In vitro phosphorylation of medial vestibular nucleus and prepositus hypoglossi proteins during behavioural recovery from unilateral vestibular deafferentation in the guinea pig

Brain Research 778 Ž1997. 166–177

Research report

In vitro phosphorylation of medial vestibular nucleus and prepositus hypoglossi proteins during behavioural recovery from unilateral vestibular deafferentation in the guinea pig Andrew J. Sansom

a,b

, Vicki A. Brent c , Paula E. Jarvie c , Cynthia L. Darlington a , Paul F. Smith Richard Laverty b , John A.P. Rostas c

b, )

,

a

b

Department of Psychology, Neuroscience Research Centre, UniÕersity of Otago, Dunedin, New Zealand Department of Pharmacology, School of Medical Sciences, Otago Medical School and Neuroscience Research Centre, UniÕersity of Otago, P.O. Box 913, Dunedin, New Zealand c The Neuroscience Group, Faculty of Medicine and Health Sciences, UniÕersity of Newcastle, Newcastle, NSW 2308, Australia Accepted 19 August 1997

Abstract Unilateral removal of vestibular nerve input to the vestibular nuclei Že.g. by unilateral labyrinthectomy, UL. results in severe ocular motor and postural disorders which disappear over time Žvestibular compensation.. We investigated whether recovery of ocular motor function is temporally correlated with changes in protein phosphorylation in the medial vestibular nucleus ŽMVN. and prepositus hypoglossi ŽPH; MVNrPH. in vitro. Bilateral MVNrPH were dissected from 48 guinea pigs following decapitation at 10 h, 53 h or 2 weeks post-UL, or -sham operation and frozen. Tissue extracts were incubated with w g- 32 PxATP " Ca2q plus phorbol 12,13-dibutyrate and phosphatidylserine. UL resulted in a significant bilateral increase in the 32 P-incorporation into a 65–85 kDa band Žprobably the myristoylated alanine-rich C kinase substrate, MARCKS. in compensated animals Ž53 h post-UL. under conditions which favoured the activation of protein kinase C. Under identical conditions, the labelling of a 42–49 kDa protein ŽP46. was increased significantly in the bilateral MVNrPH between either 10 h or 53 h and 2 weeks post-UL; there were no significant changes over time in sham controls. These results show that later stages of vestibular compensation are accompanied by changes in the phosphorylation of several likely protein kinase C substrates in the MVNrPH in vitro. q 1997 Elsevier Science B.V. Keywords: Vestibular compensation; Medial vestibular nuclei; Guinea pig; Neural plasticity; Protein phosphorylation; Protein kinase C

1. Introduction The central nervous system plasticity which follows unilateral removal of primary afferent input to the vestibular nuclei Že.g. by unilateral labyrinthectomy, UL. is known as vestibular compensation Žsee w42,8,11x, for reviews.. The vestibular system is an ideal system in which to investigate the cellular mechanisms of behavioural change because the neural pathways which generate the behaviour Ži.e. vestibular reflexes. are both anatomically and electrophysiologically well characterised w51x. In addition, individual behavioural symptoms of UL can be reliably and easily quantified w19x. The vestibular nuclei are thought to play a critical role in compensation since they mediate

)

Corresponding author. Fax: q64 Ž3 . 479-9140; E-mail: [email protected] 0006-8993r97r$17.00 q 1997 Elsevier Science B.V. All rights reserved. PII S 0 0 0 6 - 8 9 9 3 Ž 9 7 . 0 1 0 5 9 - 7

normal vestibular ocular motor and postural reflexes w51x and lesions to the vestibular nuclei prevent the development of compensation w46x. Furthermore, the disappearance of the static behavioural symptoms of UL Ži.e. symptoms which persist in the absence of head movement. are closely correlated with the recovery of resting activity in type I medial vestibular nucleus ŽMVN. neurons ipsilateral to the UL w41,32,35x. It has been suggested that following UL, remaining inputs to the MVN on the ipsilateral side undergo a change in synaptic efficacy which may replace the lost VIIIth nerve input Žw15x; see w44,8,11x, for reviews.. Alternatively, it is possible that compensation may be achieved through a change in the intrinsic properties of vestibular nucleus neurons themselves Žthe ‘intrinsic mechanism hypothesis’, proposed by Darlington and Smith w9x.. At present there is little empirical data to support these hypotheses. However, either mechanism could involve a change in second mes-

A.J. Sansom et al.r Brain Research 778 (1997) 166–177

senger levels which would be reflected by a change in the phosphorylation state of one or more proteins in the vestibular nucleus. Changes in the phosphorylation of specific proteins are correlated with several types of neural plasticity; for example, axonal sprouting following axotomy and long term potentiation Žsee w49x, for a review.. Results from a number of studies suggest that Ca2q-dependent enzymes w39x, N-methyl-D-aspartate ŽNMDA. receptors Žwhose ion channels are permeable to Ca2q w38,12,34,28,29x., and NMDA receptor sub-unit mRNA w13,37x in the vestibular nuclei, undergo changes during compensation. A large body of evidence now exists which suggests that the function of NMDA receptor ion channels and Ca2q-dependent enzymes can be regulated directly and indirectly by protein phosphorylation and dephosphorylation Žsee w49x, for a review.. In the only investigation of its type to date, Flohr et al. w16x and Janssen et al. w23x showed that the early stages of compensation in frogs are associated with a change in the in vitro phosphorylation of a number of endogenous proteins in whole brain homogenates. In particular, the phosphorylation of four proteins with apparent molecular weights of 20, 45, 52 and 58 kDa was increased compared to control frogs Ži.e. labyrinthine-intact. at 2, 4, 7 and 12 days post-UL. The 20 and 45 kDa phosphoproteins from control frog brains were shown to be regulated by an endogenous Ca2qrcalmodulin-dependent protein kinase ŽCaMK. and protein kinase C, respectively w24x. However, since these studies were carried out using whole brain homogenates, it is unclear whether the phosphorylation changes involve the vestibular nuclei. Furthermore, since there were no age-matched controls in these experiments, it is not possible to determine whether the observed phosphorylation changes are related to vestibular compensation or some other time-dependent process Žfor example, brain aging.. The aim of the present study was firstly to characterise the major endogenous phosphoproteins and their kinases in the MVNrprepositus hypoglossi ŽPH. and secondly, determine whether the pattern of phosphorylation is changed following UL. The protein kinases involved in such changes were determined by comparing phosphorylation patterns from UL- and sham-operated guinea pigs following the addition of various protein kinase activators. To our knowledge, this is the first investigation to characterise protein phosphorylation patterns in the vestibular nucleus ŽMVNrPH. following UL.

