therefore seems reasonable to infer that this was the way in which demyelination had occurred earlier throughout the endoneurium. Almost certainly, profound demyelination of motor axons was the basis for the patient's paralysis. If the changes in the biopsied nerve are representative of those occurring in other affected nerves in this patient, it is unlikely that axonal degeneration (i.e., an acute axonal neuropathy) played an important part in the production of symptoms. This investigation was supported by the Special Trustees of Guy's Hospital. We thank D r J. Payan for the neurophysiological results, and D r D. Bihari and the staff of the intensive care unit who helped to look after the patient.

References I . Asbury AK, Arnason BG, Adams RD. The inflammatory lesion in idiopathic polyneuritis. Its role in pathogenesis. Medicine 1969;48:17j-215 2. Honavar M, Tharakan JKJ, Hughes RAC, et al. A clinicoparhologicd study of Guillain-Barre syndrome: nine cases and literature review. Brain 1991;114:1245-1270 3. Asbury AK, Arnason BGW, Karp HR, McFarlin DG. Criteria for diagnosis of Guillain-Barre syndrome. Ann Neurol 1978i3.565-566 4. Feasby TE, Gilbert JJ, Brown WF, e t al. An acute axonal form of Guillain-Barre polyneuropathy. Brain 1986;109: 1115-1 126. 5. Prineas JW. Pathology of the Guillain-Barre syndrome. Ann Neurol 1981;9(suppl):6-19 6. Brechenmacher C, Viral C, Deminiere C, et al. Guillain-BarrC syndrome: an ultrastructural study of peripheral nerve in 65 patients. Clin Neuropathol 1987;6:19-24 7 . Hughes RAC, Atkinson P, Coates P, et al. Surd nerve biopsies in Guillain-BarrC syndrome: axonal degeneration and macrophage-mediated demyelination and absence of cytornegalovirus genome. Muscle Nerve 1991 (in press) 8 Gainsborough N , Hall SM, Hughes RAC, Leibowirz S. Sarcoid neuropathy. J Neurol 1991;238:177-180 9. Brosnan CF, Selmaj K, Raine, CS. Hypothesis: a role for tumor necrosis factor in immune-mediated demyelinarion and its relevance to multiple sclerosis. J Neuroimmunol 1988;18:87-94 10. Dyck PJ, Lambert EH. Numbers and diameters of nerve fibers and compound action potential of sural nerves: controls and hereditary neuromuscular disorders. Trans Am Soc Neurol 19669 1 : 2 14 11. Dyck PJ, Karnes J, Lais A, et al. Pathologic alterations of the peripheral nervous system of humans. In: Dyck PJ, Thomas PK, Lamberr EH, Bunge R, eds. Peripheral neuropathy. Pliiladelphia: WB Saunders 1984:760-870 12. Brown WF, Feasby TE. Conduction block and denervation in Guillain-Barri. polyneuropathy. Brain 1984; 107:219-239 13. Albers JW, Donofrio PD, McGonagle TK. Sequential elecrrodiagnostic abnormalities in acute inflammatory demyelinating polyrddiculoneuropathy. Muscle Nerve 1985;6:504-509 14. Van der Mechi. FG, Meulsree J, Vermeulen M, l e v i t A. Parterns of conduction failure in the Guillain-Barre syndrome. Brain 1988;111:405-416

Localization of Herpes Simplex Virus and Varicella Zoster Virus DNA in Human Ganglia R. Mahalingam, PhD," M. C. Wellish, BS," A. N. Dueland, MD," R. J. Cohrs, PhD,' and D. H. Gilden, MD'I-

