Altered physiology of purkinje neurons in cerebellar slices from transgenic mice with chronic central nervous system expression of interleukin-6

Pergamon

Neuroscience Vol. 89, No. I, pp. 127-136, 1999 Copyright ~; 1998 IBRO. Published by ElsevierScience Ltd Printed in Great Britain. All rights reserved PII: S0306-4522(98)00316-9 0306~-522/99 $19.00+0.00

ALTERED PHYSIOLOGY OF PURKINJE NEURONS IN CEREBELLAR SLICES FROM TRANSGENIC MICE WITH CHRONIC CENTRAL NERVOUS SYSTEM EXPRESSION OF INTERLEUKIN-6 T. E. N E L S O N , I. L. C A M P B E L L and D. L. G R U O L * Department of Neuropharmacology and AIDS Research Center, The Scripps Research Institute, La Jolla, CA 92037, U.S.A. Abstract--The cytokine interleukin-6 is produced at elevated levels within the central nervous system in a number of neurological diseases and has been proposed to contribute to the histopathologic, pathophysiologic, and cognitive deficits associated with such disorders. In order to determine the effects of chronic exposure of interleukin-6 on the physiology of central neurons, we compared the firing properties of cerebellar Purkinje neurons from control mice and transgenic mice that chronically express interleukin-6 within the central nervous system. Extracellular recordings from cerebellar slices revealed that the mean firing rate of spontaneously active Purkinje neurons was significantly reduced in slices from transgenic mice compared to control mice. In addition, a significantly greater proportion of Purkinje neurons from transgenic slices exhibited an oscillatory pattern of spontaneous firing than neurons in control slices. Orthodromic stimulation of climbing fiber afferents evoked similar excitatory synaptic responses (complex spikes) in Purkinje neurons of both transgenic and control mice. However, the inhibitory period following the complex spike (climbing fiber pause) was significantly longer in slices from transgenic mice. Using immunohistochemistry, we also showed that Purkinje neurons express high levels of both the interleukin-6 receptor and its intracellular signaling subunit, gpl30, indicating that interleukin-6 could act directly on Purkinje neurons to alter their physiological properties. The interleukin-6 expressing transgenic mice have been shown previously to exhibit a number of histopathological changes in the central nervous system including injury and loss of cerebellar Purkinje neurons. The present data show that these transgenic mice also have altered physiology of cerebellar Purkinje neurons, potentially through a direct activation of interleukin-6 receptors expressed by this neuronal type. Interleukin-6 induced alterations of Purkinje neuron physiology would ultimately affect the flow of information out of the cerebellum, and could thus contribute to the motor deficits observed in the transgenic mice. © 1998 IBRO. Published by Elsevier Science Ltd. AYe), word~: interleukin-6, Purkinje neuron, cerebellum, transgenic, electrophysiology, immunohisto-

chemistry.

Interleukin-6 (IL-6), is one of a number of cytokines known to be an important mediator of inflammatory and immune responses. 2'21 Cytokines such as IL-6 are produced in the periphery and act systemically to induce growth and differentiation of cells in the immune and hematopoietic systems and to induce and co-ordinate the different elements of the acutephase inflammatory response. 19 Cytokines can also act centrally to activate neural aspects o f the immune response such as the hypothalamic-pituitary axis, thus serving as a regulatory signal between the immune system and the C N S . 5'37 There is now evidence that IL-6 is also produced within the CNS and may be important for cell-to-cell signaling associated with such functions as the coordination of neuroimmune responses, protection of *To whom correspondence should be addressed. Abbreviathms: ACSF, artificial cerebrospinal fluid; AIDS,

acquired immunodeficiency syndrome; GFAP, glial fibrillary acidic protein; IL-6, interleukin-6; LTD, long-term depression; PBS, phosphate-buffered saline.

neurons from insult, as well as neuronal differentiation, growth and survival. 5 IL-6 is produced within the CNS by astrocytes and microglia 6'14 and under normal conditions the CNS expression of IL-6 is highly regulated. 5'23 However, it is now apparent that dysregulation of IL-6 production in the brain is associated with a number of neurological diseases and that chronic exposure to increased amounts of IL-6 may contribute to the neuropathological and pathophysiological sequelae of these disorders. 5"34 Elevated levels of IL-6 are common to several neurological disorders such as acquired immunodeficiency syndrome (AIDS) dementia complex, 17~44 Alzheimer's disease, 3 multiple sclerosis, 33 systemic lupus erythematosus, 22 CNS trauma, 47 and viral ~4 and bacterial 26 meningitis. While much is known about the function of IL-6 as a peripheral mediator of inflammatory and immune responses, little is known about the effect of IL-6 on the normal physiology of CNS neurons or how chronic exposure of neurons to cytokines alters their function in various neurological diseases. 127

128

T.E. Nelson et al.

