Open Access Open Badges Research article

Bovine leukemia virus p24 antibodies reflect blood proviral load

Gerónimo Gutiérrez1*, Hugo Carignano2, Irene Alvarez1, Cecilia Martínez1, Natalia Porta1, Romina Politzki1, Mariela Gammella1, Marina Lomonaco1, Norberto Fondevila1, Mario Poli2 and Karina Trono1

Author Affiliations

1 Instituto de Virología. Centro de Investigaciones en Ciencias Veterinarias y Agronómicas, INTA, C.C. 1712, Castelar, Argentina

2 Instituto de Genética. Centro de Investigaciones en Ciencias Veterinarias y Agronómicas, INTA, C.C. 1712, Castelar, Argentina

For all author emails, please log on.

BMC Veterinary Research 2012, 8:187  doi:10.1186/1746-6148-8-187

Published: 9 October 2012



Bovine leukemia virus (BLV) is worldwide distributed and highly endemic in Argentina. Among the strategies to prevent BLV dissemination, a control plan based on the selective segregation of animals according to their proviral load (PVL) is promising for our dairy productive system. The objective of this work was to study the relationship between the blood PVL and the antibody level, in order to identify whether the individual humoral response, i.e. the anti-p24 or anti-whole-BLV particle, could be used as a marker of the blood level of infection and thus help to recruit animals that may pose a lower risk of dissemination under natural conditions.


The prevalence of p24 antibodies on the 15 farms studied was over 66%. The prevalence of p24 and whole-BLV antibodies and PVL quantification were analyzed in all the samples (n = 196) taken from herds T1 and 51. ROC analysis showed a higher AUC for p24 antibodies than whole-BLV antibodies (Zreactivity: 3.55, P < 0.001; Ztiter: 2.88, P < 0.01), and as consequence a better performance to predict the proviral load status in herd 51. No significant differences were found between the performance of p24 and whole-BLV antibodies in herd T1. A significant positive correlation was observed between PVL values and p24 antibody reactivity in both farms (r T1 = 0.7, P < 0.001, r 51 = 0.71, P < 0.0001). The analysis was extended to the whole number of weak p24 antibody reactors (n = 311) of the other 13 farms. The mean of high PVL reactors within weak p24 reactors was 17.38% (SD = 8.92). In 5/15 farms, the number of weak p24 reactors with high PVL was lower than 10%.


We found that the humoral response reflected the level of in vivo infection, and may therefore have useful epidemiological applications. Whereas the quantitative evaluation of blood proviral load using real-time PCR is expensive and technically demanding, the measurement of antibodies in blood by ELISA is relatively straightforward and could therefore constitute a cost-effective tool in a BLV control intervention strategy, especially in highly infected herds such as Argentinean dairy ones.

BLV; Proviral load; Control measures; Serological marker; ROC curve; p24 antibodies