Human dorsal root ganglia from 14 randomly autopsied adults and 1 infant (all seropositive for both herpes simplex virus [HSV) and varicella zoster virus [VZV]) were examined for latent HSV-1 and VZV DNA by polymerase chain reaction. Thoracic ganglionic DNA from all subjects and trigeminal ganglionic DNA from 1 1 adults were analyzed. HSV-1 DNA was detected in trigeminal ganglia from 8 of 1 1 (73%) adults and in thoracic ganglia from 2 of 14 (14%) adults. VZV DNA was detected in trigeminal ganglia from 10 of 11 (91%) adults and in thoracic ganglia from 12 of 14 (86%)adults. None of the DNA samples were positive with primers specific for HSV-2. These findings indicate the presence of latent HSV-1 and VZV DNA in trigeminal ganglia and latent VZV DNA in thoracic ganglia of most seropositive adults. Furthermore,although HSV-1 latency most commonly develops in trigeminal ganglia, we also show for the first time the presence of HSV-1 latency in thoracic ganglia. Finally, both viruses can become latent in the same trigeminal ganglion.

Mahalingam R, Wellish MC, Dueland AN, Cohrs RJ, Gilden D H . Localization of herpes simplex virus and varicella zoster virus D N A in human ganglia. Ann Neurol 1992;31:444-448

Establishment of latent infection is common to all human herpes viruses. To date, the mechanisms of latency and reactivation are not known. During latency, virions cannot be detected, but certain herpes viruses, such as herpes simplex virus type 1 (HSV-l), can often be rescued by explantation of latently infected tissues. For example, explantation of latently infected ganglia leads to reactivation of HSV- 1 from crigcminal [ I } and other cranial nerve ganglia [2) and of HSV-2 from sacral ganglia { 3 ] . In addition, H.SV-1 D N A has been detected in latently infected human rrigeminal ganglia [4}. The development of an HSV cytopathic effect in

From the Departments of 'Neurology and tMicrobiology and Immunology, University of Colorado Health Sciences Center, Denver.

co.

Received Aug 9, 1991, and in revised form Oct 1 I. Accepted for publication Oct 12, 1991. Address correspondence to D r Mahalingam, Department of Neurology, University of Colorado Health Sciences Center, 4200 East Ninth Ave, Campus Box B-183, Denver, C O 80262.

444 Copyright 0 1992 by the American Neurological Association

explants of trigeminal ganglia precludes the isolation of varicella zoster virus (VZV) from the same ganglion, from human thoracic ganand attempts to rescue glia have f i l e d 157. Nevertheless, VZV DNA has been detected in both trigemind and thoracic ganglia by nucleic acid hybridization and polymerase chain reaction

vzv

both HSV and

(PCR)

vzv

DNA

have been found in t r i g e m i d ganglia, it is not known whether both viruses become latent in the same ganglion, which is a question of clinical relevance because establishment of latency by the first herpes virus might affect that of the second. We used PCR technology to study latent and HSV-1 DNA in human ganglia.

vzv

diseases observed in the subjects. None had a history of recent varicella, zoster, or skin lesions consistent with recent infection with VZV or HSV-1. Ganglia were removed aseptically, washed mice in Dulbecco's modified Eagle's medium (Grand Island Biological Co, Grand Island, NY), and stored at - 70°C. Ganglia were analyzed individually or pooled. DNA extracted from four pooled thoracic ganglia from a neonate who died 1 day after birth was used as a control extracted from the human gang,a as specimen. DNA described and adjusted to Fg/pL,

Cells, Viruses, and Antibodies

Propagation of vzv and HSV-1 in African green monkey (BSC-1) cells and extraction of D N A have been described {8, 91. The presence of antibody to both viruses in human sera was determined by immunoprecipitation ( 101.

