- Email: [email protected]
Immunohistochemical localization of calcium-binding proteins in the brainstem vestibular nuclei of the jaundiced Gunn rat
Hearing Research 173 (2002) 82^90 www.elsevier.com/locate/heares
Immunohistochemical localization of calcium-binding proteins in the brainstem vestibular nuclei of the jaundiced Gunn rat Wayne T. Shaia a , Steven M. Shapiro a;b;c; , Andrew J. Heller a , David L. Galiani d , Aristides Sismanis a , Robert F. Spencer a;d a
Department of Otolaryngology ^ Head and Neck Surgery, School of Medicine, Medical College of Virginia Campus, Virginia Commonwealth University Health System, Richmond, VA 23298-0599, USA b Department of Neurology, School of Medicine, Medical College of Virginia Campus, Virginia Commonwealth University Health System, P.O. Box 980599, Richmond, VA 23298-0599, USA c Department of Pediatrics, School of Medicine, Medical College of Virginia Campus, Virginia Commonwealth University Health System, Richmond, VA 23298-0599, USA d Department of Anatomy, School of Medicine, Medical College of Virginia Campus, Virginia Commonwealth University Health System, Richmond, VA 23298-0599, USA Received 2 October 2001; accepted 23 July 2002
Abstract Vestibular gaze and postural abnormalities are major sequelae of neonatal hyperbilirubinemia. The sites and cellular effects of bilirubin toxicity in the brainstem vestibular pathway are not easily detected. Since altered intracellular calcium homeostasis may play a role in neuronal cell death, we hypothesized that altered expression of calcium-binding proteins may occur in brainstem vestibular nuclei of the classic animal model of bilirubin neurotoxicity. The expression of the calcium-binding proteins calbindinD28k and parvalbumin in the brainstem vestibular pathways and cerebellum of homozygous recessive jaundiced (jj) Gunn rats was examined by light microscopy and immunohistochemistry at 18 days postnatally and compared to the findings obtained from agematched non-jaundiced heterozygous (Nj) littermate controls. Jaundiced animals exhibited decreased parvalbumin immunoreactivity specifically in synaptic inputs to superior, medial, and inferior vestibular nuclei, and to oculomotor and trochlear nuclei, whereas the neurons retained their normal immunoreactivity. Jaundiced animals also demonstrated a decrease in calbindin expression in the lateral vestibular nuclei and a paucity of calbindin-immunoreactive synaptic endings on the somata of Deiters’ neurons. The involved regions are related to the control of the vestibulo-ocular and vestibulospinal reflexes. Decreased expression of calcium-binding proteins in brainstem vestibular neurons may relate to the vestibulo-ocular and vestibulospinal dysfunction seen with clinical kernicterus, and may provide a sensitive new way to assess bilirubin toxicity in the vestibular system. 6 2002 Elsevier Science B.V. All rights reserved. Key words: Parvalbumin; Calbindin-D28k ; Vestibulo-ocular; Vestibulospinal ; Posture; Balance; Gaze; Gunn rat; Jaundice ; Kernicterus ; Hyperbilirubinemia
1. Introduction * Corresponding author. Tel.: +1 (804) 828-7416; Fax: +1 (804) 828-5654. E-mail address: [email protected] (S.M. Shapiro). Abbreviations: BSA, bovine serum albumin; IVN, inferior vestibular nucleus; ip, intraperitoneal; jj, homozygous recessive; LVN, lateral vestibular nucleus; MLF, medial longitudinal fasciculus; MVN, medial vestibular nucleus; NHS, normal horse serum; Nj, heterozygous; NN, homozygous dominant; PBS, phosphate-bu¡ered saline; SVN, superior vestibular nucleus
Neonatal hyperbilirubinemia is common in newborn infants who develop physiological jaundice due to the immaturity of the hepatic enzyme uridine diphosphate glucuronosyl transferase, which converts unconjugated bilirubin to its conjugated form (Gourley, 1997). More extreme neonatal jaundice may be caused by pathological causes including fetomaternal blood incompatibility such as Rh disease, congenital anomalies, and enzyme
0378-5955 / 02 / $ ^ see front matter 6 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 5 9 5 5 ( 0 2 ) 0 0 6 3 1 - 7
HEARES 3990 1-10-02
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
de¢ciencies. Regardless of its physiologic or pathologic etiology, neonatal hyperbilirubinemia places the infant at risk for the development of bilirubin encephalopathy, which results in the pathological condition of kernicterus. Speci¢c central nervous system regions are particularly susceptible to bilirubin toxicity. Acute bilirubin encephalopathy is characterized by lethargy, impairment of upgaze (setting sun sign), high pitched cry, increased extensor tone, opisthotonus, retrocollis, fever, seizures and death (Larroche, 1968; Connolly and Volpe, 1990). The chronic sequelae of bilirubin toxicity, also known as kernicterus, include motor, auditory, oculomotor and dental disturbances. The motor manifestations include athetosis, dystonia and spasticity ; the auditory disturbances include hearing loss, deafness, auditory neuropathy or dys-synchrony and central dysfunction; the ocular abnormalities consist of impaired upgaze and strabismus ; and the dental ¢ndings is enamel dysplasia of deciduous teeth (Larroche, 1968; Connolly and Volpe, 1990). Many of the clinical symptoms relate to pathological changes in brainstem auditory and vestibular nuclei, subthalamic nucleus and basal ganglia, especially the globus pallidus, and the cerebellum (Ahdab-Barmada, 1987; Malamud, 1961; Martich-Kriss et al., 1995). The Gunn rat is a well-known animal model for bilirubin encephalopathy. This mutant of the Wistar strain of rats has an autosomal recessive genetic de¢ciency of the uridine diphosphate glucuronosyl transferase enzyme. The normal homozygous Gunn rat (NN) possesses normal enzyme activity. The heterozygote (Nj), with approximately 50% enzyme activity, is non-jaundiced and una¡ected. The homozygous recessive jaundiced Gunn rat (jj) has virtually no enzyme activity (Strebel and Odell, 1971), is unable to conjugate bilirubin, and displays many of the same neurological symptoms and histopathology that are exhibited by hyperbilirubinemic human newborns (Blanc and Johnson, 1959; Johnson et al., 1959, 1961; Schutta and Johnson, 1969). Blood bilirubin levels are elevated in the jaundiced Gunn rat (Schutta and Johnson, 1969) and the unbound or free unconjugated fraction correlates signi¢cantly with electrophysiological abnormalities in uninjected rats (Ahlfors and Shapiro, 2001). Following sulfonamide injection, bilirubin is forced out of the circulation into tissue acutely increasing the brain bilirubin concentration (Diamond and Schmid, 1966), lowering blood bilirubin measurements in rats as in humans (Silverman et al., 1956), and producing the consistent pathological changes of kernicterus (Blanc and Johnson, 1959; Rose and Wisniewski, 1979). In a previous study, we demonstrated a reduction in the density of immunoreactive staining for the calciumbinding proteins calbindin-D28k and parvalbumin in the lower brainstem auditory relay nuclei in the jaun-
83
diced Gunn rat in comparison to non-jaundiced Nj littermate control rats (Spencer et al., 2002) in the absence of kernicteric lesions. These changes in the expression of calcium-binding proteins in the brainstem auditory nuclei correspond to electrophysiological de¢cits in the brainstem auditory evoked responses described in this model (Shapiro, 1988, 1993; Shapiro and Hecox, 1988, 1989; Uziel et al., 1983; Shapiro, 1991; Shapiro and Conlee, 1991). Bilirubin toxicity is associated with speci¢c motor symptoms, including ataxia, athetosis, spasticity, retrocollis (backward arching of the neck), opisthotonus (backward arching of the trunk), and gaze de¢cits (Connolly and Volpe, 1990). The vestibular system is responsible for maintaining the body’s equilibrium through the control of posture (lateral vestibulospinal tract for extensors muscles in£uenced by proprioception mediated by GABAergic input from the vermis of the cerebellum), balance (medial vestibulospinal tract for proximal axial musculature in£uenced by the saccule and utricle, and, to a lesser extent, the semicircular canals) and coordination of gaze via vestibulo-ocular and vestibulocollic pathways (medial longitudinal fasciculus in£uenced by the semicircular canals and, to a lesser extent, the saccule and utricle). The present study has been undertaken to determine whether brainstem regions that are related to vestibular function are vulnerable to bilirubin toxicity.