2. Materials and methods 2.1. Materials The following compounds were used in this study: Tris Žhydroxymethyl.methylamine ŽTris–HCl, BDH., ethylenediaminetetraacetic acid ŽEDTA, BDH., leupeptin, phenyl-

167

methyl-sulfonyl fluoride ŽPMSF, Sigma., 1,4-dithiothreitol ŽDTT, Boehringer., sodium molybdate ŽBDH., tetrasodium pyrophosphate ŽSigma., microcystin, ethylene glycol-bis Ž b-aminoethyl ether. N, N, N X , N X-tetraacetic acid ŽEGTA, Sigma., w g- 32 Px adenosine triphosphate ŽATP, G 3,000 Cirmmole, Amersham., cyclic AMP, purified calmodulin ŽDiscipline of Medical Biochemistry, University of Newcastle, NSW., phosphatidylserine ŽPS. and phorbol 12,13dibutyrate ŽPBD.. Fentazin anaesthetic contains: 0.4 mgrml fentanyl citrate, 58.3 mgrml xylazine HCl and 3.2 mgrml azaperone ŽParnell Laboratories, Auckland.. 2.2. Surgical procedures and post-operatiÕe measurements Forty eight adult male and female pigmented guinea pigs Ž CaÕia porcellus; 180–600 g. were used in these experiments. Twenty four guinea pigs received a right surgical UL Žsee w19x, for details.. Briefly, guinea pigs were anaesthetised with fentazin Ž0.4 mlrkg, i.m. w40x., the bony labyrinth exposed, and under microscopic control the horizontal, anterior and posterior canal ampullae and otoliths were destroyed with a dental drill and their contents aspirated. The temporal bone cavity was sealed with dental acrylic. The remaining 24 animals received a sham operation which consisted of exposing the right bony labyrinth only Žas above.. Animals recovered in light and were given food and water ad libitum. Spontaneous nystagmus was quantified in all animals in order to verify the state of ocular motor compensation at 10 h post-UL andror immediately pre-decapitation w19x. Previous studies, using temporal bone histology, have shown that the UL technique described above results in the complete destruction of vestibular hair cells within the operated labyrinth w7x. When correctly performed, a surgical UL does not result in damage to the brainstem itself w7,39,10x. All surgical procedures were approved by the University of Otago Committee on Ethics in the Care and Use of Laboratory Animals. 2.3. Tissue preparation Guinea pigs were decapitated following cervical dislocation at either 10 h ŽUL, n s 9; sham, n s 9., 53 h ŽUL, n s 9; sham, n s 9. or 2 weeks ŽUL, n s 6; sham, n s 6. following UL or sham operation. Bilateral MVNrPH were dissected from each brainstem in addition to a small sample of occipital cortex Žapprox. 90 mg wet weight. using the following procedure. A coronal brainstem slice Ž2–3 mm thick. containing the rostro-caudal extent of the MVN was cut from the brainstem following removal of the cerebellum. Contralateral Žcontra-. and ipsilateral Žipsi-. MVNrPH Ž10–15 mg each wet weight. to the UL were dissected from the slice using a razor blade, frozen in liquid nitrogen and stored at y808C. Each MVNrPH sample also contained small amounts of lateral vestibular nuclei, descending vestibular nuclei and medial longitudi-

168

A.J. Sansom et al.r Brain Research 778 (1997) 166–177

nal fasciculus. The total time between decapitation and freezing of MVNrPH samples ranged between 2 and 4 min. 2.4. Subcellular fractionation and protein phosphorylation in Õitro Tissue samples from three guinea pigs were pooled for each experiment and homogenised Ž10% homogenate. in an ice-cold solution containing: 0.32 M sucrose, 1 mM EDTA, 20 mgrml leupeptin, 50 mM PMSF, 0.4 mM DTT, 0.08 mM molybdate, 2 mM pyrophosphate, and 2.5 mM microcystin. Homogenates were centrifuged at 1000 g Ž48C. for 10 min, the supernatant S1 fractions removed and their protein concentrations determined by the Lowry method. S1 fractions were diluted to 1 mgrml protein with ice-cold lysis buffer Ž30 mM Tris–HCl, 1 mM EGTA, 20 mgrml leupeptin, 50 mM PMSF, 0.4 mM DTT, 0.08 mM molybdate, 2 mM pyrophosphate, and 2.5 mM microcystin.. Phosphorylation reactions were initiated by the addition of ice-cold protein Ž100 ml @ 1 mgrml. to a reaction buffer containing the following Žat final concentrations.: 30 mM Tris–HCl ŽpH 7.4.; 1 mM MgSO4 Žfor cAMP- and Ca2qrcalmodulin-stimulated phosphorylation., or 5 mM MgSO4 Žfor Ca2q, PS and PDB-stimulated phosphorylation.; 1 mM EGTA; 40 mM ATP Ž50 mCirml.. In addition, either 50 mM cAMP; or 1.2 mM CaCl 2 plus 100 mgrml calmodulin; or 1.2 mM CaCl 2 plus 50 mgrml PS and 100 ngrml PDB were added to basal incubation conditions. The final protein concentration in each reaction solution was 0.5 mgrml Žin a 200 ml volume.. Phosphorylation reactions were carried out at 308C for 20 s and terminated by the addition of 100 ml sodium dodecyl sulphate-ŽSDS. stop buffer Žcontaining 12% mercaptoethanol. before being boiled for 2 min. Under these conditions, the labelling of bands was linear with respect to time and protein. The optimal assay concentration for Mg 2q was determined in preliminary experiments Ždata not shown.. 2.5. Gel electrophoresis and autoradiography Protein mixtures were separated by one dimensional SDS-polyacrylamide gel electrophoresis ŽPAGE. on 9– 16% linear gradient gels Žapproximately 60 mg protein per track. and stained with coomassie blue. Phosphoproteins were detected using autoradiography by exposing dried gels to X-ray film ŽKodak.. For densitometry, only autoradiographs that were within the linear range of the emulsion were used. The level of 32 Pi-incorporation into individual bands was quantified by optical densitometry and measured in optical density units. All phosphorylation data are mean ratios Žphosphorylation in the presence of protein kinase activators% basal phosphorylation. expressed as a percentage of basal from three independent experiments. Phosphorylation data were corrected for minor differences