Materials and Methods Tissue Specimens and DNA Isolation

Polymerase Chain Reaction

Trigeminal and thoracic ganglia were obtained from 14 adult subjects 2 to 24 hours after death. The Table lists the clinical

Oligonucleotide primers (Operon Inc, Alameda, CA) were chosen from published sequences of VZV Ell}, HSV-1 1121,

Results of Polymerase Chain Reaction Analysis of DNA from Human Ganglia Subject No. 1 2

Ageisex

Clinical Diagnosis

Ganglia"

1 day/M 54/M

Lung failure Cardiac arrest

TH-4 TG-right TG-left TH-1 TH- 1 TG-right TG-left TH-5 TG-right TG-left TH-6 TG-2 TH-4 TG-2 TH-4 TG-2 TH-7 TG-2 TH-4 TG-2 TH-4 TG-2 TH-6 TG-2 TH-6 TG-2 TH-5 TH- 1 TH-1 TH-1 TH-7 TH-5

3

671M

Myocardial infarction

4

72lM

Lung cancer

5

63/M

Metastatic carcinoma

6

82/F

Stroke

7

36/F

Stokes-Adam syndrome

8

75lM

Multiple organ failure

9

54/F

Hypoxic encephalopathy

10

59lM

Chronic obstructive pulmonary disease

11

39iM

Renal failure

12

541F

Peritonitis

13

36/F

Amyloidosis

14 15

75/M 55/F

Myocardial infarction Pulmonary hypertension

~~

~

HSV- 1 DNA

HSV-2 DNA

vzv DNA +b

+b +b +b

+

+ + + + + +

+c

+

-

+ + + + + + + + + + + + + + +

~~~

'Number of pooled rhoracic (TH) or rrigeminal (TG) ganglia. bAnalyzed using varicella zosrer virus origin-of-replicarion primers only. 'DNA from pooled thoracic ganglia was positive after amplification with the varicella zosrer virus origin-of-replicarion primers, bur nor with rhe varicella zoster virus UL28 primer.

HSV = herpes simplex virus; VZV

=

varicella zosrer virus; N D = nor derermined.

Brief Communication: Mahalingam et al: HSV and VZV DNA in Human Ganglia 445

and HSV-2 [I 31 DNA. The VZV origin of replication (ORI) UL IRLIRS TRS primers used for amplification and detection have been de* HSV-1 UL30 scribed 17). The specificity of the VZV OR1 primers for the VZV genome has been confirmed by the lack of amplifica-3 -2 tion of cellular D N A isolated from human brain [71.The U‘i UL IRL IRS 32 VZV UL28 primer T (5’-CGGAAC’lTCITTTTCCAnHSV-2 TRL us4 ACAGTA-3’) is located between nucleotides 50390 and 50414, primer U (5’-TAAAATGGCGATCAGAACGGGGTT-3’) is located between nucleotides 50640 and UL IRL IRS TRS 50626, and VZV UL28 internal oligonucleotide UL28 / h. ORI X (5‘-TTCTCTGlTACTACCGCGCC-3’) is located between n d e o t i d e s 50544 and 50564 on the VZV genome. -c -b The HSV-1 UL30 primers used for amplification and detecX U tion have been described [l2}. HSV-2 US4 primer 4 Fig 1. Location of regions ulithin the herpes simplex viru.r types (5’-AGCGATG’lTGTIlTCCCGGGAGGT-3’), primer 6 (5’-AGCAGGGAAGCAmACGAGAGCG-3’), and 1 and 2 (HSV-I, HSV-2) and varicella zoster virus (VZV) geprimer 5 (5’-A?TGTTATGCCTATCCCCGG’lTGG-3’)nomes used for deteciion by polymerase chain reaction. The physare located in the HindIII-L region in the short unique comical map of HSV-1, HSV-2, and VZV consists of two coz’aponent of the HSV-2 genome 1131 (Fig 1). The size of the l’ently linked segments, a unzque long (UL) and LZ unique short products of amplification of VZV ORI, VZV UL28, HSV-1 (US) segment bounded by inoerted repeat sequences (IRL/TRL; UL30, HSV-2 US4, and human beta globin D N A segments IRS/TRS). The HSV-1 UL30 encodes the DNA polymerase are 325, 250, 90, 401, and 110 base pair5 (bp), respectively. gene. The HSV-2 gene U 4 encodeJ the glycoprotein G gene 113). PCR and the detection of virus-specific sequences in the amThe VZV UL28, located in the UL, is predicted t o encode the plified products were performed as described earlier 17,121. VZV DNA pohmerase gene { l 1). The VZV origin of repkcation (ORI) is present as two copies located in the IRS