2. Materials and methods 2.1. Experimental animals Non-jaundiced (Nj) and jaundiced (jj) rats were obtained from a Gunn rat breeding colony in the Division of Animal Resources at the Virginia Commonwealth University. Nj female heterozygotes were mated with jj male homozygotes to produce litters with approximately equal numbers of jj and Nj pups. At 5^10 days of age, pups were characterized phenotypically by their skin color. Healthy jj and Nj pups were matched within litters by weight and sex at 15 N 1 (mean N S.D.) days of age and injected intraperitoneally (ip) with 100^200 mg/kg sulfadimethoxine which competes with bilirubin for albumin-binding sites in the circulation promoting its transfer into brain (Diamond and Schmid, 1966). Control jjs were given an equivalent volume of saline, and control Njs received either sulfadimethoxine or saline. At 18 N 1 days, animals were re-weighed and their clinical symptomatology was characterized clinically on a rating scale of 0^5 as described previously (Spencer et al., 2002). Brie£y, 0 = normal, 1 = very mildly abnormal with slight ataxia of hindlimbs,
HEARES 3990 1-10-02
84
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
2 = mildly abnormal with mild hindlimb ataxia and dystonia, and mildly impaired righting re£ex, 3 = moderately abnormal as in 2 but with a more severe movement disorder and more prolonged righting re£ex, 4 = severely abnormal with failure of locomotion, general lack of spontaneous movement with occasional bursts of hyperactivity and no righting re£ex, and 5 = moribund including seizures and agonal respiration. The Gunn rats studied, 44 jjs and 26 Njs from 20 litters, and the results of examining their auditory brainstem structures were previously described (Spencer et al., 2002). Brie£y, they were injected at an average of 14.9 N 0.5 days of age (mean N S.D.) and sacri¢ced at 17.7 N 0.8 days. Experimental jjs were given 100 (N = 11), 125 (N = 4), 150 (N = 13) or 200 (N = 9) mg/ kg of sulfadimethoxine ip. Control jjs were injected with saline (N = 7), and control Njs were injected with either sulfadimethoxine 100 (N = 7), 125 (N = 1), 150 (N = 8) and 200 (N = 4), or saline (N = 6). The jjs injected with sulfadimethoxine gained signi¢cantly less weight than jj controls, 0.06 N 3.0 g vs. 5.9 N 2.4 g (P = 0.00002, independent two-tailed t-test), while Njs gained 5.1 N 2.5 g. There were no di¡erences in weight gain between Njs given sulfadimethoxine or saline or between jj controls and either Nj group. The jjs were more symptomatic than the Njs. Behavioral scores were 0 for all Njs, a median of 4 and range 2^5 for jjs given sulfadimethoxine, and a median of 1.5 and range 1^2 for jjs given saline. Overall group di¡erences in behavioral scores were statistically signi¢cant, and post-hoc comparisons showed jj sulfa worse than Nj sulfa groups (P = 0.0001), and jj saline worse than Nj saline (P = 0.04) (Spencer et al., 2002). There were no statistically signi¢cant di¡erences in clinical scores between di¡erent sulfadimethoxine doses in jjs. 2.2. Immunoperoxidase localization of calcium-binding proteins At 18 N 1 days of age, littermate jj rats and Nj rats were euthanized with an overdose of Nembutal and perfused transcardially with a ¢xative solution containing 4.0% paraformaldehyde and 0.1% glutaraldehyde in 0.1 M phosphate bu¡er, pH 7.2. Coronal sections were cut through the brainstem at 50 Wm thicknesses with a Vibratome and collected serially in 0.1 M sodium phosphate bu¡er. Sections were divided into three groups representing 150 Wm intervals through the rostral^caudal extent of the brainstem. Two groups of sections were immersed in 5% sodium borohydride for 15 min and then blocked in 10% normal horse serum (NHS) in phosphate-bu¡ered saline (PBS) containing 1.0% bovine serum albumin (BSA) and 0.1% Triton X-100, pH 7.2, for 1 h. Sections were incubated in mouse monoclonal antibodies to calbindin-D28k (Sigma ;
1:400) or parvalbumin (Sigma ; 1:2000) diluted in PBS with NHS and BSA for 18 h at 4‡C with constant agitation. Sections then were incubated in a biotinylated horse anti-mouse IgG (Vector, 1:200) in PBS with NHS for 60 min followed by an avidin-biotinylated horseradish peroxidase complex (Vector; 1:100) in PBS. Sections were reacted with 3,3P-diaminobenzidine tetrahydrochloride and hydrogen peroxide in 0.1 M phosphate bu¡er. Most brainstem sections were mounted on glass microscope slides pretreated with chrome alum-gelatin, dehydrated in graded alcohols and xylenes, and coverslipped with Cytoseal 60. Sections were examined and photographed with a Nikon Microphot FXA microscope using bright¢eld and Normarski di¡erential interference contrast illumination. All experimental procedures were reviewed and approved by the Institutional Animal Care and Use Committee of Virginia Commonwealth University and were performed in compliance with the NIH Guide for the Care and Use of Laboratory Animals.
3. Results 3.1. Immunohistochemical localization of calbindin-D28k The immunohistochemical localization of calbindin in Nj Gunn rats demonstrated dense immunoreactive staining in the lateral vestibular nucleus (LVN), but little, if any, labeling in the superior (SVN), medial (MVN), and inferior (IVN) vestibular nuclei (Fig. 1A). Immunoreactive labeling in the LVN was associated predominantly with synaptic boutons that encircled the somata of the large Deiters’ neurons, as well as distributed throughout the surrounding neuropil (Fig. 1B). In addition to the immunoreactive synaptic boutons, labeling was associated with axons descending from the cerebellum that presumably represent the Purkinje cell input to Deiters’ neurons (Fig. 1A). In the jj Gunn rats, the density of calbindin-immunoreactive staining in the LVN was reduced substantially (Fig. 1C). The paucity of calbindin-immunoreactive synaptic endings on the somata of Deiters’ neurons, as well as their sparse distributed throughout the neuropil (Fig. 1D), contrasted strikingly with the high density observed in the Nj animals. Furthermore, the decreased number of immunoreactive axons descending from the cerebellum to the LVN also was remarkable (Fig. 1C). This reduction in the number of immunoreactive axons was related speci¢cally to a substantial decrease of the number of Purkinje cells in the cerebellar vermis in the jj animals (Fig. 2B) in comparison to the Nj animals (Fig. 2A). There was also a reduction in immunoreactive staining in the inferior olivary nucleus of the jj
HEARES 3990 1-10-02
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
85
Fig. 1. Immunohistochemical localization of calbindin-D28k in the vestibular nuclei in Nj (A) and jj (B) Gunn rats. Calbindin is localized predominantly to the lateral vestibular nucleus and sparsely in the medial (MVN) and inferior (IVN) vestibular nuclei. In the LVN in Nj animals (C), immunoreactive synaptic endings outline the somata of Deiters’ neurons (asterisks). Note paucity of immunoreactive endings in the jj animal (D). Calibration: A and B, 500 Wm; C and D, 100 Wm.
