Sensory ganglia, including the trigeminal ganglion, are collections of thousands of cells including a heterogeneous group of primary sensory neurons. When sensory ganglia are infected with HSV-1 or HSV-2, infected neurons follow either of two pathways of viral gene expression. In some neurons, a productive cycle of viral gene expression occurs while in other neurons, the virus establishes a latent infection 
. Not all of the neurons in sensory ganglia are equally susceptible to productive infection and specific types of neurons are more likely than others to harbor latent HSV-1 and HSV-2 
. In previous studies, we have shown that HSV-1 and HSV-2 preferentially establish latent infection and express LAT in different subtypes of sensory neurons after ocular infection of mice 
. Approximately 50% of the latent HSV-1 LAT+ sites are in A5+ neurons, while about 50% of the latent HSV-2 LAT+ sites are in KH10+ neurons. This preferential accumulation of LAT is not due to preferential LAT promoter activity in different types of neurons (11). Through the use of viral chimerae, we mapped the function(s) responsible for differential establishment of latency of HSV in A5+ and KH10+ neurons to the LAT region 
. In the present studies, we have presented evidence that mechanisms responsible for LAT-mediated differential establishment of latency in vivo
are likely cis
-acting functions, since co-infection of mice with HSV-1 and HSV-2 did not alter preferential establishment of latent infection. We have further demonstrated that the mechanisms that regulate differential permissiveness for productive infection in vitro
are also likely to be cis
-acting in nature, since HSV-1 and HSV-2 do not appear to produce factors that act independently in trans
to regulate preferential permissiveness for productive infection in A5+ and KH10+ adult trigeminal neurons in vitro
. While there are neuronal populations that support productive infection of both HSV-1 and HSV-2 (25.2% of neurons in culture were productively infected with both HSV-1 and HSV-2), there are also populations in which each virus cannot efficiently express lytic cycle genes. A5+ neurons do not support productive infection of HSV-1 and infection of these neurons results in latent infection representing half of the latent reservoir in vivo
. Similarly, KH10+ neurons do not support productive infection with HSV-2, which leads to latent infection in this population 
, also representing half of the latent HSV-2 reservoir in vivo
. Thus, approximately half of the latent HSV reservoir appears to be maintained in neurons incapable of supporting productive infection. Populations of neurons other than A5+ or KH10+ neurons make up about 75% of all of the neurons in the TG, and as a group these populations appear to be capable of maintaining latent HSV and also supporting productive infection of the virus, and thus may serve as the reservoir for reactivation-competent HSV.
The role of the HSV LAT region in the establishment and maintenance of latent infection has been extensively investigated, largely through the use of deletion viruses. The mechanisms by which this region accomplishes these functions are less well understood, but include effects on IE gene expression, heterochromatin formation, and apoptosis 
. A number of RNAs are expressed from this region, including microRNAs, the stable LAT intron, and RNAs antisense to LAT 
. Our data suggest that the RNAs that are transcribed from the 2.8 kb of the LAT region investigated in our studies do not play a direct trans
-acting role in the establishment of latent infection of the heterologous HSV serotype in distinct neuronal populations. However, it is possible that trans
-acting factors expressed from this region are not functional in the heterologous serotype or that specific trans
-acting factors may not be expressed in all neuronal populations. It is also possible that previously identified transcripts expressed from this LAT region are important for functions other than neuronal specificity.
A theory on how the LAT region of the HSV genome might act in a cis–regulated manner to transcriptionally regulate key lytic gene promoters has been previously proposed by one of the authors (3). To summarize, Bloom and colleagues proposed that the B1 and B2 CTCF insulators in the 5′ LAT region might interact, forming a loop domain containing the LAT enhancer. Formation of this DNA loop would enable the LAT enhancer to interact with local promoters that regulate key immediate early gene expression, such as those involved in the transcription of ICP4, ICP0, and ICP27. Furthermore, the transcriptional outcome of any given LAT enhancer/IE promoter interaction (positive or negative) would depend on a given cell’s transcriptional milieu, which is clearly different amongst different types of TG sensory neurons.
The studies described herein provide an important piece of the puzzle related to HSV pathogenesis; the HSV-1 and HSV-2 2.8 kb LAT regions investigated in these studies do not produce trans
-acting factors that regulate preferential viral behavior in different types of neurons. However, co-infection with HSV-1 and HSV-2 in the murine ocular infection model presents several challenges that limit the interpretations of the results from these studies. Ocular infection of mice with HSV-1 strain 17syn+ or any HSV-2 strain requires post-infection treatment with acyclovir to prevent undue mortality rates, although we determined in previous studies that acyclovir treatment starting at 40 hpi does not significantly alter the patterns of preferential establishment of latent infection. In the current studies, co-infection with any strain of HSV-1 and any strain of HSV-2 increased the probability of mortality despite treatment with acyclovir, thus limiting the range of viral inocula we were able to investigate, particularly after co-infection with HSV-1 17syn+ and HSV-2 333. The high mortality rate of mice co-infected with HSV-1 and HSV-2 suggests that these viruses likely interact in some manner, but an interaction was not evident in ganglionic A5+ and KH10+ neurons, in which half of the latent HSV-1 and HSV-2 reservoirs are detected, respectively 
. HSV-1 and HSV-2 also demonstrate different kinetics during acute infection, both in vivo
and in vitro
, which may have permitted one virus to out-compete the other in specific types of neurons. Although we tested staggered infection in vitro
with no differences in outcome (unpublished observations), our in vitro
infections were carried out in the absence of an adaptive immune response and in the absence of the normal three-dimensional structure of the trigeminal ganglion.
Taken together, our results to date show that 1) certain populations of trigeminal ganglion neurons are more likely to support productive infection in an HSV-serotype-specific manner, which results in the establishment of HSV latent infection in a serotype-specific manner 
; 2) there is a viral function that regulates this set of phenotypes and it appears to map to the LAT region of the viral genome 
; and 3) this function is not mediated in trans
by factors produced by the 2.8 kb region of the HSV-1 and HSV-2 LAT regions investigated in these studies. Although exchanging this 2.8 kb region of the LAT between HSV-1 and HSV-2 also exchanges viral preferences for productive or latent infection in specific types of trigeminal neurons, the mechanism by which this occurs is not trans