and TRS Results 114). The lines numbered 1,2 and 4,5 and letters T,U and a,b indicate the location of the oligonucleotideprimers 124-mers) The location of the oligonucleotide primers used for usedfor the amplification of regions from the HSV-I UL30, the detection of HSV-1, HSV-2, and VZV DNA on HSV-2 US4, VZV UL28, and VZV ORI, respectively;numthe respective genomes are presented in Figure 1. The bers 3 and 6 and letters X and c indicaie the location of the specificity of the PCR was demonstrated by amplificainternal oligonucleotide primers used t o detect the amplified prodtion of a 325-bp DNA fragment located along the ucts from the HSV-1 UL30, HSV-2 US4, and VZV lJL28 VZV OR1 and by amplification of a 92-bp HSV-1 and ORI , respectively.

4 -

LJ

DNA fragment located within the HSV UL30 in DNA from the respective virus-infected BSC- 1 cells, but not in DNA from uninfected BSC-1 cells or when template DNA was omitted from the PCR (Fig 2). Using HSV-1 UL30 primers capable of amplifying either HSV-1 or HSV-2 DNA sequences [12), DNA from each of two trigeminal and from each of two thoracic ganglia of a normal 54-year-old man (see the Table; Subject 2) was analyzed. HSV-specific DNA sequences were found in each of the two trigeminal ganglia, but not in either of the two thoracic ganglia (see Fig 2). VZV ORI-specific sequences were detected in each of the two trigeminal ganglia and in both thoracic ganglia (see Fig 2). The abundance of VZV-specific sequences varied among the two thoracic ganglia (as shown by a difference in the intensity of the hands of thoracic ganglia-a and ganglia-b of Fig 2). A longer exposure of the autoradiogram resulted in a more intense band in thoracic-b. The distribution of latent HSV and VZV DNA was then prospectively analyzed in 22 trigeminal and 68 thoracic ganglia of 14 adults and in 4 thoracic ganglia from an infant who died 1 day after birth (see the Table). All 15 subjects were seropositive for HSV and VZV. Seropositivity in the l-day-old infant was presumably due to maternally derived antibody. The

446 Annals of Neurology

Vol 31 No 4 April 1992

DNA from all ganglia was analyzed in triplicate by PCR using oligonucleotide primers representing the HSV UL30, VZV ORI, and VZU UL28 (60 Kb upstream from the ORI). All of the DNA samples (except those from Subject 2) were analyzed with both VZV primer sets (OR1 and gene 28). Trigeminal ganglia from each subject were pooled for DNA extraction, and thoracic ganglia from each subject were also pooled for DNA extraction. This pooling enabled analysis of ganglia from many subjects. No VZV o r HSV DNA sequences were detected in DNA from thoracic ganglia of rhe infant (see the Table; Subject 1). VZV DNA was detected (with at least one set of primers) in 10 of 11 (7104) adult trigeminal ganglia. Twelve of 14 individuals (Sly{ ) had VZV DNA in their thoracic ganglia. DNA pooled from four thoracic ganglia of 1 human (Subect 5 ) was positive for the VZV ORI, but negative for VZV UL28. HSV DNA was detected in trigeminal ganglia from 8 of 11 (73%) adults and in thoracic ganglia from 2 of 11 (18%) subjects (Fig 3; Subjects 11 and 3). O n e of the individuals (Subject 8) did nor have HSV DNA in trigeminal ganglia, but did have VZV DNA in both trigeminal and thoracic ganglia.

z z

P c c LL 0

w

U

4

c

0

v tn

CI

z

I

m

c C

z P

c

I

w

-I

3 -I U 0 z 5

z

SUBJECT 2

SUBJECT 2

xP

a

n

I

'1 0 ua

0

a c

HERPES SIMPLEX VIRUS

