Acta Neuropathol (1992) 84: 39- 44

Am Nwopatholosica @ Springer-Verlag 1992

Latent herpes simplex virus type 1 in human geniculate ganglia* Y. Furuta ~,2, T. Takasu 1,2, K. C. Sato ~, S. Fukuda z, Y. Inuyama 2, and K. Nagashima 1 Departments of 1Pathology, and 2Otolaryngology, Hokkaido University School of Medicine, Kita 15, Nishi 7, Kita-ku, Sapporo 060, Japan Received July 30, 1991/AcceptedJanuary 20, 1992

Summary, Viral infection, especially by reactivation of herpes simplex virus (HSV) has been considered to be a possible explanation for the pathogenesis of idiopathic peripheral facial nerve palsy (Bell's palsy). We investigated whether the geniculate ganglia of man contain latent HSV type 1 (HSV-1), and compared the frequency of HSV-infected ganglia and that of latently infected neurons in human geniculate ganglia and in trigeminal ganglia. From autopsy specimens of eight adults 15 geniculate ganglia and 16 trigeminal ganglia were examined by means of in situ hybridization and immunohistochemical staining. The HSV-1 genome was detected in 11 of the 15 ( 7 1 % ) geniculate ganglia and in 13 of the 16 ( 8 1 % ) trigeminal ganglia. No HSV antigen was n o t e d in any of the ganglia. The incidence of latently infected neurons was 0.9 % in the trigeminal ganglia and 5.3 % in the geniculate ganglia. The difference in percentages between the two types of ganglia was significant. Our results suggest that reactivation of latent HSV in the geniculate ganglia is a probable cause of some cases of herpetic stomatitis and of idiopathic peripheral facial nerve palsy. Key words: Herpes simplex virus - Latent infection H u m a n geniculate ganglion - In situ hybridization Bell's palsy

The humart geniculate ganglion is located in the geniculate portion of the facial nerve and is surrounded by the bony wall of the fallopian canal. Several types of sensory fibers which supply the taste buds of the anterior two thirds of the tongue and palate, as well as some somatic

* Supported by Special Grant-in-Aid for Promotion of Education and Science in Hokkaido University (to K.N.) and Grant-in-Aid for Scientific Research (C) (03670803 to S.E) provided by the Ministry of Education, Science and Culture Correspondence to: Y. Furuta (address see above)

fibers from the skin around the concha, have their cell bodies in this ganglion. Latent herpes simplex virus (HSV) has been isolated from human trigeminal ganglia of unselected cadavers by a cocultivation m e t h o d [2, 3]. However, it is difficult to analyze the geniculate ganglion in man because it is located in the temporal bone. There has been no report thus far that demonstrates the latency of HSV in human geniculate ganglia. The cause of idiopathic peripheral facial palsy (Bell's palsy) remains unknown, but some reports [1, 13] have suggested that reactivation of latent HSV in the sensory neurons of the geniculate ganglia could lead to damage of the facial nerve such as that seen in Bell's palsy. In the present study, we have demonstrated latent HSV type 1 (HSV-1) in human geniculate ganglia for the first time using in situ hybridization. In addition, we have compared the frequency of HSV-infected ganglia and that of latently infected neurons in human geniculate ganglia and in trigeminal ganglia.

Methods Tissue specimens

Geniculate ganglia were obtained during autopsies of eight adults 2-13 h after death. At autopsy, the temporal bones were dissected from the base of the cranium, and the facial nerves were exposed from the internal auditory canal to the horizontal portion with the use of an electric drill and a microscope, as illustrated in Fig. 1. The left and right facial nerves, including the geniculate ganglia, were removed from the fallopian canals, immersed in 10 % buffered formalin or 4 % paraformaldehyde, and embedded in paraffin. The trigeminal ganglia were also removed from the subjects. As a positive control, a brain tissue specimen was obtained from a patient with HSV-induced encephalitis. A ganglion from a newborn baby was used as a negative control. HSV-1 D N A probes

