De Novo Herpes Simplex Virus VP16 Expression Gates a Dynamic Programmatic Transition and Sets the Latent/Lytic Balance during Acute Infection in Trigeminal Ganglia

Abstract

The life long relationship between herpes simplex virus and its host hinges on the ability of the virus to aggressively replicate in epithelial cells at the site of infection and transport into the nervous system through axons innervating the infection site. Interaction between the virus and the sensory neuron represents a pivot point where largely unknown mechanisms lead to a latent or a lytic infection in the neuron. Regulation at this pivot point is critical for balancing two objectives, efficient widespread seeding of the nervous system and host survival. By combining genetic and in vivo in approaches, our studies reveal that the balance between latent and lytic programs is a process occurring early in the trigeminal ganglion. Unexpectedly, activation of the latent program precedes entry into the lytic program by 12 -14hrs. Importantly, at the individual neuronal level, the lytic program begins as a transition out of this acute stage latent program and this escape from the default latent program is regulated by de novo VP16 expression. Our findings support a model in which regulated de novo VP16 expression in the neuron mediates entry into the lytic cycle during the earliest stages of virus infection in vivo. These findings support the hypothesis that the loose association of VP16 with the viral tegument combined with sensory axon length and transport mechanisms serve to limit arrival of virion associated VP16 into neuronal nuclei favoring latency. Further, our findings point to specialized features of the VP16 promoter that control the de novo expression of VP16 in neurons and this regulation is a key component in setting the balance between lytic and latent infections in the nervous system.

Fig 1. Mutants generated for these studies.

Schematic representations of the genomic structures of all new viral mutants generated for these studies. Open reading frames and promoter sequences are indicated by filled arrows (Black = genes resident in the locus, Red = transgene). Selected restriction endonuclease sites and base pair lengths are indicated above and below each diagram. All transgene expression cassettes were inserted after base pair 138047 in the HSV-1 strain 17syn+ genome. Insertion at this site does not perturb expression of any viral genes [11,54].

Fig 2. In vivo phenotypes of mutants expressing extra copies of ICP0 and VP16.

A) Viral replication in trigeminal ganglia (TG). Outbred Swiss Webster mice were infected via both corneas with 1x10⁵ pfu of the indicated virus isolates. Tissues from three mice per viral mutant per time point were harvested and analyzed for infectious virus titers. Bars represent the fold increase in titer relative to the parental strain 17syn+. Viral titers appear in the text. B) Survival was determined in groups of mice infected on the eyes with 1x10⁵ pfu. The color code is the same as that in Fig 2A with the addition of a gray colored line which represents mice infected with the wild type strain McKrae (a virulent laboratory strain). C) Replication and spread of virus within the central nervous system was assessed by examining virus titers in four roughly equal divisions as depicted below the graphs. Viral titers detected on days 5, 7 and 9 pi in the CNS of three mice from each group are shown. 17LATpVP16R is a genomically restored isolate in which the VP16 transgene in the 17LATpVP16 mutant was removed (detailed in methods). Each bar represents the average of the amount of virus found in each of the brain sections. There was no significant difference between 17syn+ and 17LATpVP16R infected brain regions (bar labeled NS for not significant). There was significantly more virus detected in all four brain sections infected with 17LATpVP16 compared to 17syn+ and 17LATpVP16R (bar labeled ** is p<0.05; labeled *** is p<0.001). Both Fisher’s exact test and unpaired Student’s t-test were employed.

Fig 3. Quantification of the number of neurons expressing the lytic and/or the latent transcriptional program markers.

A) Quantification of neurons positive for viral protein at 40 hrs pi. This experiment included 17LATpVP16R, the genomic rescue of 17LATpVP16, as well as 17LATpICP0 and two mutants made on the LAT null background parent 17AH (17AHLATpLacZ and 17AHLATpICP0). There was no significant difference in the number of neurons positive for viral proteins [11] in whole TG except for the TG infected with 17LATpVP16, which contained significantly more positive neurons (p<0.0004, Fisher’s exact test). B) Quantification of neurons positive for viral protein, LATp activity, or both at 46 hrs pi. Each point represents the number of positive neurons in a TG. C) TG were fixed and processed for the simultaneous detection of b-gal activity (blue neurons:red arrows) and viral protein expression (brown neurons:black arrows). Shown are representative photomicrographs of whole ganglia at the same magnification. The number of neurons in TG expressing either viral proteins (lytic), LacZ from the LAT promoter (“latent”), or both (lytic+“latent”) were enumerated (shown in B).

Fig 4. Quantification of viral lytic and acute stage latency gene transcription in individual neurons at very early times pi.