Fig 2. Detection of herpes simplex virus IHSV) and varicella zoster virus (VZV) D N A in human ganglia by polymerase chain reaction. Total D N A was extracted from BSC-1 cells, two trigeminal and two thoracic ganglia from 1 adult (Subject Z), HSV- I -infected BSC-1 cells and VZV-infected BSC-I cells. Template D N A was omitted from one of the reaction tubes (no DNA). One nanogram of D N A from uninfected and virusinfected cells, and 1 pg of D N A from the remaining samples were usedfor amplification of the HSV-1 U W O and VZV origin of replication regions. The ampl$cation products were analyzed as described earlier 17).

To distinguish between HSV-1 and HSV-2 D N A in positive trigeminal and thoracic ganglia, we used not only HSV-1 UL30 primers capable of amplifying either HSV-1 or HSV-2, but also HSV-2 US4 primers specific for HSV-2. Figure 3 shows the specificity of HSV-2 primers by amplification of a 401-bp HSV-2 US4 segment in D N A isolated from HSV-2-infected BSC-1 cells, but not in D N A isolated from HSV-1infected BSC- 1 cells. HSV-2-specific sequences were absent in all trigeminal and thoracic ganglia analyzed (see the Table). Of particular interest was the analysis of DNA for HSV from thoracic ganglia of 2 individuals (Subjects 11 and 3; see Fig 3). Because the HSV-1 UL30 primers but not the HSV-2 US4 primers amplified HSV D N A in these thoracic ganglia, we conclude that the two HSV-positive thoracic ganglia contained HSV-1 DNA.

VARICELLA ZOSTER VIRUS

Discussion Our findings support previous studies that showed VZV establishes latency in both trigeminal and thoracic ganglia 16, 7). Furthermore, we confirm not only that HSV-1 latency is established predominantly in trigeminal ganglia [I}, but also demonstrate HSV-1 latency in thoracic ganglia for the first time. Earlier attempts to recover herpes viruses 13, 51 from human thoracic ganglia failed. Because even a low HSV burden would be expected to reactivate during explantation, the inability to rescue latent HSV from thoracic ganglia could be due to limited sampling. Another possibility is the presence of an incomplete HSV genome, which can be tested by PCR using primers representing different regions of the virus genome. We had previously demonstrated latent VZV D N A in thoracic ganglia from 50% of humans examined [7}. The detection of latent VZV in thoracic ganglia from 86% of humans in this study most likely reflects the use of more thoracic ganglia (5-6) from each subject examined. Taken together with the detection of VZV in trigeminal ganglia from 71% of humans examined in this study, our findings are consistent with the classic epidemiological studies of Hope-Simpson [15], in which the trigeminal and thoracic dermatomes were shown to be the most frequent sites of zoster. The differential localization of HSV and VZV latency may reflect the anatomical site of primary infec-

Brief Communication: Mahalingam et al: HSV and VZV DNA in Human Ganglia

447

D N A from the trigeminal and thoracic ganglia of humans are d u e to the presence of HSV-1.

C

.-0 o

.l-

Supported in part by Public Health Service Grants AG 06127, AG 07347, and N S 07321 from the National lnsrirutes of Health, and a grant from the Beatrice and Roy Backus Foundation. A. N. D. is the recipient of an advanced postdocroral fellowship award from the National Multiple Sclerosis Society.

a, C .-

.4-

I

Thoracic

r-

I

> cn

I


ereanahzed with internal olzgonuc/eotide probes speci;ficfor the regions of the respective imiral genomes as described 17).