A model of CNS expression of IL-6 has been developed recently using transgenic mice in which constitutive IL-6 expression within the CNS is targeted to astrocytes and is regulated by the glial fibrillary acidic protein (GFAP) promoter. 8 These mice (GFAP-IL6 mice) exhibit many of the neuropathologic characteristics associated with diseases in which IL-6 levels are elevated, as well as pathophysiologic properties that are pertinent to some of the cognitive deficits observed in these conditions. 4"8'2°'42 For example, the hippocampal formation of GFAP-IL6 mice undergoes several alterations, including neuronal injury and loss, changes in dendritic morphology, and gliosis.8 Coincident with these anatomical abnormalities are several functional changes within the hippocampus, including decreased synaptic plasticity4 and increased recurrent inhibition in the dentate gyrus, 42 anomalous hippocampal electroencephalogram activity,42 and the onset of hippocampal seizure activity. 8'42 Moreover, G F A P IL6 mice exhibit decreased performance in an avoidance learning paradigm that progresses with age and corresponds to the level of neural damage and inflammation in the CNS. 2° GFAP-IL6 mice also exhibit neurological symptoms that suggest a deficit in the cerebellar control of movement. 8 These mice exhibit motor incoordination and cerebellar ataxia, which have also been reported in a number of neurological disorders associated with elevated CNS levels of IL-6, such as AIDS dementia complex, 18,35 Alzheimer's disease, 1,45 systemic lupus erythematosus, 41"43 and multiple sclerosis.l ~,46 The CNS expression of IL-6 m R N A in the GFAP-IL6 mice is highest within the cerebellum relative to other brain regions. 8 Moreover, the cerebella of G F A P - I L 6 transgenic mice feature a number of histopathologic abnormalities, including atrophy and loss of Purkinje neurons, disorganization of the Purkinje and granule cell layers, astrocytosis, angiogenesis, spongiosis, and increased levels of mononuclear inflammatory cells in the cerebellar sulci. 8 Such pathology induced by chronic IL-6 expression could contribute to the abnormal cerebellar function in these mice. However, the effects of IL-6 on the physiology of CNS neurons, both at physiological and pathological levels of expression, need to be elucidated. Thus, we have begun to investigate whether chronic IL-6 expression within the CNS disrupts the physiology of cerebellar Purkinje neurons using the GFAP-IL6 transgenic mouse model. An in vitro slice preparation was utilized for these experiments in order to minimize the effects of circulating hormones and other factors that could potentially confound the results in an in vivo preparation, particularly in light of the severe compromise of the blood-brain barrier found in the GFAP-IL6 transgenic mice. 7 Using extracellular recording techniques, we compared the spontaneous firing activity of Purkinje neurons, as well as the response of Purkinje neurons

to synaptic input from climbing fibers, in cerebellar slices from control and GFAP-IL6 transgenic mice. The spontaneous firing activity and the response of Purkinje neurons to climbing fiber activation have been well characterized and are dependent on endogenous physiological properties of Purkinje neurons. 24'25'27'31'32 Taken together, these two measures, spontaneous activity and synaptic function, can serve as sensitive detectors of the physiological state of Purkinje neurons. Extracellular recording provided an effective means for studying these neuronal properties. In addition, using this non-invasive recording technique avoided the potential confound of differential sensitivities of Purkinje neurons in control and GFAP-IL6 slices to neuronal damage that can occur with intracellular recording. The Purkinje neuron is the sole output neuron of the cerebellar cortex and it plays an important role in motor control. Bergmann glial cells, which are in close contact with Purkinje neurons, express IL-6 m R N A in the GFAP-IL6 mice. a Thus, in addition to being a critical component in cerebellar function, Purkinje neurons are likely to be directly exposed to the transgene-encoded IL-6 in this model. IL-6 produces its biological effects by acting at the IL-6 receptor which, in turn, activates the membranebound signal transduction peptide, gpl30, leading to stimulation of multiple intracellular signal transduction pathways that eventually regulate gene transcription. 29 IL-6 receptor m R N A has been reported to be expressed within the cerebellum. 39'4° However, the cell types that express the IL-6 receptor have not been clearly identified. Thus, in order to determine if IL-6 can act directly on Purkinje neurons, we used immunohistochemistry to ascertain the cellular localization of the IL-6 receptor and gp 130 within the cerebella of control and GFAP-IL6 mice.

EXPERIMENTAL PROCEDURES

Transgenic mice

Production of the GFAP-IL6 transgenic mice has been described in detail elsewhere.8 Briefly, IL-6 expression was targeted to astrocytes by an expression vector derived from the murine GFAP gene. Full-length murine IL-6 cDNA was modified and inserted into the GFAP gene. The genes were then microinjected into fertilized eggs of F1 generation hybrid mice (C57BL/6J x SJL). After weaning (three- to four-weeks-old) transgenic animals were identified by slot blot analysis of tail DNA and separated from their nontransgenic littermates. Heterozygous mice of the G369 transgenic line, which express high levels of IL-6, were used in this study. A second transgenic line (G167) has also been produced which expresses lower levels of IL-6 and exhibits similar histopathological and neurological features as the G369 mice, although to a lesser extent. ~ However, we have focused on the G369 transgenic mice because of their higher level of neurologic disease. Age-matched littermates which did not express the IL-6 transgene were used as controls. All efforts were made to minimize animal suffering, to reduce the number of animals used, and to utilize alternatives to in vivo techniques.

Purkinje neurons from interleukin-6 transgenic mice

Slice preparation The ages of the mice (n=29) used for preparation of cerebellar slices ranged from 31 to 73 days. Control and GFAP-1L6 mice were anesthetized with halothane and killed by decapitation. The cerebellum and brainstem were removed rapidly and immersed in ice-cold, oxygenated artificial cerebrospinal fluid (ACSF) to cool the tissue and slow down metabolism. Following several minutes of immersion, the cerebellum was separated from the underlying brainstem. On one side of the cerebellum, the lateral hemisphere was transected sagittally and the tissue block containing the vermis was glued with cyanoacrylate (Super Glue Corporation) to the stage of a vibrating tissue slicer (Campden Instruments). Sagittal slices of the vermis were then cut at a thickness of 350 ~m and transferred to an incubation chamber containing oxygenated ACSF at room temperature. Slices remained in this chamber for a minimum of 1.5 h until they were used for recording. ACSF consisted of (in mM) 130 NaCI, 3.5 KCI, 1.25 NaH2PO4, 24 NaHCO3, 2.2 CaC1y2H20, 2 MgSO4.7H20, and 10 glucose. The ACSF was adjusted to pH 7.3-7.4 and the osmolarity was approximately 300 mOsm. The ACSF was bubbled continuously with a mixture of 95% 02/5% CO~. Prior to recording, each slice was transferred to the recording chamber. The chamber was a gas-fluid interface perfusion chamber maintained at approximately 35°C. A constant flow (1.5-2.0 ml/min) of oxygenated ACSF was maintained through the chamber by a peristaltic pump. Slices were allowed to equilibrate in the recording chamber for at least 15 rain before recording was started.