animals (Fig. 2D) in comparison to the Nj animals (Fig. 2C). 3.2. Immunohistochemical localization of parvalbumin Parvalbumin-immunoreactive staining was distributed densely throughout the neuropil in the MVN and SVN, and to a lesser extent in the LVN and IVN, in the Nj rats (Fig. 3A). At least some of this neuropil labeling appeared to be derived from immunoreactive axons of the vestibular nerve that innervate the semicircular canals and the saccule and utricle and enter the vestibular nuclei ventral to the restiform body of the inferior cerebellar peduncle. In addition, many neurons within the SVN, MVN, IVN, and, to a lesser extent, the LVN were parvalbumin-immunoreactive. In the jj animals, parvalbumin immunoreactivity in the vestibular nuclei was reduced substantially (Fig. 3B,D). Much of this decrease was attributable to a reduction in the immunoreactive labeling associated with
the neuropil, whereas immunoreactive neurons appeared to be una¡ected (Fig. 3D). The decreased immunoreactive labeling in the neuropil, however, apparently was unrelated to the peripheral vestibular input, since the number and density of parvalbumin-immunoreactive axons of the vestibular nerve entering the vestibular nuclei were comparable to the Nj animals (Fig. 3A,B). Since the SVN and rostral portions of the MVN and IVN are the site of termination of primary vestibular axons that innervate the semicircular canals and second-order neurons in these nuclei have projections to the extraocular motor nuclei for the control of the vestibulo-ocular re£ex, oculomotor and preoculomotor nuclei in the midbrain were examined for changes in the expression of parvalbumin. In the Nj animals, the trochlear and oculomotor nuclei exhibited dense immunoreactive labeling associated with both the motoneurons and the surrounding neuropil (Fig. 4A). Neurons in both the rostral interstitial nucleus (Ri N) of the
HEARES 3990 1-10-02
86
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
median longitudinal fasciculus (MLF) and the interstitial nucleus of Cajal also were parvalbumin-immunoreactive (Fig. 4B). By contrast, in the jj animals the density of parvalbumin-immunoreactive labeling of both the neuropil and, to a lesser extent, the motoneurons in the oculomotor and trochlear nuclei was substantially reduced (Fig. 4C). The decreased neuropil labeling appeared to be attributed largely to a reduction in parvalbumin immunoreactivity in the rostral interstitial nucleus of the MLF and the interstitial nucleus of Cajal (Fig. 4D).
4. Discussion Bilirubin encephalopathy resulting from neonatal hyperbilirubinemia, also known as kernicterus, is asso-
ciated with a variety of sensory and motor de¢cits, including both vestibular and auditory symptoms. Previously, we demonstrated changes in the expression of the calcium-binding proteins calbindin-D28k and parvalbumin in speci¢c regions of the auditory brainstem pathway (Spencer et al., 2002) that are correlated with changes in the amplitude, wave latencies, and interwave intervals of the auditory brainstem responses in jaundiced Gunn rats (Shapiro and Conlee, 1991; Shapiro, 1994). The present study extends these ¢ndings to include changes in the expression of these calcium-binding proteins in the brainstem vestibular nuclei that presumably are related to functional de¢cits in posture, balance, and gaze. As noted in the previous report utilizing material from the same group of animals (Spencer et al., 2002), the extent of functional de¢cits is directly proportional to the severity of the clinical symptoms
Fig. 2. Immunohistochemical localization of calbindin-D28k in the cerebellum of Nj (A) and jj (B) Gunn rats. Note the decreased number and density, but increased size, of Purkinje cells in the jj animal in comparison to the Nj animal. In the inferior olivary nucleus, calbindin immunoreactivity is reduced in the dorsal (IOD), principal (IOPr), and medial (IOM) nuclei in the jj animal (D) in comparison to the Nj (C) animal. Calibrations: A and B, 250 Wm; C and D, 400 Wm.
HEARES 3990 1-10-02
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
87
Fig. 3. Immunohistochemical localization of parvalbumin in the vestibular nuclei in Nj (A) and jj (B) Gunn rats. Parvalbumin is localized predominantly to the superior (SVN), medial (MVN) and inferior (IVN) vestibular nuclei, as well as the vestibular nerve (8v ) axons. In the MVN in Nj animals (C), immunoreactivity is associated with both neurons and synaptic endings. Note the decrease in the density of immunoreactive synaptic endings in the jj animal (D). Calibration: A and B, 500 Wm; C and D, 100 Wm.
and occurs in the absence of kernicterus lesions. Furthermore, these changes are speci¢c, since neighboring non-vestibular regions of the brainstem are una¡ected. Finally, the abnormalities we describe are observed only with the acute encephalopathy caused by the competitive binding of sulfadimethoxine for bilirubin-binding sites on albumin, since the ¢ndings with jj saline control animals are comparable to the Nj saline and Nj sulfadimethoxine control animals. This study describes important new qualitative observations in this model of bilirubin neurotoxicity. Future studies using quantitative image analysis, Western blots, and £uorescently labeled secondary antibodies would provide additional data to support our qualitative ¢ndings. Bilirubin is forced out of the circulation into tissue following sulfadimethoxine injection. This acutely increases brain bilirubin concentration (Diamond and Schmid, 1966), and lowers blood bilirubin measure-
ments in rats as in humans (Silverman et al., 1956), and produces the pathological changes of kernicterus (Blanc and Johnson, 1959; Rose and Wisniewski, 1979). Thus, blood bilirubin measurements would not provide useful information regarding the severity of the bilirubin toxicity, and they were not performed in this study. In addition, it is possible that the abnormalities described in the experimental jj sulfa group may in part be due to malnutrition and dehydration. The impairment of motor function that occurs in this model interferes with suckling and leads to weight loss and dehydration, and the experimental group had signi¢cantly less weight gain than controls. However, acute brainstem auditory evoked potential changes do occur as early as 2 h after sulfadimethoxine injection into jj Gunn rats, before any e¡ects of malnutrition and dehydration would be expected to occur (Shapiro, 1988). Calcium-binding proteins have an important role in maintaining intracellular calcium homeostasis (Celio,
HEARES 3990 1-10-02
88
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
1990; McBurney and Neering, 1987). In the auditory brainstem, the expression of calcium-binding proteins changes during the postnatal period (Enderlin et al., 1987; Solbach and Celio, 1991; Friauf, 1993, 1994; Lohmann and Friauf, 1996), coinciding with the time of neuronal growth, the formation of synaptic connections (Kandler and Friauf, 1993; Kotak and Sanes, 1995) and the elaboration of neurotransmitter utilization (Kandler and Friauf, 1995; Kotak and Sanes, 1996). Recent ¢ndings indicate that bilirubin inhibits calcium/calmodulin-dependent kinase II, which is a calcium-activated second messenger system that regulates neuronal function (Churn et al., 1995). In other systems, the impaired calcium homeostasis resulting from the loss of calcium-binding proteins is related to both neuronal and behavioral de¢cits (Airaksinen et al., 1997).