tion. Our findings suggest that initial infection of the oropharynx by HSV results in latency primarily in cranial nerve ganglia, whereas the widespread involvement of skin during primary infection by VZV (i.e., chicken pox) parallels concurrent infection of most or all ganglia. Furthermore, both HSV-1 and VZV can establish latency in the Same trigeminal ganglion. Subsequent in situ hybridization studies will determine if the same o r different cells in dually infected trigeminal ganglia are latently infected. The primers used for the detection of HSV-1 D N A were from a region of the viral genome that is identical to that of HSV-2 [16}. Therefore, w e also used primers specific for HSV-2 from a 1,400-bp segment of the HSV-2 genome that is absent in the HSV-1 genome {IJ]. N o n e of the samples analyzed contained HSV-2 D N A , indicating that the HSV sequences detected in

448

Annals o f Neurology

Vol i1

No

4 April I092

We are grateful to the Departments of Pathology at University Hospital, the Veterans Administration Hospital, Denver General Hospital, the Presbyterian-St Luke’s Medical Center, the Children’s Iiospital, and the Fitzsimmons Army Medical Center in Denver for allowing us access to autopsy material We thank D r Abbas Vafai for performing the immunoprecipitarion analysis, D r Wroblewska for providing HSV-2 virus, Mary Devlin for technical assistance, Marina Hoffman for editorial review, and Cathy Allen for preparation of the manuscript.

References 1. Baringer JF, Swoveland P. Recovery of herpes simplex v m s from human trigeminal ganglions. N Engl J Med 1973;2X8: 648-650 2. Warren KG, Brown SM, Wroblewska 2, et d.Isolation of latent herpes virus from the superior cervical and vagus ganglions of human beings. N Engl J Med 1978;298:1008-1069 3 . Baringer JR. Recovery of herpes simplex virus from human sacral ganglions. N Engl J Med 1974;291:828-830 4. Efsrarhiou SE, Minson AC, Field HJ, et al. Detection of herpcs simplex virus-specific D N A sequences in latently infected mice and humans. J Virol 1986;57:440-455 5 . Plotkin SA, Stein S, Snyder M, lnimesoete P. Attempts to recover varicella virus from ganglia. Ann Neurol 15)7’;2:249 6. Gilden D H , Vafai A, Shrram Y, et al. Varicella zosrer virus D N A in human sensory ganglia. Nature 13Xi,i06:478-4X0 7. Mahalingam R, Wellish M, Wolf W, et a]. Latent varicella-zosrcr viral D N A in human rrigeminal and thoracic ganglia. N Eogl J Med 1990;323:627-63 1 8. Gilden DH, Hayward AR, Krupp J, et al. Varicella zoster virus infection o f human inononuclear cells. Virus Res 198 . 7 . 117-129 9. Gilden DH, Shtram Y , Friedmann A, et al. Extraction of cellassociated varicella zoster virus D N A with triton X-lOO-Na< I. J Virol Methods 1982;4:263-275 10. Vafai A, Mahalingam R, Zerbe G, er al. Detection of antibodies to varicella zoster virus proteins in sera trom the elderly. Gerontology 1988;34:242-249 11. Davison AJ, Scott JE. Thc complete D N A sequence of vartcclla zoster virus. J Gen Virol 1986;67:1759-1810 12. Cao M, Xiao X, Egberr B, er al. Rapid detection of cutaneous herpes simplex virus infection with the polymerase chain reaction. J Invest Dermatol 1989;82:191-192 13. McGeoch DJ, Moss HWM, McNah D, Frame MC. D N A sequence and genetic content of the Hindlll-L region in the short unique component of the herpes simplex virus type 2 genome identification of the gene encoding glvcoprotein G, and cvolutionary comparisons. J G r n Virol 1987;68:19-38 14. Stow N D , Davison AJ. Identification of a varicella zoster virw origin of D N A replication and its activation by herpes simplex virus type 1 gene products. J Gen Virol 1986;07:161 3-102 3 15. Hope-Simpson RE. The nature of herpes zoster: a long-term study and a new hypothesis. Proc K Soc Med 1965;589-20 16. Tsurumi T, Maeno K, Nishiyama Y. Nucleotidc sequence of the D N A polymerase gene of herpes simplex virus type 2 ancl comparison with the type 1 counterpart. Gene I987;52: I2l)I 37