in protein loadings between tracks on gels Žthe total protein in each track was determined by the area under the curve.. Bands are referred to as PXX, where XX is the mean apparent molecular weight of the band estimated from protein standards on each gel by linear regression. It should be noted that some phosphoprotein bands from the 2 week condition ŽP22 and P18. were unreadable and therefore could not be presented. 2.6. Statistical analysis Only bands within the same autoradiograph were subjected to statistical analyses in order to minimise possible differences in optical density between X-ray films and acrylamide densities. Phosphorylation ratios were analysed statistically using a one factor ANOVA Ž a s 0.05.. Posthoc comparisons were carried out where appropriate using Scheffe F-tests Ž a s 0.05; w30x..

3. Results 3.1. Choice of post-operatiÕe times and behaÕioural obserÕations There is a large body of evidence to show that 10 h, 53 h and 2 weeks post-UL represent critical stages of static compensation in the guinea pig Žsee w42,44,8,11x, for reviews.. At 10 h post-UL animals have recovered from fentazin anaesthesia sufficiently to display severe ocular motor and postural symptoms Žthe uncompensated stage.. These symptoms disappear rapidly over the next 40 h or so, such that few symptoms are visible by 50 h post-UL Žthe acute compensated stage.. We chose the 2 week post-UL time point Žthe chronic compensated stage. since there is evidence that further neurochemical changes take place between the acute and chronic compensated stages w43,17,28x. The static postural and ocular motor symptoms of UL and their pattern of compensation in guinea pig have been described extensively elsewhere Žw7x; see w42,8x, for reviews.. We quantified levels of spontaneous nystagmus for each group of animals in order to verify the behavioural state of compensation immediately prior to decapitation. Briefly, at 10 h post-UL, all guinea pigs displayed high frequency spontaneous nystagmus and severe postural symptoms. By 53 h these symptoms had virtually disappeared as a result of compensation. The behaviour of animals at 2 weeks post-UL was difficult to distinguish from that of sham-operated, or labyrinthine-intact animals. It should be noted that the magnitude of behavioural symptoms for individual animals within a group was similar at each time period and therefore brain tissue from animals within groups was pooled Žsee Methods.. UL-like symptoms were never observed in sham-operated guinea pigs at any post-operative time. The lack of behavioural

A.J. Sansom et al.r Brain Research 778 (1997) 166–177

169

Fig. 1. Autoradiographs ŽA, C and E. and coomassie stained gradient gels ŽB, D and F. showing the pattern of endogenous phosphoproteins and proteins in the contra-ŽC. and ipsi-ŽI. MVNrPH Žtracks 1–8. and cortex Žtracks 9–12. at 10 h ŽA and B., 53 h ŽC and D. and 2 weeks ŽE and F. post-UL ŽUL. or -sham ŽSh.. S1 fractions of MVNrPH or cortex were incubated in w g- 32 PxATP " 1.2 mM CaCl 2 plus 50 mgrml PS and 100 ngrml PDB.

170

A.J. Sansom et al.r Brain Research 778 (1997) 166–177

symptoms seen in sham animals strongly suggests that the sham operation did not damage the vestibular labyrinth, since even superficial damage to the vestibular apparatus results in transient UL-like symptoms Žw7x Sansom et al., unpubl. obs... 3.2. Design of gels Aliquots of labelled ipsi- and contra-MVNrPH were arranged on gels to permit two types of within-gel comparison: Ži. the effect of UL ŽUL vs. sham at each of three post-operative times.; and Žii. the effect of compensation Ž10 h vs. 53 h vs. 2 weeks post-UL or -sham.. Samples of occipital cortex were also run on some gels as internal markers to facilitate the identification of MVNrPH bands. We identified the major occipital cortex phosphoproteins by a comparison of their apparent molecular weights and phosphorylation characteristics Žin the presence of kinase activators. with those of a previous phosphorylation study of guinea pig cortex w25x. The following phosphoprotein bands were tentatively identified: the synapsins 1a and 1b Ž81 and 75 kDa respectively., the myristoylated alanine-rich C kinase substrate ŽMARCKS, which often migrated with synapsin 1b; 65–85 kDa., a Ž50 kDa. and brb X subunits of CaMK II Ž63 and 61 kDa, respectively.. 3.3. Characterisation of phosphoproteins in the MVN r PH Preliminary results with occipital cortex showed that frozen samples could be used for the analysis of in vitro protein phosphorylation patterns. Although freezing induced some changes in the relative degree of phosphoprotein labelling, the overall pattern of labelling was similar between frozen and fresh samples. In particular, cAMP-, Ca2qrcalmodulin- and Ca2qrPSrPDB-stimulated activity could still be detected in frozen samples and major substrate proteins were all labelled in a similar manner Ždata not shown.. In frozen MVNrPH samples, the phosphorylation of a number of bands increased in the presence of Ca2q alone, Ca2q plus calmodulin, or Ca2q plus PS and PDB, indicating the presence of endogenous Ca2qrcalmodulin- and Ca2qrphospholipid-dependent protein kinases in this tissue. In contrast, there was little or no cAMP-stimulated activity in the MVNrPH, even though frozen cortex samples from the same animals showed obvious cAMP-stimulated activity Ždata not shown.. Although protein patterns from cortex and MVNrPH samples were similar, the corresponding phosphoprotein patterns were markedly different Žsee Fig. 1.. The MVNrPH contained fewer phosphoproteins, with most showing lower levels of basal phosphorylation compared to their cortex counterparts. The most heavily and consistently labelled phosphoprotein bands in MVNrPH samples were: a 65 to 85 kDa band ŽP75; see below., a 42–49 kDa band ŽP46. which was not present in cortex tracks, and two bands whose apparent molecular weights were 20–25 kDa ŽP22. and 17–21 kDa ŽP18..