Electrophysiology Glass microelectrodes were filled with 3 M NaCl (1.01.5 M resistance) and lowered through the Purkinje cell layer until a single neuron could be recorded in isolation. The Purkinje cell layer of the cerebellar cortex was easily identifiable under low magnification using a dissecting microscope. Single-unit extracellular recordings of spontaneously-occurring action potentials were obtained from neurons in the Purkinje cell layer to characterize their rate and pattern of firing. Sample recordings were taken from all cerebellar lobules. However, no differences in the physiological properties were observed between lobules and, thus, all data were pooled for analysis. In order to assess changes in the response of Purkinje neurons to afferent input, the climbing fiber pathway was stimulated electrically using a concentric bipolar electrode placed in the white matter of the respective lobule. Stimulus intensities sufficient to elicit synaptic responses from Purkinje neurons were normally between 20-40 V, and the stimulation rate was either 0.5 or 1 Hz. Approximately three slices per animal were used from both control and transgenic mice and about three Purkinje neurons were recorded from each cerebellar slice. Ratemeter histograms were produced by summing action potentials over 500 ms intervals using a window discriminator and displaying the data on a chart recorder (Gould). Firing rate was quantified by digitizing selected periods of recording using an IBM-compatible computer equipped with data acquisition hardware and software. Data were also stored on magnetic tape for off-line analysis using Axotape and Axograph software (Axon Instruments).

Immunoh&tochemistrv Brains of control (n=2) and GFAP-IL6 transgenic (n= 3) mice (three months-of-age) were removed and immersed in Bouin's fixative. The brains were embedded in paraffin and 5-gm-thick sagittal sections were cut. The sections were deparaffinized in AmeriClear (Baxter), rehydrated through a series of alcohols (100%, 95%, 80%, 50%) into phosphatebuffered saline (PBS). Following initial blocking with

129

PBS/I% goat serum (Pharmingen), the sections were incubated overnight in the primary antibody (diluted 1:100 in PBS+I% goat serum). Antibodies used (Santa Cruz Biotechnology, Inc.) were to the precursor proteins for the IL-6 receptor and gpl30. The IL-6 receptor antibody was an affinity-purified rabbit polyclonal antibody raised against amino acids 441-460 of the carboxy terminus of mouse IL-6Ra protein. The gp 130 antibody was an affinity-purified rabbit polyclonal antibody raised against amino acids 895914 of the carboxy terminus of mouse gpl30 protein. As a control for non-specific staining, prior to staining primary antibodies were co-incubated with the respective antigenic peptides (hl0 dilution) used to produce the primary antibodies. After rinsing in PBS and PBS/I% goat serum, the sections were incubated in biotinylated secondary antibody (Vectastain anti-rabbit) for l h. The sections were rinsed again in PBS and PBS/I% goat serum and incubated in avidin/biotin complex (Vectastain) for 1 h. After rinsing in PBS the sections were exposed to diaminobenzidine and H202 for approximately 12rain, until the reaction was stopped by rinsing in water. The sections were then counterstained with Hematoxylin for 30 s, rinsed in water, dehydrated in a series of ethanol, cleared in xylene, and coverslips were added. Some sections were not counterstained in order to more clearly visualize the immunostaining. The slides were examined by light microscopy. RESULTS

Altered spontaneous firing of Purkinje neurons in glial fibrillary acidic protein-interleukin-6 transgenic mice Purkinje neurons were identified visually by their position in the Purkinje cell layer and electrophysiologically by their firing characteristics. Purkinje neurons in control slices exhibited a regular, high frequency firing pattern and large action potentials, or "simple spikes" (Fig. 1A), similar to that described for Purkinje neurons in other preparations. J3,~5,30 In general, Purkinje neurons in G F A P - I L 6 slices showed similar spontaneous firing properties (Fig. 1A). However, visual identification was more difficult in G F A P - I L 6 slices due to the loss of Purkinje neurons and a less organized Purkinje cell layer. Although Purkinje neurons in slices from both control and G F A P - I L 6 mice (31 73 days postnatal) exhibited a relatively rapid spontaneous firing rate, G F A P - I L 6 Purkinje neurons fired at a significantly lower frequency than those of control mice (Fig. 1B). Purkinje neurons of G F A P - I L 6 mice fired spontaneously at a mean rate of 3 6 + 2 Hz ( + S . E . M . ; n= 115) whereas control Purkinje neurons fired at a mean rate of 62 + 2 Hz (n= 130). A range of ages of both control and G F A P - I L 6 mice was used in this study in order to determine if changes in firing rate were dependent on the age of the mice. G F A P expression begins within the first week following birth in the mouse cerebellum] ° In addition, IL-6 m R N A has been detected in the brain of G F A P - I L 6 mice as early as seven days postnatal and its expression increases progressively with age, peaking at around three months postnatal. 9 Thus, the cerebellum of older G F A P - I L 6 mice would have been exposed to higher levels of IL-6, and for a longer period of time, than younger mice. F o u r age groups were studied: 31-34, 4 0 4 2 , 4 6 4 9 , and 70 73 days

T.E. Nelson et al.

130

A

Control

'IFII'I F71.rT'I'I'r?

A

IL-6 Oscillatory Purkinje Neuron Firing

2 mS

50 mS

B

Incidence of Oscillatory Firing

Purkinje Neuron FR

B

.

