Figure 1.
Infected DRG fulfill classical definitions of HSV latency.
(A) Plaque forming units assay of HSV-infected DRG (T5 to L1) at various time points post inoculation. (B) Copy number of viral DNA from HSV-infected DRG (day 50) was determined by quantitative PCR. (C) Relative expression of HSV 2 kb LAT and lytic genes in HSV-infected DRG (days 5 and 50) were determined by quantitative PCR. (D) Expression (Et) of HSV 2 kb LAT and lytic genes in HSV-infected DRG (> day 240) were determined by quantitative PCR. Total RNA from each sample was aliquoted into four tubes and RT and pre-amplification were done as indicated beneath each graph. Pre-amplified products were used without dilution. Data in (A–C) are pooled from two independent experiments with 4–5 mice per group, (D) from one experiment with 5 mice, and plotted as mean ± S.E.M (A–C) or showing each individual result (D).
Figure 2.
The number of infected cells remains stable during latency.
(A) Number of infected (β-galactosidase+) cells from KOS/pCMV/eGC infected DRG at the times shown post inoculation. (B) DRG explants from ROSA-YFP mice (day 20) infected with KOS0152 or WT HSV, or uninfected mice were examined under the 2 photon intra-vital microscope. Arrows identify HSV-infected cells. (C) DRG sections from KOS0152- or WT HSV-infected ROSA-YFP mice (day 50) were examined under the laser capture microscope for YFP expression and Hoechst nuclear stain. Data in (A) are pooled from two independent experiments and represented as mean ± S.E.M. Images in (B and C) are from two independent experiments.
Figure 3.
Latent HSV resides in a subpopulation of sensory neurons.
(A) YFP+ and YFP− cells were identified and captured from DRG sections of infected ROSA-YFP mice. (B) DRG sections from KOS0152- or WT HSV-infected ROSA-YFP mice (day 20) were stained for Nissl bodies and examined under the confocal microscope. Arrows identify HSV-infected neurons. (C) Copy number of viral DNA in individual YFP+ cells (day 15) captured by LCM is shown in a histogram. (D) Number and percent of Ntrk1+ and Ntrk1− neurons from populations of infected (YFP+) and uninfected (YFP− and uninfected DRG) neurons in the main set of single cell expression data are shown in pie charts. Data in (A and C) are pooled from two independent experiments. Images in (B) are from one experiment. (D) As for all single cell gene expression data, the neuron sections analysed were combined from 4 independent infections (YFP+ and YFP− neurons) or 3 independent experiments (neurons from uninfected DRG). Numbers in brackets show the number of individual cells analyzed.
Table 1.
Number and percentage of YFP+ and YFP− cells containing HSV DNA.
Table 2.
Number and proportion of YFP+ cells containing various copy number of HSV DNA.
Figure 4.
HSV gene expression in individual neurons during latency.
(A) Heatmap showing HSV genes determined using quantitative RT-PCR in single YFP+ neurons. (B) Number and percent of all infected YFP+ neurons containing 2 kb LAT. (C) Number and percent of all infected YFP+ neurons with different HSV lytic gene expression profiles. (D and E) As per B and C, but restricted to Ntrk1+YFP+ neurons. Numbers in brackets show the number of individual cells analyzed. All data are from the main single cell gene expression dataset.
Figure 5.
Single cell gene expression profiling reveals the transcriptional response of infected neurons towards latent HSV.
(A) Principal components (PC) analysis of 48 cellular genes in single Ntrk1+neurons – uninfected, YFP− and YFP+. (B) PC analysis on Ntrk1+YFP+ neurons – YFP− versus uninfected. All data are from the main single cell gene expression dataset.
Figure 6.
Increasing viral activity is matched by progressive host neuronal transcriptional response.
(A) Violin plot representation of selected cellular gene expressions in LAT+Ntrk1+YFP+ neurons categorized based on their lytic gene expression profile. Log2Ex represents expression threshold (Et). Numbers in brackets show the number of individual cells analyzed. * both proportion (p<0.05) and expression levels (p<0.0167) are significant, ‡ only expression levels significant, # proportion and expression levels are not significant when comparing full-lytic to partial- and non-lytic subsets. (B) k-means clustering of Kendall tau rank correlation coefficients (τ) of every pair of 48 genes from single LAT+Ntrk1+YFP+ neurons. Correlation coefficient matrix of non-lytic neurons was clustered according to the optimal clustering observed in full-lytic neurons. (C) Complete-linkage clustering of Kendall tau rank correlation coefficients (τ) between expression profiles of every pair of 48 genes from single LAT+Ntrk1+YFP+ neurons. Correlation coefficient matrices of non-lytic and full-lytic neurons were independently clustered. All data are from the main single cell gene expression dataset.
Table 3.
k-means clustering of gene expression correlation coefficients.