The following evidence suggests that MVNrPH P75 is probably the MARCKS phosphoprotein. First, the apparent molecular weight of P75 on SDS-PAGE Ž65–85 kDa. was within the range published for the MARCKS Ž68–87 kDa, see w4x for a review.. Second, P75 co-migrated with a band previously identified in guinea pig cortex as the MARCKS w25x and its phosphorylation characteristics were identical to those of the MARCKS; i.e. markedly increased in the presence of Ca2q alone, or the protein kinase C activators Ca2q plus PS and PDB, but strongly inhibited in the presence of calmodulin w1x. Third, P75 had a phosphopeptide map identical to that of cortex MARCKS Ždata not shown.. Fourth, the presence of MARCKS immunoreactivity in the rat MVN and PH has been demonstrated w33x. However, further characterisation of this band is necessary to provide definitive evidence that P75 is MARCKS; therefore, P75 will be referred to as the MARCKS-like protein throughout the remainder of this paper. The identity of P46 Ž42–49 kDa. is unknown. P46 phosphorylation was markedly stimulated by the addition of either Ca2q, or Ca2qrPSrPDB, or the CaMK activators Ca2qrcalmodulin to the incubation medium. Although MVNrPH P46 had a similar apparent molecular weight to occipital cortex a-CaMKII, it did not co-migrate with this band. P46 also has a similar apparent molecular weight on SDS-PAGE to that of rat growth associated protein ŽGAP43; 43–57 kDa; w6x.. Although GAP-43 mRNA is present in high concentrations in the rat MVN and PH w31x, guinea pig P46 is unlikely to be the GAP-43 for the following reasons: Ži. it was not a diffuse band, Žii. it was absent from the occipital cortex, and Žiii. its phosphorylation was not inhibited by calmodulin Ždata not shown.. However, these may reflect species differences since guinea pig brain does not have a phosphoprotein that displays the same in vitro phosphorylation characteristics as rat brain GAP-43 w25x. The identity of P22 Ž20–25 kDa. is also unknown. P18 Ž17–21 kDa. was the most heavily labelled protein in both MVNrPH and cortex; its phosphorylation was both Ca2qrcalmodulin- and Ca2qrPSrPDB-dependent and its apparent molecular weight on PAGE gels is similar to that of myelin basic protein Žapproximately 18 kDa; w47x.. Therefore, it is possible that P18 represents one or more species of myelin basic protein. In addition to these major phosphoproteins, two minor bands were identified which correspond to the neuronal phosphoproteins synapsin 1b and b-CaMKII. Synapsin 1b Ž75 kDa. was identified on the basis of its molecular weight on PAGE and its phosphorylation characteristics in the presence of Ca2q plus calmodulin or PSrPDB. bCaMKII Ž63 kDa. was identified on the basis of its molecular weight on PAGE, its phosphorylation characteristics in the presence of Ca2qrcalmodulin and 5 mM ZnSO4 Ža specific marker for CaMKII subunit autophosphorylation w25x; results not shown.. b-CaMKII was not always visible on 9–16% gradient gels. Interestingly, there were no bands in the MVNrPH with similar molecular weight or phos-

A.J. Sansom et al.r Brain Research 778 (1997) 166–177

phorylation characteristics to that of a-CaMKII Ža 50 kDa band w25x; see Fig. 1.. 3.4. Ca 2 qr PS r PDB-stimulated phosphorylation in the MVN r PH following UL Results of preliminary experiments indicated there were no obvious changes in the levels of Ca2qrcalmodulin-

171

stimulated phosphorylation of the four phosphoproteins investigated between UL and sham controls at any post-operative time Ždata not shown.. Therefore, we chose to investigate Ca2qrPSrPDB-stimulated phosphorylation following UL in the remaining experiments described in this paper. Phosphorylation mediated by endogenous protein kinase C was assessed by calculating the ratio of phosphorylation in the presence of protein kinase C activa-

Fig. 2. Phosphorylation ratios of four MVNrPH proteins ŽA–D. at 10 h, 53 h, and 2 weeks post-UL ŽUL. or -sham. Ratios were calculated as follows: Ca2qrPSrPDB-stimulated phosphorylation ŽOD. % basal phosphorylation ŽOD.. Ratios are means" 1 S.E.M. Žbars. and are based upon results from three Ž10 h and 53 h. or two Ž2 weeks. independent experiments. The increased P75 phosphorylation at 53 h was statistically significant compared to sham controls Ž p - 0.05 ) ..

172

A.J. Sansom et al.r Brain Research 778 (1997) 166–177

Fig. 3. Percent change in the phosphorylation ratios of four MVNrPH proteins relative to sham Ž100%. at each post-operative time Žvalues are calculated from data in Fig. 2.. Note that compared to shams, the phosphorylation ratios of all proteins tended to decrease at 10 h and increase at 53 h post-UL.

Fig. 4. Autoradiographs ŽA and C. and coomassie stained gradient gels ŽB and D. showing the pattern of endogenous phosphorylation and proteins respectively, in the contra- and ipsi-MVNrPH at 10 h Žtracks 1, 2 and 7, 8., 53 h Žtracks 3, 4 and 9, 10. and 2 weeks Ž2 weeks; tracks 5, 6 and 11, 12. post-sham ŽA and B. or -UL ŽC and D.. Refer to Fig. 1 legend for assay conditions. The phosphorylation ratios for P75 and P46 over time are shown in Fig. 5.

