Nature 502, 100–104 (2013); doi:10.1038/nature12519; corrigendum Nature 514, 654 (2014)

Here we provide additional information about the specificity and validation of the simian immunodeficiency virus (SIV) DNA and RNA assays reported in this Letter. As described, we used quantitative PCR and quantitative PCR with reverse transcription (qPCR and qRT–PCR, respectively) assays targeting sequences in the SIV gag gene for quantification of SIV RNA and DNA. For measurements of plasma viraemia (SIV RNA), we used a standard qRT–PCR assay that has been extensively used in the field. For measurements of cell- and tissue-associated SIV DNA and RNA, we used nested qPCR and qRT–PCR assays. In our Letter, we report the evaluation of rhesus cytomegalovirus (RhCMV)-based vaccine vectors that include one vector with an SIV gag sequence insert. As shown in Fig. 1, sequence identity between the SIV gag insert in the RhCMV/Gag vector and the SIV gag-targeted primer and probe sequences for the standard qRT–PCR assay used for plasma SIV RNA measurements (sGAG21, sGAG22 and sGAG23 probes) raises the possibility that transcripts from the vaccine insert could have been detected by the assay we used.

Figure 1: Sequence of SIV gag PCR/RT–PCR assay primers/probe and vaccine insert.
figure 1

The nucleic acid sequence for the first 640 base pairs of the native SIVmac239 gag gene (‘SMM239’) is shown aligned with the sequence of the SIV gag insert used in the RhCMV/Gag vector (‘GAG.FLAG’). The gag-targeted primers and probe are indicated in colour on the strands that were used in qPCR- and qRT–PCR-based viral nucleic acid quantification assays. Dashes indicate sequence identity and only base differences from the SIVmac239 gag gene are shown in the sense strand of the RhCMV/Gag vector. The primers and probe are listed as follows. The ‘dP’ and ‘dK’ bases1 were incorporated to allow non-biased coverage of non-SIVmac239 isolates with divergent sequences at the relevant positions in other studies using different viruses. Forward primer, sGAG21: 5′-GTCTGCGTCAT(dP)TGGTGCATTC-3′; reverse primer, sGAG22: 5′-CACTAG(dK)TGTCTCTGCACTAT(dP)TGTTTTG-3′. Probe, psGAG23: 5′-FAM-CTTC(dP)TCAGT(dK)TGTTTCACTTTCTCTTCTGCG-BHQ1-3′. (BHQ1, black hole quencher-1; FAM, 6-carboxyfluorescein.) Nested primers: SIVnestF01: 5′-GATTTGGATTAGCAGAAAGCCTGTTG-3′, SIVnestR01: 5′-GTTGGTCTACTTGTTTTTGGCATAGTTTC-3′.

PowerPoint slide

Full size image

For the nested assay, designed mismatching of the outer reverse primer (SIVnestR01) and the codon-optimized portion of the vaccine gag insert sequence and specific priming of the reverse transcription step of the qRT–PCR assay is intended to prevent potential cross reactivity. Biological features of the RhCMV vectors used are expected to minimize the prospects of detection of vaccine insert-derived SIV gag RNA in plasma or insert-derived SIV gag DNA or RNA sequences in cell or tissue samples.

To test empirically for possible detection of vaccine insert-derived SIV gag RNA sequences in plasma, we evaluated 394 plasma samples from RhCMV/Gag-vaccinated, SIV-unexposed rhesus macaques using our standard (non-nested) SIV RNA qRT–PCR assay. All 394 samples were negative. These results support the interpretation that under the experimental conditions reported, the standard assay does not detect vaccine insert-derived sequences in the plasma of vaccinated animals, and that SIV RNA detected after SIV challenge in such animals is derived from SIV infection.

To evaluate empirically the potential detection of vaccine insert sequences in tissues by the nested qPCR and qRT–PCR assays that we used, we surveyed a total of 291 tissue samples obtained at necropsy from five rhesus macaques that were vaccinated with a regimen that included RhCMV/Gag vectors, but were not challenged with SIV, including three animals described in our Letter, and two additional animals not previously reported. Out of 2,550 total nested PCR or RT–PCR reactions, we detected 4 (0.16%; 95% confidence interval (CI) 0.06–0.40%) and 42 (1.65%; 95% CI 1.22–2.22%) reactions positive for SIV DNA and RNA, respectively. We compared these frequencies to results obtained from the five rhesus macaques described in our Letter that were vaccinated with RhCMV/Gag vectors and challenged with SIV, that then showed evidence of SIV infection, and were necropsied shortly after demonstrating control of plasma viraemia (<5 SIV gag RNA copies per millilitre at necropsy). The frequencies of positive reactions for the vaccinated but non-SIV challenged animals were significantly lower than the results observed for tissue specimens from the vaccinated and SIV-challenged animals necropsied shortly after demonstrating control of SIV infection (early protected monkeys), which showed 59 (2.65%; 95% CI 2.06–3.40%) and 531 (23.81%; 95% CI 22.09–25.62%) positive DNA and RNA reactions, respectively, out of 2,230 total nested PCR and RT–PCR reactions (P < 2.2 × 10−16 for DNA, P = 3.7 × 10−16 for RNA, by Fisher’s exact test). These results support the interpretation that under the experimental conditions reported, the exquisitely sensitive nested qPCR and qRT–PCR assays we used detect only very low frequencies of reactions positive for SIV DNA and RNA in tissues from animals vaccinated with SIV gag insert-containing vaccines but not challenged with SIV, probably reflecting background false positive reactions, and that the significantly higher levels of SIV DNA and RNA detected after SIV challenge of vaccinated animals are derived from SIV infection. This interpretation is corroborated by other evidence of an initial SIV infection in the challenged, vaccine-protected monkeys including T cell responses to SIV antigens not included in the vaccine (for example, SIV Vif), and rescue of replication-competent virus by culture.