Intrathecal humoral immune reaction in zoster infections

Journal of the Neurological Sciences, 103 (1991) 101-104

© 1991 Elsevier Science Publishers B.V. 0022-510X/91/$03.50

101

JNS 03519

Intrathecal humoral immune reaction in zoster infections H . - J . Sch~tdlich, M . N e k i c , J. J e s k e a n d H . K a r b e Neurological Clinic of the Universityof Cologne, Cologne (F.R.G,)

(Received 8 July, 1990) (Revised, received 23 November, 1990) (Accepted 11 December, 1990) Key words: Zoster infections; Intrathecai humoral immune reaction; Immunoblotting

Summary Intrathecal humoral immune reaction in 26 patients with a reactivation of varicella-zoster virus was analyzed. 11 suffered from ganglionitis, in 7 cases an additional affection of the lower motor neuron was demonstrable. In 8 patients, meningitis, myelitis or cerebral infarctions by zoster angiitis were diagnosed. Intrathecal immune reaction in ganglionitis was weak whereas an intense IgG synthesis became demonstrable in all meningomyelitis/cerebral infarction cases. As demonstrated by immunoblotting, in early stages of the disease immune reaction in serum and cerebrospinal fluid differed only quantitatively. In the further course, most intrathecally synthesized antibodies were directed against low molecular antigens whereas serum pattern did not change. In some patients, additional antibodies not detectable in serum were demonstrable. Though intensity markedly differed, no qualitative differences between the immune response in ganglionitis and more widespread zoster infections of the CNS were detectable.

Introduction After primary infection with varicella-zoster virus (VZV) the infectious agent persists in the central nervous system (CNS; Esiri and Tomlinson 1972). If reactivation occurs, ganglionitis mostly develops with typical cutaneous lesions, a clinical inapparent meningitis and minimal cerebrospinal fluid (CSF) changes (Barnes and Whitley 1986). Rarely inflammatory lesions are more extended and cause severe clinical syndromes like meningoencephalitis, myelitis or cerebral infarctions (Jemsek etal. 1983; Barnes and Whitley 1986). In these cases, an intense intrathecal humoral immune response is demonstrable with marked pleocytosis and a local IgG synthesis (Echevarria et al. 1987; Mathiesen et al. 1989). It is unresolved whether the intensity of the humoral immune response, which obviously depends upon the site and extension of the inflammatory lesions, is the only distinguishing factor in these clinical syndromes or if there are any qualitative differences. By immunoblotting it has become possible to gain more precise information about intrathecally synthesized antibodies. Therefore that method

Correspondence to: Dr. H.-J. Sch~ldlich, Klinik und Poliklinik far Neurologie und Psychiatrie der Universit~t zu KSln, Joseph-StelzmannStrage 9, D-5000 KOln41, F.R.G.

has been used to compare C S F alterations in zoster ganglionitis and myelomeningitis cases.

Patients and methods 26 patients with zoster infections were investigated. 11 had ganglionitis, 3 an additional facial palsy and 4 an affection of the lower motor neuron. Meningitis was found in 4 cases, 2 suffered from myelitis and in 2 patients cerebral infarctions due to inflammatory lesions of the middle cerebral artery were diagnosed. In 24 patients, typical zoster effiuorescences were demonstrable, in 2cases (1 myelitis, 1 cerebral infarction) diagnosis was based upon CSFfindings as reported previously (Karenberg et al. 1988; Sandmann et al. 1989). No patient suffered from disease comprising the immune system or was treated with an immunosuppressive therapy. CSF was taken by lumbar puncture. In patients with ganglionitis, facial palsy and affections of the lower motor neuron mean time between the onset of clinical symptoms and C S F examination was 33 days. Only in 3 cases was a single puncture done within the first week. In all cases of meningomyelitis, repeated punctures were done. Cell numbers and total protein were determined, the detection of an intrathecal IgG synthesis was performed by end-point laser nephelometry according to Felgenhauer and