80

o .jL

~

15-

~ O..

S-

*

60 ~) 40 i~ u_

0

2O

Control C

IL-6

Purkinje Neuron FR by Age 80

6o 40 == iLL ~-

Control

IL-6

]Fig. 2. (A) Oscillatory firing patterns of two Purkinje neurons in slices taken from GFAP-IL6 mice. Oscillatory firing patterns varied from short periods of spontaneous simple spike activity (as in cell no. l) to much longer periods (as in cell no. 2). (B) Incidence of oscillatory firing activity in control and GFAP-IL6 slices. Twelve percent (n = 14/116) of GFAP-IL6 Purkinje neurons exhibited oscillatory simple spike firing patterns compared to only 1.5% (n=2/130) of control Purkinje neurons. The asterisk indicates a significant difference between GFAP-IL6 and control groups (P<0.001, chi-square).

2O 0

31-34

40-42 46-49 A g e G r o u p (days)

70-73

were found more frequently in GFAP-IL6 slices than Fig. 1. (A) Spontaneous activity of Purkinje neurons in in control slices (Fig. 2B). cerebellar slices taken from control and GFAP-IL6 mice. To the right are single action potentials viewed at a faster sweep speed. There were no significant changes in the Altered synaptic responses o f Purkinje neurons in glial appearance of action potentials between the two strains of fibrillary acidic protein-interleukin-6 transgenic mice mice. (B) Mean spontaneous firing rate was significantly Electrical stimulation of the white matter usually reduced in the GFAP-IL6 Purkinje neurons compared to controls (P<0.0001, unpaired t-test). (C) Mean spontaneous resulted in a large excitatory synaptic response by firing rate of Purkinje neurons at different age groups. The Purkinje neurons in both GFAP-IL6 and control asterisks indicate a significant difference between GFAPIL6 and control in each age group (P<0.0001, two-factor slices. This response, known as the climbing fiber response, is a classical synaptic response of Purkinje ANOVA). neurons and is caused by the activation of afferents (i.e. climbing fibers) that originate in vivo in the inferior olivary nucleus. Because the position of the postnatal. In each age group the transgenic GFAP- stimulating electrode in the white matter was critical IL6 mice exhibited a significantly reduced mean firing for activating climbing fibers and many of these rate compared to age-matched controls (Fig. 1C). afferents were likely cut during the preparation of the However, using regression analysis, no correlation slices, not all Purkinje neurons exhibited climbing between age and firing rate was observed for either fiber responses. Thus, although eliciting a climbing control (R2=0.001) or GFAP-IL6 (R2=0.037) fiber response helped to confirm that the recorded neurons within these age groups. neuron was a Purkinje neuron it was not used as the In addition to the alteration of the spontaneous definitive criterion for identifying Purkinje neurons in firing rate, there was a difference in the pattern of this study. There was no difference in the percentage spontaneous simple spike activity between control of cells exhibiting climbing fiber responses during and GFAP-IL6 Purkinje neurons. Purkinje neurons white matter stimulation between control (67178, were occasionally observed to exhibit oscillatory pat86%) and GFAP-IL6 slices (58/68, 85%). terns of spontaneous firing. Such oscillatory patterns In both GFAP-IL6 and control slices, the climbing were characterized by periods of normal firing inter- fiber response consisted of an excitatory phase, rupted by short periods of quiescence (Fig. 2A). featuring a burst of three to five spikes (termed a Purkinje neurons exhibiting oscillatory firing patterns complex spike) (Fig. 3A), which was often followed

Purkinje neurons from interleukin-6 transgenic mice

A Con,ro,

GFAP-IL6

B Control

I

50 mS

GFAP-IL6

C

Climbing Fiber Pause 300 "~ 250, ,~ 200, c: 150 O "~ 100 50' 0

Control

I L-6

Fig. 3. (A) Electrical stimulation of climbing fibers elicited complex spikes in Purkinje neurons of control and GFAPIL6 slices. Responses were similar in control and transgenic slices. Arrows indicate the stimulus artifact. (B) Inhibition of simple spike activity (i.e. climbing fiber pause) shown in representative traces from control and GFAP-IL6 Purkinje neurons. Arrows indicate the stimulus artifact. (C) The duration of the climbing fiber pause (interspike interval subtracted) was significantly longer in GFAP-IL6 Purkinje neurons. The asterisk indicates a significant difference between GFAP-IL6 and control groups (P=0.003, MannWhitney U-test).

by a post-burst inhibitory period during which spontaneous simple spike firing was interrupted (Fig. 3B). The post-burst inhibitory period following the complex spike has been termed the "climbing fiber pause". 27'38 Studies of the climbing fiber response in other preparations have yielded varying reports on the duration of the pause, depending largely on the experimental preparation; 38 it can be very short, similar to the interspike interval during spontaneous simple spike firing, or very long, lasting several hundred milliseconds. In the present study, complex spikes were normally followed by a relatively long pause in simple spike firing in both control and GFAP-IL6 slices. Also, in previous studies the climbing fiber pause was often succeeded by either a facilitation or reduction of the simple spike

