), the results indicate that, in comparison to serum, the gE-blocking ELISA is highly sensitive for the testing of milk samples (sensitivity, 0.96), irrespective of whether the samples are collected from unvaccinated or vaccinated cows. In contrast, the modified gE ELISA is less sensitive for the testing of milk samples (sensitivity, 0.76). The results of the determination of the detection limit for milk samples (Table 2) also showed a higher geometric mean titer for the gE-blocking ELISA than for the gE ELISA. These data underline the fact that the gE-blocking ELISA is more sensitive than the gE ELISA for the testing of individual milk samples.

The specificities of both gE ELISA systems for the testing of milk samples is very high. This indicates that the number of false-positive reactions during the eradication program will be very low in both gE ELISA systems.

In case milk samples will be used in large-scale screening programs, the addition of a preservative to the milk and storage at 20°C are necessary, because milk samples cannot always be analyzed on the day of collection. In the gE-blocking ELISA and in the modified gE ELISA, the addition of the sodium azide mixture has no influence on the BHV1 gE results. The results of the gE-blocking ELISA also indicate that defatted milk samples can be stored at 20°C for at least 32 days without having any influence on the BHV1 gE antibodies.

Although our data indicate that milk can be used instead of serum, it should be noted that for the detection of BHV1 antibodies VNT is more sensitive than both gE ELISA systems. This study indicates that serum samples with low VNT titers (8) could be negative in both gE ELISA systems. Perrin et al. (17) also showed that serum samples with low VNT titers can give negative results in the gE-blocking ELISA. The results of the detection limit for serum (Table 2) underline these published data and indicate that in both gE ELISA systems the titers in serum are lower than the VNT titers in serum. It must be assumed that some of the negative BHV1 gE reactions are not due to an intrinsic lack of sensitivity of both gE ELISA systems but are mainly due to a lower antibody response to the antigenic epitope of BHV1 gE than to neutralizing epitopes. However, VNT can obviously not be used to detect infected cattle in vaccinated herds. For that purpose only gE ELISA systems, which detect gE-specific antibodies against wild-type BHV1, are suitable. During BHV1 eradication programs, wherein cattle must be monitored for the presence of BHV1, the lower sensitivity of the gE ELISAs compared with that of VNT can be compensated for by a more frequent testing of the cattle (4). For that purpose milk is the specimen of choice because it is much cheaper and easier to collect milk instead of blood.

In summary, this study shows that in spite of the lower BHV1 gE antibody levels in milk samples than in serum samples, individual milk samples can replace serum samples for the detection of BHV1-positive animals in infected herds, irrespective of whether the herd is vaccinated with a BHV1 gE-negative marker vaccine or not.