102 coworkers (1982). Intrathecally synthesized zoster antibodies were demonstrated by an enzyme-linked immunosorbent assay (ELISA) after adjusting IgG levels in serum and CSF to 1 mg/dl as previously reported (Felgenhauer et al. 1985). Immunoblotting was done in all patients by the following method: 1 ml VZVantigen (EIA, Virion, Wtlrzburg, F.R.G.) was dissolved in 150 #1 distilled water and a 1 : 2 dilution was made with 0.5 M Tris-HC1, pH 6.8, 2.5 ml; SDS 600mg; fl-mercaptoethanol 1 ml; glycerin 2ml; bromophenol red 20 #1; made up to 10 ml. Probe volume was 20 #1. SDS-PAGE was performed as described previously (Sch~dlich et al. 1990). Gel length was 10 cm (collection gel 3 cm, separation gel 7 cm), thickness 1 mm. A current of 1.5 mA/cm was applied until the dye reached the separation gel and was increased to 2 mA/cm gel throughout the whole procedure. After SDS-PAGE, semidry-blotting with a discontinuous buffer system on nitrocellulose membranes was done (LKB Novablot, Broma, Sweden). According to the manufacturer's recommendations, the following buffer solutions were used: Anode I (pH 10.4) 0.3 M Tris 36.6 g, methanol 20% 200 ml, made up to 1000 ml. Anode II (pH 10.4) 25 mM "Iris 3.03 g, methanol 20% 200 ml, ad 1000 ml. Cathode (pH 7.6) 40 mM 6-aminohexanoic acid, methanol 20% 200 ml, ad 1000 ml. A current of 1 mA/cm 2 was applied for 1.5 h. Blocking was done with bovine serum or albumine. After blotting, nitrocellulose membranes were incubated with serum and CSF after adjusting both fluids to identical IgG concentrations (0.1-1.0 mg/dl) with phosphate buffer. Both probes were examined side by side. After 1 h at room temperature, 3 washes were done. Membranes were then incubated with peroxidase-labeled goat anti-human IgG antibodies for 1 h. Staining was done with AEC (3-amino-9-ethylcarbazol) for 15-20 min. After staining, copies were made on transparent foil. Evaluation was done by densitometry (Vitatron, Dieren, The Netherlands). Since patients with ganglionitis, facial palsy and other affections of the lower motor neuron exhibited very similar CSF changes, these cases will be combined as "ganglionitis/lower motor neuron involvement (LMI)".

Cetls

Ig(3~oc

QAIb (5e/CSF)

(/at-~) 730

500

> 160 400 +

150

o

8° 80

300

o

100 200

0%°o°00oo o

50

ooooo

100 oo

"o

oo o

oo

o

Ooo~

Fig. 1. CSF cellnumber,permeabilityof the blood-CSF barrier (normal value > 160) and intrathecal IgG synthesis in zoster ganglionitis/lower motor neuron affection (O) and meningomyelitis/cerebral infarctions (Q). as in 2 ganglionitis/LMI and both angiitis cases, cell count was slightly elevated or normal. Increased permeability of the blood-CSF barrier (albumine Se/CSF < 160) was demonstrable in 61% (11/18) ganglionitis/LMI patients as compared to 63% (5/8) of meningomyelitis and cerebral infarction cases. By laser-nephelometry, intrathecal IgG synthesis was found in all patients with meningomyelitis and cerebral infarctions. In ganglionitis/LMI local IgG production could be detected in only one case. With ELISA, intrathecal zoster antibody synthesis became detectable in 44 % (8/18) ofganglionitis/LMI cases.

Results

Exclusively mononuclear cells were demonstrable in all cases. Cell count was slightly elevated in patients with ganglionitis/LMI (max. 66 cells. #1 - ~, Fig. 1). In meningomyelitis, cell elevation was more pronounced (maximum 730 cells. #1- ~). If puncture was done in late disease stages

i

II

Fig. 2. Varicella-zostervirus after SDS-PAGE. (a)By electrophoresis and staining with Coomassie blue numerous distinct bands becomevisible. (b)After immunoblottingand reaction with CSF, most antibodies are directed against antigenic determinants with a molecular mass of 89-110 kDa (antigenI) and 50-56 kDa (antigenII).