Consistent with previous studies in other species (Spencer and Baker, 1990; Spencer and Wang, 1996; Kevetter, 1996; Horn et al., 1995), calbindin and parvalbumin are di¡erentially localized in the four principal nuclei (medial, lateral, superior, and inferior) of the brainstem vestibular complex in the Nj control Gunn rat. Di¡erent calcium-binding proteins furthermore are associated with individual axons in the vestibular nerve (Kevetter, 1996) and presumably not only are related to di¡erent peripheral targets in the semicircular canals and the saccule and utricle (Dechesne et al., 1988; Dememes et al., 1993), but also re£ect their di¡erent discharge characteristics (regular versus irregular) that are transferred by second-order neurons to the ascending kinetic vestibulo-ocular and descending tonic vestibulospinal systems (Boyle et al., 1992). The di¡erent patterns of localization of calbindin and parvalbumin in
Fig. 4. Immunohistochemical localization of parvalbumin in the oculomotor nucleus in Nj (A) and jj (B) Gunn rats. Although motoneurons in the jj animal retain immunoreactivity, the surrounding neuropil exhibits a marked decrease in immunoreactive axons and synaptic endings. Neurons in the interstitial nucleus of Cajal in Nj animals (C) exhibit intense parvalbumin immunoreactivity that is substantially reduced in the jj animal (D). Calibration: A and B, 100 Wm; C and D, 400 Wm.
HEARES 3990 1-10-02
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
the vestibular nuclei in the Nj Gunn rat at 18 days of age also appear to re£ect a speci¢c relationship to these dichotomous projections. In Nj Gunn rats, calbindin is localized predominantly in the lateral vestibular nucleus and is associated with preterminal axons and synaptic boutons that are distributed in a high density on the somata of Deiters’ neurons. Most, if not all, of the calbindin-immunoreactive input to the lateral vestibular nucleus is derived from the GABAergic inhibitory Purkinje cells in the vermis of the cerebellum (Baurle and Grusser-Cornehls, 1994). The loss of calbindin-immunoreactive terminals in the jj animals appears to re£ect an indirect cascade e¡ect due to the loss of Purkinje cells in the cerebellum rather than a direct e¡ect of bilirubin toxicity in the lateral vestibular nucleus. The pattern of changes in the lateral vestibular nucleus, the cerebellum, and the inferior olivary nucleus in the jaundiced Gunn rat is remarkably similar to the Leaner mutant mouse, in which the genetically determined loss of Purkinje cells in alternating sagittal bands of the cerebellar cortex is associated with a decrease in GABA-immunoreactive terminals in the lateral vestibular nucleus (Grusser-Cornehls et al., 1995) and loss of calbindin-immunoreactive staining in distinct subdivisions of the inferior olivary nucleus (Heckroth and Abbott, 1994). In the Leaner mouse, the loss of the Purkinje cell inhibition results in an increase in the activity of vestibular nucleus neurons that contributes to severe motor de¢cits (GrusserCornehls et al., 1995). Similarly, a loss of Purkinje cell inputs to Deiters’ neurons in the jj Gunn rat is likely to produce an increase in the tonic activity of the lateral vestibulospinal tract and axial extensor muscles that is manifested as postural de¢cits, such as opisthotonus, that are associated with bilirubin toxicity. In contrast to calbindin, in the Nj Gunn rats parvalbumin is localized predominantly in both neurons and synaptic boutons in the superior, medial, and inferior vestibular nuclei, which contain second-order neurons that have ascending connections to the extraocular motor nuclei for the vestibulo-ocular re£ex and descending connections to the cervical spinal cord for the vestibulocollic re£ex (McElligott and Spencer, 1999). In the jj animals, the changes in the expression of parvalbumin involve predominantly, if not exclusively, synaptic inputs to these vestibular nuclei, whereas the neurons retain their immunoreactivity. Although one source of parvalbumin-immunoreactive input to the vestibular nuclei is derived from the vestibular nerve, primary vestibular axons exhibit a normal intensity of immunoreactivity in the jaundiced animals. Consequently, any de¢cits in vestibulo-ocular function resulting from bilirubin toxicity must be indirect and involve other brainstem and/or cerebellar inputs. Furthermore, despite the presence of parvalbumin-immunoreactive neurons in
89
the vestibular nuclei in the jj animals, immunoreactive staining of synaptic terminals, but not of the motoneurons, in the oculomotor and trochlear nuclei is reduced. This reduction in terminal immunoreactive labeling is likely attributable to the dramatic reduction in the expression of parvalbumin in the rostral interstitial nucleus of the medial longitudinal fasciculus and the interstitial nucleus of Cajal, which contain premotor neurons that are related to the control of vertical gaze (Horn et al., 1995; Spencer and Wang, 1996). In conclusion, bilirubin toxicity in the jaundiced Gunn rat produces changes in the expression of the calcium-binding proteins, calbindin-D28k and parvalbumin, in the vestibular nuclei that are related to deficits in posture and gaze. In contrast to the auditory system, in which sensorineural hearing loss and auditory processing de¢cits are attributable to direct e¡ects on neurons in the brainstem auditory pathway, vestibular dysfunction is likely attributable to a cascade e¡ect involving other brainstem and/or cerebellar structures that are related to the control of posture and gaze through connections with neurons in the vestibular nuclei.
Acknowledgements Supported by a Grant from the National Institutes of Health NIDCD R01-DC00369. The technical assistance of Barbara Mann and Michael Nguyen is greatly appreciated.