A. Relative promoter strength was assayed by transfection of rabbit skin cells with the constructs employed to generate the viral promoter/reporter mutants and analysis of the amount of b-gal determined with a CPRG assay as detailed in methods. Transfection efficiencies were normalized by including a renilla luciferase plasmid in the assay. Lane 1 = LATpLacZ, 2 = VP16pLacZ, 3 = ICP0pLacZ. The LAT promoter was the weakest and the levels of expression of this construct were set to one. Where indicated, a VP16 expressing plasmid was included in the transfections. Each bar represents the average of 3 transfection experiments of 3 wells each. B. Photomicrographs of TG from mice infected with17LATpLacZ processed for LATp activity and viral protein expression. At 22 hrs pi, expression was restricted to the LATp. A 36 hrs, viral proteins are expressed but restricted to a subset of neurons that are marked by LATp activity. C. At the indicated times pi, eyes and TG were harvested. Unsectioned TG were processed for the histochemical detection of b-gal activity and IHC detection of viral proteins as detailed previously [4,11,48,54,78,81]. Bars represent the number of neurons positive in an individual TG. The number of TG positive over the number tested in each group is indicated. All TG tested from mice infected with 17LATpLacZ contained one or more neurons positive for b-gal activity at 22 hrs pi, whereas no TG infected with the lytic stage promoter reporter viruses were positive (p≤0.0001, ANOVA). At 36 hrs pi a subset of ganglia examined contained neurons positive for both b-gal and viral protein. The number of ganglia positive and the number of positive neurons in the TG was not different between groups (p≥0.5, ANOVA). Importantly, all neurons positive for viral protein were also positive for LAT promoter activity in the 17LATpLacZ group.

Fig 5. Entry into lytic infection in TG neurons is regulated by sequences in the VP16 promoter.

A) Sequence of the proximal VP16 promoter. Putative regulatory sites identified previously [83,84] are color coded and labeled above the sequence. Potential sites known to confer reciprocal regulation to neuronal genes are underlined and in red. The boxed nucleotides were altered as detailed in methods. B) Replication kinetics on eyes and in TG on days 2 through 10 pi is shown as the area under the curve (AUC). Tissues from 3 mice were examined for each time point. 1 = 17syn+, 2a, b, c are three independently derived isolates of the VP16 promoter mutant 17VP16pπRR and 3 = the genomically restored isolate 17VP16pπRR-R. C) Mice were infected on the cornea with the VP16 promoter mutant 17VP16pπRR or its genomically restored isolate 17VP16pπRR-R. At 72 hrs pi tissues were harvested and processed for the whole tissue immunohistochemical detection of VP16 and the number of VP16 positive neurons enumerated. The data are shown as a scattergram with each point representing the number of positive neurons in an individual TG. The difference between the groups was significant (p<0.0001, Student’s t-test). D) Representative photomicrographs of corneas stained for VP16 protein which is seen as a brown precipitate in characteristic lesions on the corneal surface (arrows). E) Representative photomicrographs of sectioned trigeminal ganglia infected with 17VP16pπRR-R stained immunohistochemically for VP16 (purple precipitate). Numerous cells including neurons identifiable by their large size, morphology and axonal tracts are positive in the TG infected with 17VP16pπRR-R the genomically restored isolate which is not different than WT 17syn+ (white arrows). F) Rare VP16 positive cells in the 17VP16pπRR infected TG were detected and did not appear to be neurons. One such area is boxed and shown enlarged in an inset micrograph. This region appears to be a small focus of positive support cells.

Fig 6. The VP16 promoter contains a region required for efficient exit from the default latent state.

Mice were infected on scarified corneas with 2x10⁵ pfu of (A) 17LATpLacZ or (B) 17VP16πRR+LATpLacZ. At 44 hrs pi TG were harvested and processed for whole ganglion histochemical detection of b-gal activity (blue) followed by immunohistochemical detection of viral proteins (brown) as detailed in methods. Each bar represents the number of positive neurons in a TG. Blue vertical bars are neurons positive for only b-gal, brown bars are those positive for only viral proteins, and blue and brown checkered bars are neurons positive for both. There were significantly more neurons containing viral proteins in 17LATpLacZ infected group, about 33% of the total labeled neurons vs. 2% of the total in the 17VP16πRR+LATpLacZ infected group (p<0.0001, Student’s t-test).

Fig 7. Regulation of de novo VP16 expression in neurons controls the exit from the default acute stage latent transcriptional program.

Rectangular panels showing viral program usage in TG neurons at 22, 36, and 44 hours pi. Top panel: Wild type virus (17LATpLacZ). Progression from only latent transcriptional program usage at 22 hrs pi (blue neurons only) to ~33% of the accumulated blue neurons transitioning into the lytic program at 44 hrs pi is observed. These viruses display wild type replication and virulence properties. Left panel: Disruption of the pre-immediate early regulatory region of the VP16 promoter, mutant 17VP16pπRR+LATpLacZ. Similar to WT numbers of neurons positive for the latent transcription program are seen at 22 hrs pi but the number of neurons transitioning into the lytic viral transcription program is reduced to 2%. These mutants replicate normally on eyes, but as would be expected from the reduced transition into the lytic cycle, replicate poorly in the TG and virulence was not observed. Right panel: Expression of VP16 from the LAT promoter, mutant 17LATpVP16. Expression of VP16 directly upon engagement of the latent transcription program results in entry into the lytic cycle in neurons at 22 hrs pi. Neurons continue to enter the lytic transcription program and viral replication in the nervous system and virulence are increased.