A.J. Sansom et al.r Brain Research 778 (1997) 166–177

tors ŽCa2qrPSrPDB. divided by basal phosphorylation. Phosphorylation ratios Žexpressed as % of basal. for each of the four phosphoproteins, were then compared across experimental conditions.

173

Since there were no statistically significant differences in the phosphorylation ratio of any band between ipsi- and contra-MVNrPH, at any post-operative time following UL or sham surgery, data from ipsi- and contra-MVNrPH

Fig. 5. The time course of P75 ŽA–C. and P46 ŽD–F. phosphorylation ratios in the contra- and ipsi-MVNrPH following UL ŽB and E. or sham operation ŽC and F.. The increase in P46 phosphorylation between 0.4, or 2.2 days and 14 days post-UL was statistically significant Ž p - 0.001.. There were no significant changes in P46 phosphorylation over time in sham controls. Data points are means" 1 S.E.M. Žbars; B, C, E and F.; means are based upon results from three Ž0.4 and 2.2 days. or two Ž14 days. independent experiments.

174

A.J. Sansom et al.r Brain Research 778 (1997) 166–177

were combined for statistical analyses. First, possible differences in phosphorylation ratios between sham and UL tissue were examined at each post-operative time. At 10 h following UL, the phosphorylation ratios of all four phosphoproteins were less than controls wmean % change from sham; the MARCKS-like protein Žy13%., P46 Žy4%., P22 Žy33%. and P18 Žy17%.; see Fig. 2A–D and Fig. 3x. The phosphorylation of P22 was most affected by UL at this time. However, none of these differences reached statistical significance. In contrast, at 53 h post-UL, the phosphorylation ratios of all four bands increased compared to controls wmean % change from sham; the MARCKS-like protein Ž29%., P46 Ž31%., P22 Ž14%. and P18 Ž11%.; Fig. 2A–D and Fig. 3x. However, only the increased phosphorylation of the MARCKS-like protein was statistically significant at this time Ž F Ž1,10. s 17.1, p - 0.05.. At 2 weeks post-UL there were no differences in the phosphorylation of the MARCKS-like protein and P46 compared to controls ŽFig. 2A, B.. These changes are more clearly illustrated in Fig. 3 where the same data are plotted as a percent difference from control values. We also assessed possible changes in phosphorylation of the four MVNrPH proteins between each post-operative time. Phosphoprotein and protein patterns following UL ŽFig. 4C, D respectively., or sham operation ŽFig. 4A, B respectively. from a representative experiment are shown in Fig. 4. The phosphorylation ratios for the MARCKS-like protein and P46 at each post-operative time are shown in Fig. 5. Fig. 5B and 5E shows there was a rapid increase in the phosphorylation of the MARCKS-like protein and P46 respectively, between 10 h and 53 h post-UL, followed by more gradual rates of change between 53 h and 2 weeks Žtotal mean increases of 31% and 60% respectively, between 10 h and 2 weeks.. However, only the increase in P46 phosphorylation was statistically significant Ž F Ž2,13. s 13.6, p - 0.001.. Post-hoc analysis revealed that the period of greatest change occurred between 53 h and 2 weeks post-UL Ž p - 0.05., i.e. during the compensated stage of static symptom recovery Žmean increases of 35% and 33% for ipsi- and contra-MVNrPH respectively.. In contrast, there were no statistically significant changes in P46 phosphorylation ratios over time in controls Ž9% mean increase between 10 h and 2 weeks; Fig. 5F.. There were no significant changes in phosphorylation for any of the four bands between 10 and 53 h post-UL.

4. Discussion 4.1. Endogenous protein kinases and substrates in the MVN r PH The first aim of this study was to characterise the major endogenous phosphoproteins and their kinases in the MVNrPH. The following phosphoproteins were tentatively identified: synapsins 1a and 1b, the MARCKS-like

protein ŽP75., b-CaMKII, P46 Žequivalent band not present in cortex samples; identity unknown., P22 Židentity unknown. and P18 Žpossibly a myelin basic protein.. The in vitro labelling of most bands Žincluding the MARCKSlike protein, P46, P22 and P18. was increased in the presence of Ca2q alone, Ca2q plus calmodulin, or Ca2q plus PSrPDB, indicating the presence of one or more endogenous CaMK and Ca2qrphospholipid-dependent kinaseŽs. in this tissue. Since MARCKS is a selective protein kinase C substrate Žw2x; see w4,36x, for reviews., the increased Ca2qrPSrPDB-stimulated phosphorylation of the MARCKS-like protein suggests that the MVNrPH contains an endogenous protein kinase C. In contrast, the lack of cAMP-stimulated phosphorylation in the MVNrPH is in good agreement with the finding that the activity of cAMP-dependent kinase in the brainstem is among the lowest in mammalian brain w48x. 4.2. Phosphorylation changes in the MVN r PH following UL We sort to determine whether the in vitro phosphorylation of the four MVNrPH proteins was changed in response to UL andror over the time course of compensation. These results show, for the first time, that UL results in phosphorylation changes in the MVNrPH of the guinea pig. These changes were specific to two out of the four major phosphoproteins identified and mostly likely involved an endogenous Ca2qrphospholipid-dependent kinase. The mechanismŽs. responsible for the increase in Ca2qrPSrPDB-stimulated MARCKS-like protein phosphorylation at 53 h post-UL most likely involves an endogenous protein kinase C. The fact that the phosphorylation ratios of P46, P22 and P18 also increased at this time Žalthough not statistically significant., is consistent with this possibility. These results suggest that the acute compensated stage is accompanied by a change in protein kinase C pathways in vivo. However, it is also possible that the increase in MARCKS-like protein phosphorylation occurred via another mechanism, such as an increase in the synthesis of MARCKS-like protein or an increase in phosphatase activity in vivo. The results of the present experiment cannot distinguish between these possibilities. Evidence for an involvement of protein kinase C in neuronal plasticity has been increasing in recent years and to date includes such paradigms as long-term potentiation, associative learning paradigms, and axonal regeneration in the peripheral nervous system Žw50x; see w18,49x, for reviews.. It has recently been shown that UL resulted in a transient, asymmetric change in the spatial distribution of specific protein kinase C isozymes in the flocculonodular lobe of the rat cerebellum in vivo w3,20x. Since the MVN receives direct input from the flocculus w51x, it is possible that these changes in protein kinase C expression might alter the regulation of MVN activity following UL.