References Ahdab-Barmada, M., 1987. Kernicterus in the premature neonate. J. Perinatol. 7, 149^152. Ahlfors, C.E., Shapiro, S.M., 2001. Auditory brainstem response and unbound bilirubin in jaundiced (jj) Gunn rat pups. Biol. Neonat. 80, 158^162. Airaksinen, M.S., Eilers, J., Garaschuk, O., Thoenen, H., Konnerth, A., Meyer, M., 1997. Ataxia and altered dendritic calcium signaling in mice carrying a targeted null mutation of the calbindin D28k gene. Proc. Natl. Acad. Sci. USA 94, 1488^1493. Baurle, J., Grusser-Cornehls, U., 1994. Calbindin D-28k in the lateral vestibular nucleus of mutant mice as a tool to reveal Purkinje cell plasticity. Neurosci. Lett. 167, 85^88. Blanc, W.A., Johnson, L., 1959. Studies on kernicterus; relationship with sulfonamide intoxication, report on kernicterus in rats with glucuronyl transferase de¢ciency and review of pathogenesis. J. Neuropathol. Exp. Neurol. 18, 165^189. Boyle, R., Goldberg, J.M., Highstein, S.M., 1992. Inputs from regularly and irregularly discharging vestibular nerve a¡erents to secondary neurons in squirrel monkey vestibular nuclei. III. Correlation with vestibulospinal and vestibuloocular output pathways. J. Neurophysiol. 68, 471^484. Celio, M.R., 1990. Calbindin D-28k and parvalbumin in the rat nervous system. Neuroscience 35, 375^475. Churn, S.B., DeLorenzo, R.J., Shapiro, S.M., 1995. Bilirubin induces
HEARES 3990 1-10-02
90
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
a calcium-dependent inhibition of multifunctional Ca2+/calmodulin-dependent kinase II activity in vitro. Pediatr. Res. 38, 949^954. Connolly, A.M., Volpe, J.J., 1990. Clinical features of bilirubin encephalopathy. Clin. Perinatol. 17, 371^379. Dechesne, C.J., Thomasset, M., Brehier, A., Sans, A., 1988. Calbindin (CaBP 28 kDa) localization in the peripheral vestibular system of various vertebrates. Hear. Res. 33, 273^278. Dememes, D., Eybalin, M., Renard, N., 1993. Cellular distribution of parvalbumin immunoreactivity in the peripheral vestibular system of three rodents. Cell Tissue Res. 274, 487^492. Diamond, I., Schmid, R., 1966. Experimental bilirubin encephalopathy: the mode of entry of bilirubin bilirubin-14 C into the CNS. J. Clin. Invest. 45, 678^689. Enderlin, S., Norman, A.W., Celio, M.R., 1987. Ontogeny of the calcium binding protein calbindin D-28k in the rat nervous system. Anat. Embryol. Berl. 177, 15^28. Friauf, E., 1993. Transient appearance of calbindin-D28k-positive neurons in the superior olivary complex of developing rats. J. Comp. Neurol. 334, 59^74. Friauf, E., 1994. Distribution of calcium-binding protein calbindinD28k in the auditory system of adult and developing rats. J. Comp. Neurol. 349, 193^211. Gourley, G.R., 1997. Bilirubin metabolism and kernicterus. Adv. Pediatr. 44, 173^229. Grusser-Cornehls, U., Luy, M., Baurle, J., 1995. Electrophysiology and GABA-immunocytochemistry in the vestibular nuclei of normal (C57BL/6J) and Leaner mutant mice. Brain Res. 703, 51^62. Heckroth, J.A., Abbott, L.C., 1994. Purkinje cell loss from alternating sagittal zones in the cerebellum of Leaner mutant mice. Brain Res. 658, 93^104. Horn, A.K., Buttner Ennever, J.A., Suzuki, Y., Henn, V., 1995. Histological identi¢cation of premotor neurons for horizontal saccades in monkey and man by parvalbumin immunostaining. J. Comp. Neurol. 359, 350^363. Johnson, L., Garcia, M.L., Figueroa, E., Sarmiento, F., 1961. Kernicterus in rats lacking glucuronyl transferase. Am. J. Dis. Child. 101, 322^349. Johnson, L., Sarmiento, F., Blanc, W.A., Day, R., 1959. Kernicterus in rats with an inherited de¢ciency of glucuronyl transferase. Am. J. Dis. Child. 97, 591^608. Kandler, K., Friauf, E., 1993. Pre- and postnatal development of e¡erent connections of the cochlear nucleus in the rat. J. Comp. Neurol. 328, 161^184. Kandler, K., Friauf, E., 1995. Development of electrical membrane properties and discharge characteristics of superior olivary complex neurons in fetal and postnatal rats. Eur. J. Neurosci. 7, 1773^ 1790. Kevetter, G.A., 1996. Pattern of selected calcium-binding proteins in the vestibular nuclear complex of two rodent species. J. Comp. Neurol. 365, 575^584. Kotak, V.C., Sanes, D.H., 1995. Synaptically evoked prolonged depolarizations in the developing auditory system. J. Neurophysiol. 74, 1611^1620. Kotak, V.C., Sanes, D.H., 1996. Developmental in£uence of glycinergic transmission: regulation of NMDA receptor-mediated EPSPs. J. Neurosci. 16, 1836^1843. Larroche, J.C., 1968. In: Vinken, P., Bruyn, G.W. (Eds.), Handbook of Clinical Neurology, Vol. 6. North-Holland, Amsterdam, pp. 491^515. Lohmann, C., Friauf, E., 1996. Distribution of the calcium-binding proteins parvalbumin and calretinin in the auditory brainstem of adult and developing rats. J. Comp. Neurol. 367, 90^109.
Malamud, N., 1961. In: Swinyard, C.A. (Ed.), Kernicterus and its Importance in Cerebral Pasly. Charles C. Thomas, Spring¢eld, IL, pp. 230^246. Martich-Kriss, V., Kollias, S.S., Ball, W.S., Jr., 1995. MR ¢ndings in kernicterus. Am. J. Neuroradiol. 16, 819^821. McBurney, R.N., Neering, I.R., 1987. Neuronal calcium homeostasis. Trends Neurosci. 10, 164^169. McElligott, J.G., Spencer, R.F., 1999. In: Anderson, J.H., Beitz, A.J. (Eds.), Neurochemistry of the Vestibular System. CRC Press, Miami, FL, in press. Rose, A.L., Wisniewski, H., 1979. Acute bilirubin encephalopathy induced with sulfadimethoxine in Gunn rats. J. Neuropathol. Exp. Neurol. 38, 152^165. Schutta, H.S., Johnson, L., 1969. Clinical signs and morphologic abnormalities in Gunn rats treated with sulfadimethoxine. J. Pediatr. 75, 1070^1079. Shapiro, S.M., 1988. Acute brainstem auditory evoked potential abnormalities in jaundiced Gunn rats given sulfonamide. Pediatr. Res. 23, 306^310. Shapiro, S.M., 1991. Binaural e¡ects in brainstem auditory evoked potentials of jaundiced Gunn rats. Hear. Res. 53, 41^48. Shapiro, S.M., 1993. Reversible brainstem auditory evoked potential abnormalities in jaundiced Gunn rats given sulfonamide. Pediatr. Res. 34, 629^633. Shapiro, S.M., 1994. Brainstem auditory evoked potentials in an experimental model of bilirubin neurotoxicity. Clin. Pediatr. (Phil.) 33, 460^467. Shapiro, S.M., Conlee, J.W., 1991. Brainstem auditory evoked potentials correlate with morphological changes in Gunn rat pups. Hear. Res. 57, 16^22. Shapiro, S.M., Hecox, K.E., 1988. Development of brainstem auditory evoked potentials in heterozygous and homozygous jaundiced Gunn rats. Brain Res. 469, 147^157. Shapiro, S.M., Hecox, K.E., 1989. Brain stem auditory evoked potentials in jaundiced Gunn rats. Ann. Otol. Rhinol. Laryngol. 98, 308^317. Silverman, W.A., Andersen, D.H., Blanc, W.A., Crozier, D.N., 1956. A di¡erence in mortality rate and incidence of kernicterus among premature infants allotted to two prophylactic antibacterial regimens. Pediatrics 18, 614^625. Solbach, S., Celio, M.R., 1991. Ontogeny of the calcium binding protein parvalbumin in the rat nervous system. Anat. Embryol. Berl. 184, 103^124. Spencer, R.F., Baker, R., 1990. Immunohistochemical localization of parvalbumin in cat brainstem oculomotor system. Soc. Neurosci. Abstr. 16, 1083. Spencer, R.F., Shaia, W.T., Gleason, A.T., Sismanis, A., Shapiro, S.M., 2002. Changes in calcium-binding protein expression in the auditory brainstem nuclei of the jaundiced Gunn rat. Hear. Res. 171, 129^141. Spencer, R.F., Wang, S.F., 1996. Immunohistochemical localization of neurotransmitters utilized by neurons in the rostral interstitial nucleus of the medial longitudinal fasciculus (riMLF) that project to the oculomotor and trochlear nuclei in the cat. J. Comp. Neurol. 366, 134^148. Strebel, L., Odell, G.B., 1971. Bilirubin uridine disphosphoglucuronyltransferase in rat liver microsomes: genetic variation and maturation. Pediatr. Res. 5, 548^559. Uziel, A., Marot, M., Pujol, R., 1983. The Gunn rat: an experimental model for central deafness. Acta Otolaryngol. 95, 651^656.