We used a 3.4-kilobase (kb) BarnHI-Q restriction fragment and a 2.6-kb SalI/BamHI DNA fragment derived from the BamHI-B

40

Fig. la-c. Preparation of human geniculate ganglia, a The facial nerve in the fallopian canal is exposed from the meatal segment (M) to the horizontal segment (H) with an electric drill. The geniculate portion (G) and the large petrosal nerve (L) are seen. h A removed facial nerve. Arrowhead indicates geniculate ganglion, c Hematoxylin-eosinstained section of the facial nerve. The geniculate ganglion (G) is visible in this slice, a x 10; bx5;cx60 restriction fragment of the HSV-1 strain [15]. The former fragment contains the thymidine kinase (TK) gene, and the latter contains the infected-cell protein number zero (ICP0) gene and the latency-associated transcript (LAT) gene which partially overlaps the 3' end of the ICP0 gene and is transcribed in the opposite direction [19, 20], as shown in Fig. 2. These DNA fragments were radiolabeled with [35S]dCTP (Amersham) by the random primed method [10] to specific activities of 0.7 • 109 - 1 • 109 cpm/~ug. Each probe was purified on a G-50 Sephadex column.

HSV-1 R N A probes The 2.6-kb SalI/BamHI fragment was cloned into pSPT 18. With the RNA labeling kit (Boehringer Mannheim), strand-specific HSV-I RNA probes (Fig. 2) were synthesized by in vitro transcription with the SP6 promoter (RNA probe A) and T7 promoter (RNA probe B), according to [17]. The probes were labeled with digoxigenin-uridine-triphosphateas substrate. The template DNA was digested with RNase-free DNase I (Boehringer Mannheim), and the RNA probes were then subjected to limited alkaline hydrolysis and reduced to an average size of 200 nucleotides [5].

In situ hybridization Five-micrometer-thick sections were cut from paraffin-embedded specimens and mounted on 3-aminopropyl-triethoxysilane-coated

slides [16].We made certain that the serially resected, hematoxylineosin-stained specimens contained geniculate ganglia. In situ hybridization was performed as described previously [11]. Briefly, the tissues were deparaffinized and rehydrated by sequential immersion in xylene and graded ethanols. After pretreatment with 0.2 M HC1 and 100 ~tg/ml proteinase K (Boehringer Mannheim), the sections were refixed in 4 % paraformaldehyde and immersed in 0.2 % glycine, followed by washing in 0.1 M phosphate-buffered saline. When DNA probes were used, hybridization was carried out in 50 % deionized formamide, 5 x SSPE (t x SSPE: 0.18 M NaC1, 10 mM NaH2PO4, 1 mM EDTA, pH 7.4), 5 x Denhardt's solution, 0.1% sodium dodecyl sulfate (SDS), 100 gg/ml denatured salmon sperm DNA, 10 % dextran sulfate, and 0.2 ng/~tl 35S-labeled probe at 42 ~ for 18 h, and the sections were washed in 50 % formamide, 2 x SSC (1 x SSC: 0.15 M NaCI, 15 mM sodium citrate) at 55 ~ The sections were then dipped in Konica NR-M2 emulsion for autoradiography. After 7 days of exposure, the slides were developed and counterstained with hematoxylin. For RNA probes, hybridization was done in 50 % deionized formamide, 600 mM NaC1, 10 mM Tris (pH 7.6), 1raM EDTA, 10 mM dithiothreitol, 1 • Denhardt's solution, 0.25% SDS, 200 ~g/ml Escherichia coli tRNA, and 5 ng/~tl of digoxigeninlabeled probe at 50 ~ for 18 h. After washing under high stringent conditions (50 % formamide, 2 x SSC at 60~ hybridization signals were detected by enzyme-linked immunoreaction with an anti-digoxigenin alkaline phosphatase conjugate, as described previously [11].