131

firing rate. In contrast, facilitation or reduction of the spontaneous simple spike firing rate following the climbing fiber pause rarely occurred in our study. The general appearance of complex spikes was similar in control and GFAP-IL6 slices (Fig. 3A). However, in GFAP-IL6 Purkinje neurons the climbing fiber pause was significantly longer than in control Purkinje neurons (Fig. 3B, C). In the present study, the mean climbing fiber pause of each Purkinje neuron was determined by measuring the time interval from the end of the complex spike to the subsequent simple spike from a random sample of seven to 10 climbing fiber responses. Purkinje neurons from GFAP-IL6 mice had a mean climbing fiber pause of 1 1 2 ± 2 3 m s (n=58) compared to 3 0 + 5 m s (n=67) for control mice (P<0.0001, Mann Whitney U-test). No correlation between the age of the mice and the length of the climbing fiber pause was observed for either control (R2=0.004) or GFAP-IL6 (R2=0.017) Purkinje neurons. A contributing factor to the difference in the duration of the climbing fiber pause between control and GFAP-IL6 Purkinje neurons is the longer interspike interval in the GFAP-IL6 Purkinje neurons, corresponding to the observed reduction in the spontaneous simple spike firing rate of GFAP-IL6 Purkinje neurons. To determine if the difference in the interspike interval could account for the difference in the duration of the climbing fiber pause, a second analysis of the data included only those Purkinje neurons exhibiting a climbing fiber pause that was longer than the interspike interval during spontaneous simple spike activity. In addition, for each neuron the mean interspike interval was subtracted from the mean climbing fiber pause and the difference between these measurements was used in the comparison of the duration of the climbing fiber pause. After selecting the data using this criterion, the mean difference between the duration of the climbing fiber pause and the interspike interval was 109± 30 ms (n=33) for GFAP-IL6 Purkinje neurons and 26 + 12 ms (n=27) for control neurons (Fig. 3C). Thus, the increase in the length of the climbing fiber pause in GFAP-IL6 Purkinje neurons does not appear to be due simply to the increased length of the interspike interval during spontaneous activity (i.e. reduced firing rate) relative to control Purkinje neurons. In addition to assessing the response to a single stimulus, we also examined the ability of Purkinje neurons to respond to repetitive stimuli. Purkinje neurons of both control and GFAP-IL6 mice responded to periods of repetitive climbing fiber stimulation (0.5-1 Hz) with an overall reduction in simple spike firing rate compared to baseline spontaneous activity (control, P<0.0001; GFAP-IL6, P=O.O002, paired t-test). The mean reduction in firing rate was similar in both control (12°/,,) and GFAP-IL6 (14%) Purkinje neurons.

T.E. Nelson et al.

132

Expression of interleukin-6 receptor precursor and gpl30 in control and glial fibrillary acidic proteininterleukin-6 mouse cerebellum Numerous cell types in the CNS express receptors for IL-6. 39"4° Thus, the effects observed in the current study could be due to either a direct action of IL-6 on Purkinje neurons or an indirect action mediated through other cell types affected by IL-6. As a first step to determine if a direct IL-6 action could be involved, we assessed the expression of the precursor proteins of the IL-6 receptor and its intracellular signaling subunit, gpl30, using immunohistochemical staining techniques. Purkinje neurons in both control and GFAP-IL6 mice exhibited intense immunostaining for the IL-6 receptor precursor (Fig. 4A, C). The somata and proximal dendrites of Purkinje neurons were clearly visible in the immunostained sections, as well as fibers within the cerebellar white matter which might represent the axons of Purkinje neurons. The intensity of staining was not quantified but appeared similar in the cerebella of control and transgenic mice. In comparison to the Purkinje neurons, the staining in the granule cell layer was faint, indicating that granule neurons express much lower levels of the IL-6 receptor. However, a few deeply-stained perikarya, which resemble glial cells, were also scattered throughout the granule cell layer. These cells were also present to a lesser degree in the molecular layer and in the white matter. In the presence of the IL-6 receptor antigenic peptide, little to no staining was visible, indicating that binding of the primary antibody was specific to the IL-6 receptor (Fig. 4B, D). Control and G F A P - I L 6 cerebella also showed immunostaining of similar intensity for the precursor to the intracellular signaling molecule, gpl30 (Fig. 4E, G). Again, the somata and proximal dendrites of Purkinje neurons showed the highest level of staining. In contrast to the IL-6 receptor staining, the granule cell layer contained a moderate level of gp 130 precursor immunostaining and most, if not all, of the granule neurons appeared to be stained. Neurons in the deep cerebellar nuclei were also immunoreactive for the gpl30 precursor protein (not shown). All immunoreactivity was abolished in the presence of the gpl30 antigenic peptide (Fig. 4F, H).

DISCUSSION

The results of our electrophysiological studies indicate that Purkinje neurons in cerebellar slices from

transgenic mice chronically expressing IL-6 within the CNS exhibit altered firing properties, including a reduction in the spontaneous firing rate, an increased incidence of oscillatory firing patterns, and a prolonged climbing fiber pause in response to climbing fiber activation. The GFAP-IL6 mice used in this study develop cerebellar neuropathology which is evident at one month postnatal and progresses with age throughout the life of the animal. 8 The expression of the IL-6 transgene begins early in postnatal development (postnatal day 7) and reaches a peak by about the third month of age. 9 At this age, a number of histopathological changes, such as loss of neurons and presynaptic terminals, as well as the occurrence of dendritic damage and gliosis, are observed in the brain. 9'2° In the present study, the alterations in the spontaneous activity of cerebellar Purkinje neurons were examined for age-dependent changes within a time period (31-73 days postnatal) during which neuropathologic changes might be expected to occur. However, the reduced spontaneous firing rate was evident at the earliest age studied and the magnitude did not progress with age. Several possible explanations could account for the apparent lack of agedependence of the physiological alteration, including: (i) the Purkinje neurons recorded in the GFAP-IL6 slices were selected based on their ability to fire spontaneously and thus might reflect only a subpopulation of neurons exhibiting less severe alterations in physiology than neurons that are no longer spontaneously active; (ii) the changes in spontaneous firing rate produced by IL-6 may be fully developed by the earliest age used in our study and, thus, represent a maximal effect of chronic IL-6 exposure on this physiological parameter; (iii) the histopathologic changes in the cerebellum might occur on a more protracted time-scale than can be detected electrophysiologically within the ~ 5 week range of ages used in our study. The GFAP-IL6 mice are known to express high levels of IL-6 in the cerebellum, particularly in Bergmann glial cells which are in close proximity to cerebellar Purkinje neurons, s Moreover, our immunohistochemical results reveal that Purkinje neurons express precursors to both the IL-6 receptor and its signal transduction subunit, gpl30, at high levels. Thus, the effects of elevated IL-6 expression on Purkinje neuron electrophysiology are potentially due to a direct action of IL-6 on Purkinje neurons and may not necessarily be mediated through other neuronal types or glial cells. Our results support a direct role for IL-6 in the pathophysiologic alterations that underlie