HEARES 3990 1-10-02
Immunohistochemical localization of calcium-binding proteins in the brainstem vestibular nuclei of the jaundiced Gunn rat Wayne T. Shaia a , Steven M. Shapiro a;b;c; , Andrew J. Heller a , David L. Galiani d , Aristides Sismanis a , Robert F. Spencer a;d a
Department of Otolaryngology ^ Head and Neck Surgery, School of Medicine, Medical College of Virginia Campus, Virginia Commonwealth University Health System, Richmond, VA 23298-0599, USA b Department of Neurology, School of Medicine, Medical College of Virginia Campus, Virginia Commonwealth University Health System, P.O. Box 980599, Richmond, VA 23298-0599, USA c Department of Pediatrics, School of Medicine, Medical College of Virginia Campus, Virginia Commonwealth University Health System, Richmond, VA 23298-0599, USA d Department of Anatomy, School of Medicine, Medical College of Virginia Campus, Virginia Commonwealth University Health System, Richmond, VA 23298-0599, USA Received 2 October 2001; accepted 23 July 2002
Abstract Vestibular gaze and postural abnormalities are major sequelae of neonatal hyperbilirubinemia. The sites and cellular effects of bilirubin toxicity in the brainstem vestibular pathway are not easily detected. Since altered intracellular calcium homeostasis may play a role in neuronal cell death, we hypothesized that altered expression of calcium-binding proteins may occur in brainstem vestibular nuclei of the classic animal model of bilirubin neurotoxicity. The expression of the calcium-binding proteins calbindinD28k and parvalbumin in the brainstem vestibular pathways and cerebellum of homozygous recessive jaundiced (jj) Gunn rats was examined by light microscopy and immunohistochemistry at 18 days postnatally and compared to the findings obtained from agematched non-jaundiced heterozygous (Nj) littermate controls. Jaundiced animals exhibited decreased parvalbumin immunoreactivity specifically in synaptic inputs to superior, medial, and inferior vestibular nuclei, and to oculomotor and trochlear nuclei, whereas the neurons retained their normal immunoreactivity. Jaundiced animals also demonstrated a decrease in calbindin expression in the lateral vestibular nuclei and a paucity of calbindin-immunoreactive synaptic endings on the somata of Deiters’ neurons. The involved regions are related to the control of the vestibulo-ocular and vestibulospinal reflexes. Decreased expression of calcium-binding proteins in brainstem vestibular neurons may relate to the vestibulo-ocular and vestibulospinal dysfunction seen with clinical kernicterus, and may provide a sensitive new way to assess bilirubin toxicity in the vestibular system. 6 2002 Elsevier Science B.V. All rights reserved. Key words: Parvalbumin; Calbindin-D28k ; Vestibulo-ocular; Vestibulospinal ; Posture; Balance; Gaze; Gunn rat; Jaundice ; Kernicterus ; Hyperbilirubinemia
1. Introduction * Corresponding author. Tel.: +1 (804) 828-7416; Fax: +1 (804) 828-5654. E-mail address: [email protected] (S.M. Shapiro). Abbreviations: BSA, bovine serum albumin; IVN, inferior vestibular nucleus; ip, intraperitoneal; jj, homozygous recessive; LVN, lateral vestibular nucleus; MLF, medial longitudinal fasciculus; MVN, medial vestibular nucleus; NHS, normal horse serum; Nj, heterozygous; NN, homozygous dominant; PBS, phosphate-bu¡ered saline; SVN, superior vestibular nucleus
Neonatal hyperbilirubinemia is common in newborn infants who develop physiological jaundice due to the immaturity of the hepatic enzyme uridine diphosphate glucuronosyl transferase, which converts unconjugated bilirubin to its conjugated form (Gourley, 1997). More extreme neonatal jaundice may be caused by pathological causes including fetomaternal blood incompatibility such as Rh disease, congenital anomalies, and enzyme
0378-5955 / 02 / $ ^ see front matter 6 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 5 9 5 5 ( 0 2 ) 0 0 6 3 1 - 7
HEARES 3990 1-10-02
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
de¢ciencies. Regardless of its physiologic or pathologic etiology, neonatal hyperbilirubinemia places the infant at risk for the development of bilirubin encephalopathy, which results in the pathological condition of kernicterus. Speci¢c central nervous system regions are particularly susceptible to bilirubin toxicity. Acute bilirubin encephalopathy is characterized by lethargy, impairment of upgaze (setting sun sign), high pitched cry, increased extensor tone, opisthotonus, retrocollis, fever, seizures and death (Larroche, 1968; Connolly and Volpe, 1990). The chronic sequelae of bilirubin toxicity, also known as kernicterus, include motor, auditory, oculomotor and dental disturbances. The motor manifestations include athetosis, dystonia and spasticity ; the auditory disturbances include hearing loss, deafness, auditory neuropathy or dys-synchrony and central dysfunction; the ocular abnormalities consist of impaired upgaze and strabismus ; and the dental ¢ndings is enamel dysplasia of deciduous teeth (Larroche, 1968; Connolly and Volpe, 1990). Many of the clinical symptoms relate to pathological changes in brainstem auditory and vestibular nuclei, subthalamic nucleus and basal ganglia, especially the globus pallidus, and the cerebellum (Ahdab-Barmada, 1987; Malamud, 1961; Martich-Kriss et al., 1995). The Gunn rat is a well-known animal model for bilirubin encephalopathy. This mutant of the Wistar strain of rats has an autosomal recessive genetic de¢ciency of the uridine diphosphate glucuronosyl transferase enzyme. The normal homozygous Gunn rat (NN) possesses normal enzyme activity. The heterozygote (Nj), with approximately 50% enzyme activity, is non-jaundiced and una¡ected. The homozygous recessive jaundiced Gunn rat (jj) has virtually no enzyme activity (Strebel and Odell, 1971), is unable to conjugate bilirubin, and displays many of the same neurological symptoms and histopathology that are exhibited by hyperbilirubinemic human newborns (Blanc and Johnson, 1959; Johnson et al., 1959, 1961; Schutta and Johnson, 1969). Blood bilirubin levels are elevated in the jaundiced Gunn rat (Schutta and Johnson, 1969) and the unbound or free unconjugated fraction correlates signi¢cantly with electrophysiological abnormalities in uninjected rats (Ahlfors and Shapiro, 2001). Following sulfonamide injection, bilirubin is forced out of the circulation into tissue acutely increasing the brain bilirubin concentration (Diamond and Schmid, 1966), lowering blood bilirubin measurements in rats as in humans (Silverman et al., 1956), and producing the consistent pathological changes of kernicterus (Blanc and Johnson, 1959; Rose and Wisniewski, 1979). In a previous study, we demonstrated a reduction in the density of immunoreactive staining for the calciumbinding proteins calbindin-D28k and parvalbumin in the lower brainstem auditory relay nuclei in the jaun-
83
diced Gunn rat in comparison to non-jaundiced Nj littermate control rats (Spencer et al., 2002) in the absence of kernicteric lesions. These changes in the expression of calcium-binding proteins in the brainstem auditory nuclei correspond to electrophysiological de¢cits in the brainstem auditory evoked responses described in this model (Shapiro, 1988, 1993; Shapiro and Hecox, 1988, 1989; Uziel et al., 1983; Shapiro, 1991; Shapiro and Conlee, 1991). Bilirubin toxicity is associated with speci¢c motor symptoms, including ataxia, athetosis, spasticity, retrocollis (backward arching of the neck), opisthotonus (backward arching of the trunk), and gaze de¢cits (Connolly and Volpe, 1990). The vestibular system is responsible for maintaining the body’s equilibrium through the control of posture (lateral vestibulospinal tract for extensors muscles in£uenced by proprioception mediated by GABAergic input from the vermis of the cerebellum), balance (medial vestibulospinal tract for proximal axial musculature in£uenced by the saccule and utricle, and, to a lesser extent, the semicircular canals) and coordination of gaze via vestibulo-ocular and vestibulocollic pathways (medial longitudinal fasciculus in£uenced by the semicircular canals and, to a lesser extent, the saccule and utricle). The present study has been undertaken to determine whether brainstem regions that are related to vestibular function are vulnerable to bilirubin toxicity.
