Characterization of neuronal populations in the human trigeminal ganglion and their association with latent herpes simplex virus-1 infection

Abstract

Following primary infection Herpes simplex virus-1 (HSV-1) establishes lifelong latency in the neurons of human sensory ganglia. Upon reactivation HSV-1 can cause neurological diseases such as facial palsy, vestibular neuritis or encephalitis. Certain populations of sensory neurons have been shown to be more susceptible to latent infection in the animal model, but this has not been addressed in human tissue. In the present study, trigeminal ganglion (TG) neurons expressing six neuronal marker proteins were characterized, based on staining with antibodies against the GDNF family ligand receptor Ret, the high-affinity nerve growth factor receptor TrkA, neuronal nitric oxide synthase (nNOS), the antibody RT97 against 200 kDa neurofilament, calcitonin gene-related peptide and peripherin. The frequencies of marker-positive neurons and their average neuronal sizes were assessed, with TrkA-positive (61.82%) neurons being the most abundant, and Ret-positive (26.93%) the least prevalent. Neurons positive with the antibody RT97 (1253 µm(2)) were the largest, and those stained against peripherin (884 µm(2)) were the smallest. Dual immunofluorescence revealed at least a 4.5% overlap for every tested marker combination, with overlap for the combinations TrkA/Ret, TrkA/RT97 and Ret/nNOS lower, and the overlap between Ret/CGRP being higher than would be expected by chance. With respect to latent HSV-1 infection, latency associated transcripts (LAT) were detected using in situ hybridization (ISH) in neurons expressing each of the marker proteins. In contrast to the mouse model, co-localization with neuronal markers Ret or CGRP mirrored the magnitude of these neuron populations, whereas for the other four neuronal markers fewer marker-positive cells were also LAT-ISH+. Ret and CGRP are both known to label neurons related to pain signaling.

Figure 1. Labeling of a TG with a marker mix of all markers used in the current study (Ret, TrkA, nNOS, RT97, CGRP and peripherin).

The micrograph on the left (A) was taken at 400x magnification for a more detailed view, that on the right (B) at 100x for an overview. Marker proteins are labeled with DAB (brown) resulting in complete labeling of all detected neurons. The tissue was counterstained with haematoxylin. Scale bars represent 50 µm in all cases.

Figure 2. Median cross-sectional areas of cells labeled with different marker antibodies.

Central values represent the median values inµm², whilst the bars indicate the 25th–75th percentile ranges. Filled squares  =  marker positive neurons, filled circles  =  marker negative neurons from the same experiment. n represents the number of counted neurons. Asterisks represent significant differences (*  =  p<0.01, **  =  p<0.001, Mann-Whitney U-test).

Figure 3. Fluorescence co-labeling of marker combinations

. The markers are labeled with Alexa568 (red), Alexa488 (green) and DAPI (blue). Marker combinations are indicated above each pictogram. Dotted-line circles indicate examples of single labeled neurons, solid-line circles indicate examples of double labeled neurons. The letter “U” indicates examples of unlabeled neurons, some with lipofuscin. Scale bars represent 50 µm.

Figure 4. Detection of HSV-1 LAT in neurons expressing different maker proteins in the human TG.

The micrographs in the left column are taken at 100x magnification for an overview, those in the right column at 400x for a more detailed view. The name of the marker used is indicated in each row. Marker proteins are labeled with DAB (brown), LAT using NBT (purple), and the tissue was counterstained with haematoxylin. Open arrowheads indicate marker negative cells, closed arrowheads marker positive cells. LAT positive cells are circled. Scale bars represent 50 µm in all cases.