Pseudorabies virus mutants as transneuronal markers

The transneuronal labeling properties of three genetically engineered forms of the Bartha strain of pseudorabies virus (PRV) were studied in the ocular sympathetic pathway of rats. Bartha PRV mutants in which expression of the viral glycoprotein gI (homologous to gE of herpes simplex virus type 1, HSV-1) was restored (Bartha gI +) or which express a wildtype form of glycoprotein gIII (homologous to gC of HSV-1 and referred here as Bartha gIII Ka) were analyzed. In addition, a Bartha PRV mutant (Bartha β-gal) containing the lacZ gene encoding E. coli β-galactosidase inserted into the gX gene (homologous to gG of HSV-1) was also studied. These were compared to the parental strain —Bartha PRV. The pattern of transneuronal labeling in the intermediolateral cell column was studied 4 days after 5 μl of different concentrations of viral stocks were injected into the anterior chamber of the eye. The optimal infectious dose required to produce the maximal number of cases with specific transneuronal labeling of sympathetic preganglionic neurons was determined and these were as follows: Bartha PRV = 10 7.5 pfu/ml, Bartha β-galactosidase = 10 6.5 pfu/ml, Bartha gIII Ka = 10 5 pfu/ml, Bartha gI += 10 4 pfu/ml. An inverse relationship between specificity and infectivity rate was observed. Bartha β-gal produced the greatest number of cases with specific labeling (76%); Bartha gI + produced the lowest level (10%) and thus, this virus is not useful for transneuronal labeling studies. Barthha gIII Ka labeled more sympathetic preganglionic neurons (second-order neurons) than Bartha β-gal or Bartha PRV. Bartha gIII Ka and Bartha β-gal viruses labeled more interneurons (third-order) than the standard Bartha PRV. In summary, the data demonstrate that attenuated strains of PRV are more reliable for transneuronal labeling studies because they produce a higher percentage of experiments with specific labeling patterns than do PRV strains containing wildtype gI or gIII genes.