2. Materials and methods 2.1. Experimental animals Non-jaundiced (Nj) and jaundiced (jj) rats were obtained from a Gunn rat breeding colony in the Division of Animal Resources at the Virginia Commonwealth University. Nj female heterozygotes were mated with jj male homozygotes to produce litters with approximately equal numbers of jj and Nj pups. At 5^10 days of age, pups were characterized phenotypically by their skin color. Healthy jj and Nj pups were matched within litters by weight and sex at 15 N 1 (mean N S.D.) days of age and injected intraperitoneally (ip) with 100^200 mg/kg sulfadimethoxine which competes with bilirubin for albumin-binding sites in the circulation promoting its transfer into brain (Diamond and Schmid, 1966). Control jjs were given an equivalent volume of saline, and control Njs received either sulfadimethoxine or saline. At 18 N 1 days, animals were re-weighed and their clinical symptomatology was characterized clinically on a rating scale of 0^5 as described previously (Spencer et al., 2002). Brie£y, 0 = normal, 1 = very mildly abnormal with slight ataxia of hindlimbs,
HEARES 3990 1-10-02
84
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
2 = mildly abnormal with mild hindlimb ataxia and dystonia, and mildly impaired righting re£ex, 3 = moderately abnormal as in 2 but with a more severe movement disorder and more prolonged righting re£ex, 4 = severely abnormal with failure of locomotion, general lack of spontaneous movement with occasional bursts of hyperactivity and no righting re£ex, and 5 = moribund including seizures and agonal respiration. The Gunn rats studied, 44 jjs and 26 Njs from 20 litters, and the results of examining their auditory brainstem structures were previously described (Spencer et al., 2002). Brie£y, they were injected at an average of 14.9 N 0.5 days of age (mean N S.D.) and sacri¢ced at 17.7 N 0.8 days. Experimental jjs were given 100 (N = 11), 125 (N = 4), 150 (N = 13) or 200 (N = 9) mg/ kg of sulfadimethoxine ip. Control jjs were injected with saline (N = 7), and control Njs were injected with either sulfadimethoxine 100 (N = 7), 125 (N = 1), 150 (N = 8) and 200 (N = 4), or saline (N = 6). The jjs injected with sulfadimethoxine gained signi¢cantly less weight than jj controls, 0.06 N 3.0 g vs. 5.9 N 2.4 g (P = 0.00002, independent two-tailed t-test), while Njs gained 5.1 N 2.5 g. There were no di¡erences in weight gain between Njs given sulfadimethoxine or saline or between jj controls and either Nj group. The jjs were more symptomatic than the Njs. Behavioral scores were 0 for all Njs, a median of 4 and range 2^5 for jjs given sulfadimethoxine, and a median of 1.5 and range 1^2 for jjs given saline. Overall group di¡erences in behavioral scores were statistically signi¢cant, and post-hoc comparisons showed jj sulfa worse than Nj sulfa groups (P = 0.0001), and jj saline worse than Nj saline (P = 0.04) (Spencer et al., 2002). There were no statistically signi¢cant di¡erences in clinical scores between di¡erent sulfadimethoxine doses in jjs. 2.2. Immunoperoxidase localization of calcium-binding proteins At 18 N 1 days of age, littermate jj rats and Nj rats were euthanized with an overdose of Nembutal and perfused transcardially with a ¢xative solution containing 4.0% paraformaldehyde and 0.1% glutaraldehyde in 0.1 M phosphate bu¡er, pH 7.2. Coronal sections were cut through the brainstem at 50 Wm thicknesses with a Vibratome and collected serially in 0.1 M sodium phosphate bu¡er. Sections were divided into three groups representing 150 Wm intervals through the rostral^caudal extent of the brainstem. Two groups of sections were immersed in 5% sodium borohydride for 15 min and then blocked in 10% normal horse serum (NHS) in phosphate-bu¡ered saline (PBS) containing 1.0% bovine serum albumin (BSA) and 0.1% Triton X-100, pH 7.2, for 1 h. Sections were incubated in mouse monoclonal antibodies to calbindin-D28k (Sigma ;
1:400) or parvalbumin (Sigma ; 1:2000) diluted in PBS with NHS and BSA for 18 h at 4‡C with constant agitation. Sections then were incubated in a biotinylated horse anti-mouse IgG (Vector, 1:200) in PBS with NHS for 60 min followed by an avidin-biotinylated horseradish peroxidase complex (Vector; 1:100) in PBS. Sections were reacted with 3,3P-diaminobenzidine tetrahydrochloride and hydrogen peroxide in 0.1 M phosphate bu¡er. Most brainstem sections were mounted on glass microscope slides pretreated with chrome alum-gelatin, dehydrated in graded alcohols and xylenes, and coverslipped with Cytoseal 60. Sections were examined and photographed with a Nikon Microphot FXA microscope using bright¢eld and Normarski di¡erential interference contrast illumination. All experimental procedures were reviewed and approved by the Institutional Animal Care and Use Committee of Virginia Commonwealth University and were performed in compliance with the NIH Guide for the Care and Use of Laboratory Animals.
3. Results 3.1. Immunohistochemical localization of calbindin-D28k The immunohistochemical localization of calbindin in Nj Gunn rats demonstrated dense immunoreactive staining in the lateral vestibular nucleus (LVN), but little, if any, labeling in the superior (SVN), medial (MVN), and inferior (IVN) vestibular nuclei (Fig. 1A). Immunoreactive labeling in the LVN was associated predominantly with synaptic boutons that encircled the somata of the large Deiters’ neurons, as well as distributed throughout the surrounding neuropil (Fig. 1B). In addition to the immunoreactive synaptic boutons, labeling was associated with axons descending from the cerebellum that presumably represent the Purkinje cell input to Deiters’ neurons (Fig. 1A). In the jj Gunn rats, the density of calbindin-immunoreactive staining in the LVN was reduced substantially (Fig. 1C). The paucity of calbindin-immunoreactive synaptic endings on the somata of Deiters’ neurons, as well as their sparse distributed throughout the neuropil (Fig. 1D), contrasted strikingly with the high density observed in the Nj animals. Furthermore, the decreased number of immunoreactive axons descending from the cerebellum to the LVN also was remarkable (Fig. 1C). This reduction in the number of immunoreactive axons was related speci¢cally to a substantial decrease of the number of Purkinje cells in the cerebellar vermis in the jj animals (Fig. 2B) in comparison to the Nj animals (Fig. 2A). There was also a reduction in immunoreactive staining in the inferior olivary nucleus of the jj
HEARES 3990 1-10-02
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
85
Fig. 1. Immunohistochemical localization of calbindin-D28k in the vestibular nuclei in Nj (A) and jj (B) Gunn rats. Calbindin is localized predominantly to the lateral vestibular nucleus and sparsely in the medial (MVN) and inferior (IVN) vestibular nuclei. In the LVN in Nj animals (C), immunoreactive synaptic endings outline the somata of Deiters’ neurons (asterisks). Note paucity of immunoreactive endings in the jj animal (D). Calibration: A and B, 500 Wm; C and D, 100 Wm.