41 M a p Units

Genome Structure

Barn HI Fragments

o

0.2

r,

I

TR I

L i

II

I

0.4 I

0.6

I

0.8

[

I

I

UL

I

Q 3.4Kb I Illllll IIIII

,I

IR L /R s U s TR s I

II

1.0

I

I

I

,

Ilill

I

I

Barn HI ~ I/ I Sal I

I1[I 2,6

Sal I ,

4

Patients

EZ]

B [

Results

I

Illll am H~ RNA probes

:>A

.... ICPO

T h e profile of the patients examined is presented in Table 1. Autopsies were p e r f o r m e d between N o v e m b e r 1988 and M a y 1989 in the D e p a r t m e n t of Pathology II, H o k k a i d o University School of Medicine. N o n e of the patients had a history of facial nerve palsy, and none had s y m p t o m s of H S V infection at the time of autopsy.

In situ hybridization with HSV-1 D N A probes

LAT

Fig. 2. Genomic structure of herpes simplex virus type 1 (HSV-1) and location of probes used in this study.The HSV-1 DNA consists of two unique regions (UL and Us) which are bounded by inverted-repeat sequences ( TRL to IRL and TRs to IR s). The 3.4-kb BamHI-Q restriction fragment contains the region which encodes the thymidine kinase (TK) gene. The BamHI-B restriction fragment contains the immediate early genes. The 2.6-kb SalI/BamHI DNA fragment made by additional restriction enzyme cuts encodes the infected-cell protein number zero (ICPO) gene and a part of the latency-associated transcript (LAT) gene. The approximate locations of the TK, ICPO, and LAT genes and the direction of transcription of their mRNAs are shown. RNA probe A (ICP0 antisense probe) is complementary to the ICP0 mRNA. RNA probe B (LATantisense probe) has the opposite sense to that of the LAT

Immunhistochemistry We used an antibody to HSV-1 which was purchased from Dako. Immunohistochemical staining was carried out by the avidinbiotin-peroxidase complex method on the serially sectioned specimens which were used for the in situ hybridization study.

In the positive control specimen of H S V encephalitis, positive hybridizations with the HSV-1 T K and ICP0L A T gene probes were detected in the neurons. No hybridization was seen in the trigeminal and geniculate ganglia f r o m the n e w b o r n case. In case 5, the 1eft geniculate ganglion was not available. Thus, we analyzed 15 geniculate and 16 trigeminal ganglia. In the hematoxylin-eosin-stained sections of the geniculate ganglia, we found 8 to 104 ganglion cells per section, and in the trigeminal ganglia, 73 to 1,400 ganglion cells per section. O n in situ hybridization with the HSV-1 T K probe, no definite positivity was detected in the geniculate or trigeminal ganglia. T h e in situ hybridization study with the HSV-1 I C P 0 - L A T p r o b e d e m o n s t r a t e d HSV-1 nucleic acids not only in the trigeminal ganglia, but also in the geniculate ganglia (Figs. 3, 4). Positive hybridization signals were detected mainly in the nuclei of neurons. No H S V was detected in the satellite cells or other non-neuronal cells in the ganglia.

Fig. 3. In case 2, positive hybridizations were seen in the right trigeminal ganglion (a) and the right geniculate ganglion (b) by in situ hybridization with the HSV-1 ICP0-LAT DNA probe, a,b • 100

42

as shown in Table 1. In case 6, no HSV was detected in the geniculate or trigeminal ganglia. No significant difference in the frequency of HSV-infected ganglia was observed between the two types of ganglia (P > 0.01. /42 test). The percentages of HSV-positive cells ranged from 0.1% to 4.8 % in the trigeminal ganglia and from 2.4 % to 13 % in the geniculate ganglia, as shown in Table 1. The total number of trigeminal ganglion cells analyzed was 9,049, and in these a total of 84 neurons were positive for HSV-1. Thus, the incidence of latently HSV-infected neurons was 0.9% in the trigeminal ganglia, and that in the geniculate ganglia was 5.3 % (37/704). The difference in percentages between the two types of ganglia was significant (P