Fig. 4. Immunohistochemical detection of precursors to the IL-6 receptor (A, C), and the gpl30 signal transduction molecule (E, G) in the cerebellar cortex of control (A, E) and GFAP-IL6 (C, G) mice. Corresponding sections in the presence of the blocking peptide are shown in B, D, F and H. Only counterstained neurons were visible when the blocking peptides were used. Arrows indicate Purkinje neurons; arrowheads in A and C indicate IL-6 receptor-positive glia-like cells; GCL, granule cell layer. Scale bar=25 lam.

~o

G

134

T. E. Nelson et al.

cerebellar motor deficits in neurological disorders associated with elevated CNS levels of IL-6 such as AIDS dementia complex and Alzheimer's disease. Previous studies have reported that IL-6 receptor m R N A is localized in granule neurons but not in Purkinje neurons of the rat cerebellum. 39 In contrast, we found low levels of immunostaining for the IL-6 receptor in granule neurons and high levels in Purkinje neurons in the mouse. The reason for this discrepancy is not known, although a speciesspecific difference in the distribution of IL-6 receptor expression within the CNS cannot be ruled out. At this time, the mechanism by which IL-6 alters the function of Purkinje neurons is not known. While the intracellular mechanisms of IL-6 receptor activation have been elucidated in other types of cells, it is not known if these mechanisms are also utilized by neurons. In non-neural cell systems, IL-6 is known to induce tyrosine kinase activity. 29 Upon activation, the IL-6 receptor associates with the gpl30 subunit, facilitating dimerization of the protein and enabling it to activate tyrosine kinases. Tyrosine kinase activation has been proposed to be linked to the interaction of transcription factors with promoter regions of genes. Our present data show that Purkinje neurons express the gpl30 protein. Thus, the altered physiology of Purkinje neurons in GFAP-IL6 mice could be due to a change in the expression of proteins responsible for the spontaneous activity of Purkinje neurons. Spontaneous activity is known to be endogenously generated in Purkinje neurons and to involve a variety of ion channels, regulatory proteins, and second messengersfl 6'31"32 Similarly, synaptically-evoked responses in Purkinje neurons might also be sensitive to changes in the expression of a variety of proteins. The putative signaling mechanisms of IL-6 involve processes that act on a protracted time-scale (i.e. gene transcription). Thus, it is likely that IL-6's effects on Purkinje neurons are a result of chronic exposure and would not be expected to occur immediately following acute exposure of the cytokine. Currently, data on the acute effects of IL-6 on CNS neurons is sparse. In cultures of cerebellar neurons chronic, but not acute, application of pathophysiological doses of IL-6 (500U/ml) induces alterations of granule neuron physiology consistent with the need for longer periods (i.e. several days) of exposure to influence neuronal properties. 36 However, very high concentrations of IL-6 (2000-8000 U/ml) have been shown to reversibly alter the spontaneous firing rate of hypothalamic neurons when bath-applied to tissue slices, suggesting that under some conditions acute exposure of IL-6 can alter neuronal function. 4s The exact locus and underlying mechanisms of this acute effect of IL-6 are not known although production of prostaglandins and opioids appears to be involved. Chronic IL-6 exposure results in an increase in intracellular calcium both at rest and in response to excitatory neurotransmitters in cerebellar granule

neurons maintained in culture. 36 Many of the changes in Purkinje neuron activity observed in the GFAP-IL6 mice might also be explained by elevations of intracellular calcium levels. Higher resting levels of calcium might cause an enhanced activation of a Ca2+-dependent K + conductance, resulting in a hyperpolarizing shift in the membrane potential of the Purkinje neuron and, in turn, a reduction in the intrinsic spontaneous firing rate. In addition, the oscillatory pattern of spontaneous activity observed in some Purkinje neurons (particularly in GFAP-IL6 slices) has been proposed to be a partially Ca 2+dependent process 31'32 and might also be caused by fluctuations in membrane potential mediated by a Ca2+-dependent K + conductance. Finally, the increase in the length of the climbing fiber pause observed in GFAP-IL6 Purkinje neurons could reflect greater activation of a Ca2+-dependent K + conductance which is known to be responsible for the afterhyperpolarization that contributes to the climbing fiber pause. 25'31'32 Alterations of intracellular calcium homeostasis and signaling due to chronic exposure to IL-6 could also play a role in neurotoxicity of Purkinje neurons and, thus, might explain the loss of these neurons in the GFAP-IL6 mice.