A.J. Sansom et al.r Brain Research 778 (1997) 166–177

While the physiological significance of MARCKS phosphorylation in neurons is not clear, current evidence suggests that it may be a multi functional protein Žsee w4,36x, for reviews.. In particular, it may function as a calmodulin regulating protein w36x, a regulator of the actin cytoskeleton w21x and play an essential role in neuronal development and survival w45x. MARCKS is present on membranes and in cytosol of both neurons and glial cells throughout the CNS w33x. It has been shown that the PH contains very high levels of MARCKS protein compared to other brainstem areas w33x. Therefore, it is possible that guinea pig MVNrPH MARCKS is predominantly a PH rather than an MVN protein. Conversely, a recent study has shown that the levels of MARCKS mRNA were low in PH but comparatively high in one region of the MVN of the adult rat w31x. The functional significance of changes in MARCKS-like protein phosphorylation and protein kinase C activity in the MVNrPH at 53 h following UL, remains to be determined. It is presently unclear whether the time-dependent increase in P46 phosphorylation following UL is related to the neurochemical processes underlying vestibular compensation or some other consequence of UL. However, this change cannot be related to maturation or aging since there were no significant changes in P46 phosphorylation over time in age-matched sham controls. Further experiments are necessary to substantiate a direct link between changes in P46 phosphorylation and compensation. The precise mechanismŽs. responsible for this phosphorylation change are unknown. Since MARCKS phosphorylation in tissue extracts can be used as an index of protein kinase C activity w2x, the lack of significant changes in MARCKSlike protein phosphorylation over time suggests that the change in P46 phosphorylation was not due to a change in protein kinase C activity. Possible changes in protein kinase C activity following UL could be measured directly in future experiments. The present results in guinea pig MVNrPH are generally consistent with those in frog whole brain w23,24x which show that vestibular compensation is associated with changes in the phosphorylation of several endogenous substrates in vitro. The phosphorylation of one of these proteins was shown to be Ca2qrphospholipid-dependent. However, the time course over which phosphorylation changes take place following UL is different between the two studies. In frog brain, changes in phosphorylation were associated with the early stages of postural compensation whereas in guinea pig, they were associated with the maintenance of ocular motor compensation. Although the reason for these differences is not clear, it is possible that a different balance of biochemical mechanisms is involved in achieving the same behavioural result within the two species. Furthermore, Janssen et al. w23x showed that vestibular compensation was correlated with a change in the phosphorylation of several CaMK substrates, including a 20 kDa band, immunologically identical to a myelin

175

basic protein. However, results from our preliminary experiments indicate that there were no changes in the CaMK phosphorylation of P18 Žpossibly a myelin basic protein; w47x. between 10 and 53 h post-UL Ždata not shown.. Further experiments will be necessary to clarify the role of CaMK-mediated phosphorylation during compensation in mammals. Finally, it is difficult to make direct comparisons between the present experiment and that of Janssen et al. w23x because in the latter experiment there were no sham-operated frogs run in parallel with the UL frogs to control for the effects of aging, surgery or anaesthetic. In fact, results from the present experiment showed that the phosphorylation of some bands did change over time in sham guinea pigs, however, these changes were not statistically significant Žfor example, the MARCKS-like protein and P22; see Fig. 5C and Fig. 4A, respectively.. Several studies have shown that the Fos protein and its mRNA are induced within hours of the UL in the MVN and PH, and these levels decrease to some extent over 1–3 days w26,27,5,10x. The induction of immediate early genes is thought to involve a number of second messenger pathways, including the inositol phosphaterprotein kinase C pathway Žsee w22x, for a review.. The results of the present study are consistent with these findings and suggest that protein phosphorylation systems in the MVNrPH might also be affected by UL at time points earlier than 10 h. The absence of a difference in protein phosphorylation between ipsi- and contra-MVNrPH at 10 h post-UL suggests that UL induces trans-synaptic changes in second messenger systems. Trans-synaptic regulation of protein kinase C activity has been observed in both the olfactory and visual systems following unilateral sensory deprivation in the rat w14x. An alternative explanation is that the absence of bilateral differences in phosphorylation may have been caused by the deafferentation produced by the tissue dissection procedure. The later interpretation seems unlikely given that bilateral changes in phosphorylation were observed over time following UL and not sham, in the present experiment. In summary, the in vitro phosphorylation of two out of four probable protein kinase C substrates changed significantly in the MVNrPH following UL. Behavioural studies will be required to address the functional relationship, if any, between possible changes in protein kinase C pathways in the MVNrPH and vestibular compensation in guinea pig.

Acknowledgements This research was funded by a New Zealand Neurological Foundation Project Grant Žto PFS. and a Health Research Council of New Zealand ŽHRC. Project Grant Žto CLD and PFS., an HRC Postgraduate Scholarship, a New