103 E 70 60 50 40 30 20 10 100 RF E

b

7O 6O 5O

3O 20 10

100

Fig. 3. Immunoblotting in one patient with zoster myelitis. ( a ) A t 14th day after the onset of clinical symptoms most intense antibody reaction in serum and CSF is directed against antigen complexes I and II. Antibody reaction patterns are nearly identical in both compartments. ( b ) A t the 47th day intrathecal antibody reaction against antigen I is weak. Local antibody synthesis has expanded and an additional synthesis especially against low molecular antigens has become detectable. E = extinction; R F = retention factor. Thin line = serum; heavy line = CSF.

In meningomyelitis and cerebral infarctions all patients exhibited locally synthesized antibodies. Numerous distinct bands became visible on SDSPAGE of VZV. As shown in Fig. 2, main antibody reaction was directed against antigenic determinants with a molecular mass of 50-56 and 89-110 kDa. Though these complexes obviously are composed of different viral proteins, they will be called antigen I and II in the context of this paper. The biological significance of these antigens remained unresolved since no further attempt for characterization was made. In all patients with meningomyelitis/cerebral infarctions and in 8 out of 18 ganglionitis/LMI cases intrathecal antibody synthesis was detectable by immunoblotting. If repeated CSF punctures were done, a sequential development of the intrathecal humoral immune response was detectable: early after the onset of clinical symptoms, locally synthesized antibodies were directed against identical antigenic determinants as those from serum, i.e. humoral immune response in serum and CSF differed only quantitatively (Fig. 3). Since serum and CSF had been adjusted to identical IgG concentrations and the antiviral reaction was more pronounced in CSF, these antibodies were obviously synthesized intrathecally. In later stages, intrathecal antibody synthesis against antigen I diminished whereas the reaction against antigen II increased (Fig. 4). In some

C,SF/Se 2,5 0

2,0

00 •

1,5

E 70

OlO

60

1,0 0

50 g° 0,5

30 2O 10

89 -110 kDa 0 RF

100

Fig. 4. Immunoblotting in one patient with zoster meningitis at the 45th day after the onset of clinical symptoms. Intrathecal reaction against antigen I is weak whereas a distinct antibody synthesis against antigen I I is demonstrable. E = e x t i n c t i o n , R F = retention factor. Thin line = serum; heavy line = CSF.

20- 56 kDa

Fig. 5. Comparison of the antibody reaction against antigen I and II in serum and CSF. If antibody portions (%, evaluated by densitometry as described in the text) are compared in later stages of the disease (t4th to 123th day after the onset of clinical symptoms, mean = 37 days), most pronounced reaction in CSF is directed against antigen II and low molecular antigens. There is no difference between ganglionitis/LMI (O) and meningomyelitis (O) cases.

104 patients (4 meningomyelitis, 1 gangiionitis/LMI), intrathecal humoral immune response expanded and antibody patterns against low molecular antigens ( 2 0 - 5 0 k D a ) became detectable which were not demonstrable in serum at identical I g G levels (Fig. 3). Intensity shift towards antigen II was detectable in nearly all C S F samples investigated in later disease stages irrespective of whether the patients suffered from ganglionitis/LMI or meningomyelitis/cerebral infarctions (Fig. 5). In serum, reaction pattern remained stable and only quantitative variations were found.

The present investigation confirms that there are wide variations in the intensity of the humoral immune response between ganglionitis and more extended syndromes in zoster. By more sophisticated techniques, however, noqualitative differences could be demonstrated. These results let us presume that differences in intensity are merely due to the extension o f the inflammatory process within the CNS. There is a striking resemblance in the development of the humoral immune response in zoster and herpes simplex encephalitis. At the present time, it is open to debate whether these similarities are restricted to the group of herpes viruses or whether similar results may also be obtained in C N S infections by other viruses.