CONCLUSION

In the present study we have provided evidence for alterations in Purkinje neuron function caused by chronic IL-6 exposure. Such alterations in Purkinje neuron physiology would likely cause serious alterations in the flow of information through the cerebellum. For example, the reduction of spontaneous firing rate would likely result in a decrease in the inhibitory effects of Purkinje neurons on cells of the cerebellar nuclei, thus changing the output of the cerebellum to descending motor systems. In addition, disruption of the normal response of Purkinje neurons to climbing fiber inputs might interfere with long-term depression (LTD) at the parallel fibe~ Purkinje neuron synapse. In the cerebellum, LTD is a form of synaptic plasticity in which the response of Purkinje neurons to parallel fiber input is depressed following depolarization and influx of calcium into the neuron.12'28 Such disturbances of normal cerebellar function might help to explain the symptoms of discoordination of movement and ataxia which are common to both the GFAP-IL6 transgenic mice and a number of neurological disorders. In addition, structural changes in the cerebella of GFAP-IL6 mice are likely to contribute significantly to the disturbances in cerebellar function observed in this model. The authors wish to thank Ms Carrie Kincaid for her technical assistance with the immunohistochemistry, Dr Jeffrey Netzeband for his comments on the manuscript, and Ms Floriska Chizer for her secretarial assistance. This work was supported by MH47680, AA07456, DA10187. Acknowledgements

Purkinje neurons from interleukin-6 transgenic mice

135

REFERENCES 1. Aikawa H., Suzuki K., Iwasaki Y. and Iizuka R. (1985) Atypical Alzheimer's disease with spastic paresis and ataxia. Ann. Neurol. 17, 297-300. 2. Akira S., Hirano T., Taga T. and Kishimoto T. (1990) Biology of multifunctional cytokines: IL-6 and related molecules (IL-1 and TNF). Fedn Am. Soes exp. Biol. J. 4, 2860--2867. 3. Bauer J., Strauss S., Schreiter-Gasser U., Ganter U., Schlegel P., Witt I., Yolk B. and Berger M. (1991) Interleukin-6 and c¢-2-macroglobulin indicate an acute-phase state in Alzheimer's disease cortices. Fedn Eur. bioehem. Sots Left. 285, 1ll 114. 4. Bellinger F. P., Madamba S. G., Campbell I. L. and Siggins G. R. (1995) Reduced long-term potentiation in the dentate gyrus of transgenic mice with cerebral overexpression of interleukin-6. Neurosci. Lett. 198, 95-98. 5. Benveniste E. N. (1992) Inflammatory cytokines within the central nervous system: sources, function, and mechanism of action. Am. ~ Physiol. 263, CI C16. 6. Benveniste E. N., Sparacio S. M., Norris J. G., Grenett H. E. and Fuller G. M. (1990) Induction and regulation of interleukin-6 gene expression in rat astrocytes. J. Neuroimmunol. 30, 212 301. 7. Brett F. M., Mizisin A. P., Powell H. C. and Campbell I. L. (1996) Evolution of neuropathologic abnormalities associated with blood brain barrier breakdown in transgenic mice expressing interleukin-6 in astrocytes. J. Neuropath. exp. Neurol. 546, 766-775. 8. Campbell I. L., Abraham C. R., Masliah E., Kemper P., Inglis J. D., Oldstone M. B. and Mucke L. (1993) Neurologic disease induced in transgenic mice by cerebral overexpression of interleukin 6. Proc. natn. Acad. Sei. U.S.A. 90, 10,061 10,065. 9. Chiang C.-S., Stalder A., Samimi A. and Campbell I. L. (1994) Reactive gliosis as a consequence of interleukin-6 expression in the brain: studies in transgenic mice. Devl Neurosei. 16, 212 221. 10` Chiu F. C. and Goldman J. E. (1985) Regulation of glial fibrillary acidic protein (GFAP) expression in CNS development and in pathological states. J. Neuroimmunol. 8, 283 292. 1l. Davie C. A., Barker G. J., Webb S., Tofts P. S., Thompson A. J., Harding A. E. and Miller D. H. (1995) Persistent functional deficit in multiple sclerosis and autosomal dominant cerebellar ataxia is associated with axon loss. Brain 118, 1583-1592. 12. De Schutter E. (1995) Cerebellar long-term depression might normalize excitation of Purkinje cells: a hypothesis. Trends Neurosci. 18, 291 295. 13. Eccles J. C., Ito M. and Szentagothai J. (1967) The Cerebellum as a Neuronal Machine. pp. 306. Springer+ Verlag, New York, Heidelberg. 14. Frei K., Malipiero U. V., Leist T. P., Zinkernagel R. M., Schwab M. E. and Fontana A. (1989) On the cellular source and function of interleukin 6 produced in the central nervous system in viral diseases. Eur. J. hmnunol. 19, 689-694. 15. Freund R. K., Wang Y. and Palmer M. R. (1993) Differential effects of ethanol on the firing rates of Golgi-like neurons and Purkinje neurons in cerebellar slices in vitro. Neurosci. Lett. 164, 9-12. 16. Gahwiler B. H., Mamoon A. M. and Tobias C. A. (1973) Spontaneous bioetectric activity of cultured cerebellar Purkinje ceils during exposure to agents which prevent synaptic transmission. Brain Res. 53, 71-79. 17. Gallo P., Frei K., Rordorf C., Lazdins J., Tavolato B. and Fontana A. (1989) Human immunodeficiency type I (HIV-1) infection of the central nervous system: an evaluation of cytokines in cerebrospinal fluid. ,L Neuroimmunol. 23, 109-116. 18. Graus F., Ribalta T., Abos J., Alom J., Cruz-Sanchez F., Mallolas J., Miro J. M., Cardesa A. and Tolosa E. (1990) Subacute cerebellar syndrome as the first manifestation of AIDS dementia complex. Acta neurol, stand 81, 118.... 120. 19. Heinrich P. C., Castell J. V. and Andus T. (1990) Interleukin-6 and the acute phase response. Biochem. J. 265, 621 636. 20. Heyser C. J., Masliah E., Samimi A., Campbell I. L. and Gold L. H. (1997) Progressive decline in avoidance learning paralleled by inflammatory neurodegeneration in transgenic mice expressing interleukin 6 in the brain. Proe. natn. Ac'aa! Sci. U.S.A. 94, 1500-1505. 21. Hirano T. (1992) The biology ofinterleukin-6. Chem. Immunol. 51, 153-180. 22. Hirohata S. and Miyamoto T. (1990) Elevated levels ofinterleukin-6 in cerebrospinal fluid from patients with systemic lupus erythematosus and central nervous system involvement. Arthr. Rheum. 33, 644-649. 23. Hopkins S. J. and Rothwell N. J. (1995) Cytokines and the nervous system. I: expression and recognition. Trends Neurosci. 18, 83-88. 24. Hounsgaard H. and Midtgaard J. (1988) Intrinsic determinants of firing pattern in Purkinje cells of the turtle cerebellum in vitro. J. Physiol. 402, 731 749. 25. Hounsgaard H. and Midtgaard J. (1989) Synaptic control of excitability in turtle cerebellar Purkinje cells. Z Physiol. 409, 157 170. 26. Houssiau F. A., Bukasa K., Sindic C. J. M., Van Damme J. and Van Snick J. (1988) Elevated levels of the 26K human hybridoma growth factor (interleukin 6) in cerebrospinal fluid of patients with acute infection of the central nervous system. Clin. exp. lmmunol. 71,320-323. 27. lto M. (1984) The Cerebellum and Neural Control, pp. 580. Raven, New York. 28. Ito M. (1989) Long-term depression. A. Rev. Neurosci. 12, 85-102. 29. Kishimoto T., Akira S., Narazaki M. and Taga T. (1995) Interleukin-6 family of cytokines and gpl30. Blood 86, 1243--1254. 30. Latham A. and Paul D. H. (1971) Spontaneous activity of cerebellar Purkinje cells and their responses to impulses in climbing fibers. J. Physiol. 213, 135-156. 31. Llinas R. and Sugimori M. (1980) Electrophysiological properties of in vitro Purkinje cell somata in mammalian cerebellar slices. J. Physiol. 305, 171 195. 32. Llinas R. and Sugimori M. (1980) Electrophysiological properties of in vitro Purkinje cell dendrites in mammalian cerebellar slices. J. Physiol. 305, 197 213. 33. Maimone D., Gregory S., Arnason B. G. W. and Reder A. T. (1991) Cytokine levels in the cerebrospinal fluid and serum of patients with multiple sclerosis. J. Neuroimmunol. 32, 67 74.