176

A.J. Sansom et al.r Brain Research 778 (1997) 166–177

Zealand Ministry of Foreign Affairs and Trade Travel Award and an HRC Young Investigators Award Žto AJS.. We wish to thank Dr. Darrin Gilchrist and Troy Margrie for assistance with earlier experiments; Dr. Imre Lengyel, Assoc. Professor Jim Gurd and Professor Peter Dunkley for helpful discussions. References w1x K.A. Albert, W.C. Wu, A.C. Nairn, P. Greengard, Inhibition by calmodulin of calciumrphospholipid-dependent protein phosphorylation, Proc. Natl. Acad. Sci. USA 81 Ž1984. 3622–3625. w2x L.G. Anderson, D. Zhao, K.R. Dell, D.L. Severson, R.N. Auer, Brain protein kinase C assay using MARCKS substrate reveals no translocation due to profound insulin-induced hypoglycaemia, Brain Res. 606 Ž1993. 187–194. w3x C.D. Balaban, M.M. Goto, G.G. Romero, Transient changes in cerebellar protein kinase C expression after unilateral labyrinthectomy, J. Vest. Res. Suppl. 6 Ž4S. Ž1996. S11. w4x P.J. Blackshear, The MARCKS family of cellular protein kinase C substrates, J. Biol. Chem. 268 Ž3. Ž1993. 1501–1504. w5x C. Cirelli, M. Pompeiano, P. D’Ascanio, P. Arrighi, O. Pompeiano, c-fos expression in the rat brain after unilateral labyrinthectomy and its relation to the uncompensated and compensated stages, Neurosci. 70 Ž2. Ž1996. 515–546. w6x P.J. Coggins, H. Zwiers, B-50 ŽGAP-43.: Biochemistry and functional neurochemistry of a neuron-specific phosphoprotein, J. Neurochem. 56 Ž1991. 1095–1106. w7x I.S. Curthoys, P.F. Smith, C.L. Darlington, Postural compensation in the guinea pig following unilateral labyrinthectomy, in: O. Pompeiano, J.H.J. Allum ŽEds.., Progress in Brain Research, vol. 76, Elsevier, Amsterdam, 1988, pp. 375–384. w8x I.S. Curthoys, G.M. Halmagyi, Vestibular compensation: a review of the ocular motor, neural, and clinical consequences of unilateral vestibular loss, J. Vest. Res. 5 Ž2. Ž1995. 67–107. w9x C.L. Darlington, P.F. Smith, The recovery of static vestibular function following peripheral vestibular lesions in mammals: The intrinsic mechanism hypothesis, J. Vest. Res. 6 Ž1996. 185–201. w10x C.L. Darlington, P. Lawlor, P.F. Smith, M. Dragunow, Temporal relationship between the expression of Fos, Jun and Krox-24 in the guinea pig vestibular nuclei during the development of vestibular compensation for unilateral vestibular deafferentation, Brain Res. 735 Ž1996. 173–176. w11x N. Dieringer, ‘Vestibular compensation’: neural plasticity and its relations to functional recovery after labyrinthine lesions in frogs and other vertebrates, Prog. Neurobiol. 46 Ž1995. 97–129. w12x C. de Waele, N. Vibert, M. Baudrimont, P.-P. Vidal, NMDA receptors contribute to the resting discharge of vestibular neurons in the normal and labyrinthectomized guinea pig, Exp. Brain Res. 81 Ž1990. 125–133. w13x C. de Waele, M. Abitbol, M. Chat, C. Menini, J. Mallet, P.-P. Vidal, Distribution of glutamatergic receptors and GAD mRNA-containing neurons in the vestibular nuclei of normal and hemilabyrinthectomized rats, Eur. J. Neurosci. 6 Ž1993. 565–576. w14x S. Elkabes, J.A. Cherry, A.A. Schoups, I.B. Black, Regulation of protein kinase C activity by sensory deprivation in the olfactory and visual systems, J. Neurochem. 60 Ž1993. 1835–1842. w15x H. Flohr, H. Bienhold, W. Ablen, I. Macskovics, Concepts of vestibular compensation, in Lesion-induced Neuronal Plasticity in Sensorimotor Systems, in: H. Flohr, W. Precht ŽEds.., SpringerVerlag, Berlin, 1981, pp. 153–172. w16x H. Flohr, U. Luneburg, C. Richter-Landsberg, ACTHrMSH-like neuropeptides and lesion-induced plastic processes, in: E. Keller, D.S. Zee ŽEds.., Adaptive Processes in Visual and Oculomotor Systems, Pergamon Press, Oxford, 1986, pp. 409–416.

w17x H. Flohr, U. Luneburg, Role of NMDA receptors in lesion-induced plasticity, Arch. Ital. Biol. 131 Ž1993. 173–190. w18x W.H. Gispen, A. Routtenberg, Protein kinase C and its brain substrates, Prog. Brain Res. 89 Ž1991.. w19x D.P.D. Gilchrist, C.L. Darlington, P.F. Smith, A dose-response analysis of the beneficial effects of the ACTH-Ž4–9. analogue, Org 2766, on behavioural recovery following unilateral labyrinthectomy in guinea pig, Br. J. Pharmacol. 111 Ž1994. 358–364. w20x M.M. Goto, G.G. Romero, C.D. Balaban, Transient changes in flocculonodular lobe protein kinase C expression during vestibular compensation, J. Neurosci. 17 Ž11. Ž1997. 4367–4381. w21x J.H. Hartwig, M. Thelen, A. Rosen, P.A. Janmey, A.C. Nairn, A. Aderem, MARCKS is an actin filament cross-linking protein regulated by protein kinase C and calcium-calmodulin, Nature 356 Ž1992. 618–622. w22x P. Hughes, M. Dragunow, Induction of immediate-early genes and the control of neurotransmitter-regulated gene expression within the nervous system, Pharmacol. Rev. 47 Ž1. Ž1995. 133–178. w23x U. Janssen, C. Richter-Landsberg, H. Flohr, Vestibular compensation affects endogenous phosphorylation of frog brain proteins, J. Neurochem. 58 Ž1. Ž1992. 65–71. w24x U. Janssen, C. Richter-Landsberg, A.B. Oestreicher, P.N.E. De Graaf, W.H. Gispen, H. Flohr, Identification of a B-50-like protein in frog brain synaptosomes, Brain Res. 570 Ž1992. 21–26. w25x T.M. Jeitner, P.E. Jarvie, M. Costa, J.A.P. Rostas, P.R. Dunkley, Protein phosphorylation in guinea pig myenteric ganglia and brain: presence of calmodulin kinase II, protein kinase C and cyclic AMP kinase and characterisation of major phosphoproteins, Neurosci. 40 Ž2. Ž1991. 555–569. w26x G.D. Kaufman, J.H. Anderson, A.J. Beitz, Brainstem Fos expression following acute unilateral labyrinthectomy in the rat, NeuroRep. 3 Ž1992. 829–832. w27x T. Kitahara, T. Saika, N. Takeda, H. Kiyama, T. Kubo, Changes in Fos and Jun expression in the rat brainstem in the process of vestibular compensation, Acta Otolaryngol. ŽStockholm. Suppl. 520 Ž1995. 401–404. w28x T. Kitahara, N. Takeda, T. Saika, T. Kubo, H. Kiyama, Effects of MK-801 on Fos expression in the rat brainstem after unilateral labyrinthectomy, Brain Res. 700 Ž1995. 182–190. w29x H. Li, D.A. Godfrey, A.M. Rubin, Quantitative autoradiography of 5-w 3 Hx6-cyano-7-nitro-quinoxaline-2,3-dione and Žq.-3-w 3 Hxdizocilpine maleate binding in rat vestibular nuclear complex after unilateral deafferentation, with comparison to cochlear nucleus, Neurosci. 77 Ž2. Ž1997. 473–484. w30x L.A. Marascuilo, R.C. Serlin, Statistical Methods for the Social and Behavioural Sciences, Freeman and Co, New York, 1988, pp. 501–503. w31x R.K. McNamara, R.H. Lenox, Comparative distribution of myristoylated alanine-rich C kinase substrate ŽMARCKS. and F1rGAP43 gene expression in the adult rat brain, J. Comp. Neurol. 379 Ž1997. 48–71. w32x S.D. Newlands, A.A. Perachio, Compensation of horizontal canal related activity in the medial vestibular nucleus following unilateral labyrinth ablation in the decerebrate gerbil. 1. Type I neurons, Exp. Brain Res. 82 Ž1990. 359–372. w33x C.C. Ouimet, J.K.T. Wang, I.S. Walaas, K.A. Albert, P. Greengard, Localisation of the MARCKS Ž87 kDa. protein, a major specific substrate for protein kinase C, in rat brain, J. Neurosci. 10 Ž5. Ž1990. 1683–1698. w34x V.E. Pettorossi, G. Della Torre, S. Grassi, M. Zampolini, G. Capocchi, P. Errico, Role of NMDA receptors in the compensation of ocular nystagmus induced by hemilabyrinthectomy in the guinea pig, Arch. Ital. Biol. 130 Ž1992. 303–313. w35x L. Ris, C. de Waele, M. Serafin, P.-P. Vidal, E. Godaux, Neuronal activity in the ipsilateral vestibular nucleus following unilateral labyrinthectomy in the alert guinea pig, J. Neurophysiol. 74 Ž5. Ž1995. 2087–2099.