Discussion Reactivation of VZV leads to different clinical syndromes. Besides ganglionitis, severe meningitis, encephalitis, myelitis, and cerebral infarctions may occur (Jemsek et al. 1983; Barnes and Whitley 1986). As c o m m o n pathogenetic mechanism, virus-induced angiitis has been demonstrated (Blue and Rosenblum 1983). It is well known that in ganglionitis intrathecal immune response is weak whereas severe clinical syndromes frequently are associated with marked pleocytosis and an intense immune response within the C N S (Gershon et al. 1980). These findings are confirmed by the present investigation. In no patient with ganglionitis or facial palsy and in only one case suffering from a widespread lower motor neuron involvement local IgG synthesis was demonstrable. In meningomyelitis and cerebral infarctions however, an intense humoral immune response was detectable in all cases. Even with a very sensitive E L I S A , less than half of the ganglionitis/LMI group exhibited locally synthesized zoster antibodies. Further differentiation of the local immune response with immunoblotting revealed that antibodies locally synthesized early in the disease reacted with identical antigens as those from serum. In the further course, intrathecal antibody synthesis against low molecular antigen increased. In some patients local reaction dilated and an immune reaction to additional antigens not detectable in serum became demonstrable. Although the intensity of the intrathecal immune reaction varied, no differences between ganglionitis and meningomyelitis cases were detectable. As has been previously reported (Schadlich et al. 1990), intrathecal immune reaction in herpes encephalitis develops in a very similar manner: in the first phase, immune response in serum and C S F differs only quantitatively. Later on, in some patients intrathecal immune response expands with an antibody synthesis to antigens not detectable in serum.

References Barnes, W.D. and R.J. Whitley (1986) CNS diseases associated with varicena zoster virus and herpes simplex virus infection. Neurol. Clin., 4 (1): 265-283. Blue, M.C. and W.I. Rosenblum (1983) Granulomatous angiitis of the brain with herpes zoster and varicella encephalitis. Arch. Pathol. Lab. Med., 107: 126-128. Echevarria, J.M., P. Martinez-Martin, A. Tellez, F. deOry, J.L. Rapun, A. Bernal, E. Estevez and R. Najera (1987) Aseptic meningitis due to varicella-zoster virus: serum antibody levels and local synthesis of specific IgG, IgM and IgA. J. Infect. Dis., 155 (5): 959-967. Esiri, M.M. and A.H. Tomlinson (1972) Herpes zoster: demonstration of virus in trigeminal nerve and ganglion by immunofluorescence and electron microscopy. J. Neurol. Sci., 15: 35-48. Felgenhaner, K., C. Hansen and A. Remy (1982) Laser-nephelometric evaluation of the humoral immune response in the central nervous system. Med. Lab., 11: 48-57. Felgenhaner, K., H.-J. Sch~dlich, M. Nekic and R. Ackermann (1985) Cerebrospinal fluid antibodies - a diagnostic indicator for multiple sclerosis? J. Neurol. Sci., 71: 291-299. Jemsek, J., S.B. Greenberg, L. Taber, D. Harvey, A. Gershon and R.B. Couch (1983) Herpes zoster associated encephalitis: clinicopathologic report of 12 cases and review of the literature. Medicine, 62: 81-97. Karenberg, A., H.-J. Sch~tdlich, H. Karbe, M. Neveling and F. Thun (1988) Multiple Hirninfarkte bei Zosterinfektion. Nervenarzt, 59: 45-47. Mathiesen. T., H. von Hoist, S. Fredrikson, G. Wirsen, B. Nederstedt, E. Norrby, V.A. Sundqvist and B. Wahren (1989) Total, anti-viral, and anti-myelin IgG subclass reactivity in inflammatory diseases of the central nervous system. J. Neurol., 236 (4): 238-242. Sandmann, J., H.-J. Schiidlich and J. Jeske (1989) Differential-diagnostische und therapeutische Probleme der Zosterenzephalomyelitis. Akta Neurol., 16: 165-167. Schadlich, H.-J., M. Nekic and K. Felgenhauer (1990) Differentiation of the humoral immune response in herpes simple;~ encephalitis. J. Neuroimmunol., 27: 49-53.