136 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48.

T.E. Nelson et al. Merrill J. E. (1992) Tumor necrosis alpha, interleukin 1, and related cytokines in brain development: normal and pathological. Devl Neurosci. 14, 1 10. Poser S., Luer W., Eichenlaub D., Pohle H. D., Weber T. and Jurgens S. (1988) Chronic HIV encephalitis--II. Clinical aspects. Klin. Wschr. 66, 26-31. Qiu Z., Parsons K. L. and Gruol D. L. (1995) Interleukin-6 selectively enhances the intracellular calcium response to N M D A in developing CNS neurons. J. Neurosci. 15, 6688-6699. Rothwell N. J. (1991) Functions and mechanisms of interleukin 1 in the brain. Trends pharmac. Sci. 12, 430M36. Sato Y., Miura A., Fushiki H. and Kawasaki T. (1992) Short-term modulation ofcerebellar Purkinje cell activity after spontaneous climbing fiber input. J. Neurophysiol. 68, 2051-2062. Sehobitz B., De Kloet E. R., Sutanto W. and Holsboer F. (1993) Cellular localization of interleukin-6 mRNA and interleukin-6 receptor mRNA in rat brain. Eur. J. Neurosci. 5, 1426-1435. Schobitz B., Voorhuis D. A. M. and De Kloet E. R. (1992) Localization of interleukin 6 mRNA and interleukin 6 receptor mRNA in rat brain. Neurosci. Lett. 136, 189-192. Singh R. R., Prasad K., Kumar A., Misra A., Padmakumar K. and Malaviya A. N. (1988) Cerebellar ataxia in systemic lupus erythematosus: three case reports. Ann. Rheum. Dis. 47, 954-956. Steffensen S. C., Campbell I. L. and Henriksen S. J. (1994) Site-specific hippocampal pathophysiology due to cerebral overexpression of interleukin-6 in transgenic mice. Brain Res. 652, 149-153. Tuchman A. J., Daras M. and David S. (1983) Cerebellar ataxia in systemic lupus erythematosus. N. Y. State J. Med. 83, 983-984. Tyor W. R., Glass J. D., Griffin J. W., Becker P. S., McArthur J. C., Bezman L. and Griffin D. E. (1992) Cytokine expression in the brain during the acquired immunodeficiency syndrome. Ann. Neurol. 31, 349-360. Vakili S. T. and Muller J. (1987) Juvenile Alzheimer's disease with cerebellar involvement. Archs Path. lab. Med. 111, 480482. Vergani M. I., Reimao R., Silva A. M., Muskat M., Esposito S. and Diament A. (1988) Multiple sclerosis with early childhood onset. A case report. Arq. Neuropsiquiatr. 46, 195-197. Woodroofe M. N., Sarna G. S., Wadhwa M., Hayes G. M., Loughlin A. J., Tinker A. and Cuzner M. L. (1991) Detection of interleukin-1 and interleukin-6 in adult rat brain, following mechanical injury, by in vivo microdialysis: evidence of a role of microglia in cytokine production. J. Neuroimmunol. 33, 227 236. Xin L. and Blatteis C. M. (1992) Hypothalamic neuronal responses to interleukin-6 in tissue slices: effects of indomethacin and naloxone. Brain Res. Bull. 29, 27 35. (Accepted 11 May 1998)