A.J. Sansom et al.r Brain Research 778 (1997) 166–177 w36x P.J. Robinson, The role of protein kinase C and its neuronal substrates dephosphin, B-50, and MARCKS in neurotransmitter release, Mol. Neurobiol. 5 Ž1992. 87–130. w37x N. Sans, B. Moniot, J. Raymond, Effects of a unilateral labyrinthectomy on NMDA subunits mRNA expression in the vestibular nuclei of the guinea pig, Eur. J. Neurosci., Suppl. 8 Ž1995. abstract 70.17. w38x A.J. Sansom, C.L. Darlington, P.F. Smith, Intraventricular injection of an NMDA antagonist disrupts vestibular compensation, Neuropharmacol. 29 Ž1990. 83–84. w39x A.J. Sansom, C.L. Darlington, P.F. Smith, D.P.D. Gilchrist, C.J. Keenan, R. Kenyon, Injections of calmidazolium chloride into the ipsilateral medial vestibular nucleus or IVth ventricle reduce spontaneous ocular nystagmus following unilateral labyrinthectomy in guinea pig, Exp. Brain Res. 93 Ž1993. 271–278. w40x A.J. Sansom, P.F. Smith, C.L. Darlington, Fentazin anaesthesia for labyrinthine surgery in guinea pigs, J. Vest. Res. 6 Ž1. Ž1996. 49–52. w41x P.F. Smith, I.S. Curthoys, Neuronal activity in the ipsilateral medial vestibular nucleus of the guinea pig following unilateral labyrinthectomy, Brain Res. 444 Ž1988. 308–319. w42x P.F. Smith, I.S. Curthoys, Mechanisms of recovery following unilateral labyrinthectomy: a review, Brain Res. Rev. 14 Ž1989. 155–180. w43x P.F. Smith, C.L. Darlington, The NMDA antagonists MK-801 and CPP disrupt compensation for unilateral labyrinthectomy in the guinea pig, Neurosci. Lett. 94 Ž1988. 309–313.

177

w44x P.F. Smith, C.L. Darlington, Neurochemical mechanisms of recovery from peripheral vestibular lesions Žvestibular compensation., Brain Res. Rev. 16 Ž1991. 117–133. w45x D.J. Stumpo, C.B. Bock, J.S. Tuttle, P.J. Blackshear, MARCKS deficiency in mice leads to abnormal brain development and prenatal death, Proc. Natl. Acad. Sci. USA 92 Ž1995. 944–948. w46x T. Uemura, B. Cohen, Effects of vestibular lesions on vestibulo-ocular reflexes and posture in monkeys, Acta Oto-Laryngol. ŽStockholm., Suppl. 315 Ž1973. 1–71. w47x T.V. Waehneldt, N.J. Malotka, N.J. Karin, J.-M. Matthieu, Phylogenetic examination of vertebrate central nervous system myelin proteins by electro-immunoblotting, Neurosci. Lett. 57 Ž1985. 97–102. w48x S.I. Walaas, A.C. Nairn, P. Greengard, Regional distribution of calcium and cyclic AMP-regulated protein phosphorylation systems. in mammalian brain. II. Soluble systems, J. Neurosci. 3 Ž1983. 302–311. w49x S.I. Walaas, P. Greengard, Protein phosphorylation and neuronal function, Pharmacol. Rev. 43 Ž3. Ž1991. 299–349. w50x P. Wiklund, Per A.R. Ekstrom, M. Edbladh, D. Tonge, A. Edstrom, Protein kinase C and mouse sciatic nerve regeneration, Brain Res. 715 Ž1996. 145–154. w51x V.J. Wilson, G. Melvill-Jones, Mammalian Vestibular Physiology, Plenum, New York, 1979.