animals (Fig. 2D) in comparison to the Nj animals (Fig. 2C). 3.2. Immunohistochemical localization of parvalbumin Parvalbumin-immunoreactive staining was distributed densely throughout the neuropil in the MVN and SVN, and to a lesser extent in the LVN and IVN, in the Nj rats (Fig. 3A). At least some of this neuropil labeling appeared to be derived from immunoreactive axons of the vestibular nerve that innervate the semicircular canals and the saccule and utricle and enter the vestibular nuclei ventral to the restiform body of the inferior cerebellar peduncle. In addition, many neurons within the SVN, MVN, IVN, and, to a lesser extent, the LVN were parvalbumin-immunoreactive. In the jj animals, parvalbumin immunoreactivity in the vestibular nuclei was reduced substantially (Fig. 3B,D). Much of this decrease was attributable to a reduction in the immunoreactive labeling associated with
the neuropil, whereas immunoreactive neurons appeared to be una¡ected (Fig. 3D). The decreased immunoreactive labeling in the neuropil, however, apparently was unrelated to the peripheral vestibular input, since the number and density of parvalbumin-immunoreactive axons of the vestibular nerve entering the vestibular nuclei were comparable to the Nj animals (Fig. 3A,B). Since the SVN and rostral portions of the MVN and IVN are the site of termination of primary vestibular axons that innervate the semicircular canals and second-order neurons in these nuclei have projections to the extraocular motor nuclei for the control of the vestibulo-ocular re£ex, oculomotor and preoculomotor nuclei in the midbrain were examined for changes in the expression of parvalbumin. In the Nj animals, the trochlear and oculomotor nuclei exhibited dense immunoreactive labeling associated with both the motoneurons and the surrounding neuropil (Fig. 4A). Neurons in both the rostral interstitial nucleus (Ri N) of the
HEARES 3990 1-10-02
86
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
median longitudinal fasciculus (MLF) and the interstitial nucleus of Cajal also were parvalbumin-immunoreactive (Fig. 4B). By contrast, in the jj animals the density of parvalbumin-immunoreactive labeling of both the neuropil and, to a lesser extent, the motoneurons in the oculomotor and trochlear nuclei was substantially reduced (Fig. 4C). The decreased neuropil labeling appeared to be attributed largely to a reduction in parvalbumin immunoreactivity in the rostral interstitial nucleus of the MLF and the interstitial nucleus of Cajal (Fig. 4D).
4. Discussion Bilirubin encephalopathy resulting from neonatal hyperbilirubinemia, also known as kernicterus, is asso-
ciated with a variety of sensory and motor de¢cits, including both vestibular and auditory symptoms. Previously, we demonstrated changes in the expression of the calcium-binding proteins calbindin-D28k and parvalbumin in speci¢c regions of the auditory brainstem pathway (Spencer et al., 2002) that are correlated with changes in the amplitude, wave latencies, and interwave intervals of the auditory brainstem responses in jaundiced Gunn rats (Shapiro and Conlee, 1991; Shapiro, 1994). The present study extends these ¢ndings to include changes in the expression of these calcium-binding proteins in the brainstem vestibular nuclei that presumably are related to functional de¢cits in posture, balance, and gaze. As noted in the previous report utilizing material from the same group of animals (Spencer et al., 2002), the extent of functional de¢cits is directly proportional to the severity of the clinical symptoms
Fig. 2. Immunohistochemical localization of calbindin-D28k in the cerebellum of Nj (A) and jj (B) Gunn rats. Note the decreased number and density, but increased size, of Purkinje cells in the jj animal in comparison to the Nj animal. In the inferior olivary nucleus, calbindin immunoreactivity is reduced in the dorsal (IOD), principal (IOPr), and medial (IOM) nuclei in the jj animal (D) in comparison to the Nj (C) animal. Calibrations: A and B, 250 Wm; C and D, 400 Wm.
HEARES 3990 1-10-02
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
87
Fig. 3. Immunohistochemical localization of parvalbumin in the vestibular nuclei in Nj (A) and jj (B) Gunn rats. Parvalbumin is localized predominantly to the superior (SVN), medial (MVN) and inferior (IVN) vestibular nuclei, as well as the vestibular nerve (8v ) axons. In the MVN in Nj animals (C), immunoreactivity is associated with both neurons and synaptic endings. Note the decrease in the density of immunoreactive synaptic endings in the jj animal (D). Calibration: A and B, 500 Wm; C and D, 100 Wm.
and occurs in the absence of kernicterus lesions. Furthermore, these changes are speci¢c, since neighboring non-vestibular regions of the brainstem are una¡ected. Finally, the abnormalities we describe are observed only with the acute encephalopathy caused by the competitive binding of sulfadimethoxine for bilirubin-binding sites on albumin, since the ¢ndings with jj saline control animals are comparable to the Nj saline and Nj sulfadimethoxine control animals. This study describes important new qualitative observations in this model of bilirubin neurotoxicity. Future studies using quantitative image analysis, Western blots, and £uorescently labeled secondary antibodies would provide additional data to support our qualitative ¢ndings. Bilirubin is forced out of the circulation into tissue following sulfadimethoxine injection. This acutely increases brain bilirubin concentration (Diamond and Schmid, 1966), and lowers blood bilirubin measure-
ments in rats as in humans (Silverman et al., 1956), and produces the pathological changes of kernicterus (Blanc and Johnson, 1959; Rose and Wisniewski, 1979). Thus, blood bilirubin measurements would not provide useful information regarding the severity of the bilirubin toxicity, and they were not performed in this study. In addition, it is possible that the abnormalities described in the experimental jj sulfa group may in part be due to malnutrition and dehydration. The impairment of motor function that occurs in this model interferes with suckling and leads to weight loss and dehydration, and the experimental group had signi¢cantly less weight gain than controls. However, acute brainstem auditory evoked potential changes do occur as early as 2 h after sulfadimethoxine injection into jj Gunn rats, before any e¡ects of malnutrition and dehydration would be expected to occur (Shapiro, 1988). Calcium-binding proteins have an important role in maintaining intracellular calcium homeostasis (Celio,
HEARES 3990 1-10-02
88
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
1990; McBurney and Neering, 1987). In the auditory brainstem, the expression of calcium-binding proteins changes during the postnatal period (Enderlin et al., 1987; Solbach and Celio, 1991; Friauf, 1993, 1994; Lohmann and Friauf, 1996), coinciding with the time of neuronal growth, the formation of synaptic connections (Kandler and Friauf, 1993; Kotak and Sanes, 1995) and the elaboration of neurotransmitter utilization (Kandler and Friauf, 1995; Kotak and Sanes, 1996). Recent ¢ndings indicate that bilirubin inhibits calcium/calmodulin-dependent kinase II, which is a calcium-activated second messenger system that regulates neuronal function (Churn et al., 1995). In other systems, the impaired calcium homeostasis resulting from the loss of calcium-binding proteins is related to both neuronal and behavioral de¢cits (Airaksinen et al., 1997).
Consistent with previous studies in other species (Spencer and Baker, 1990; Spencer and Wang, 1996; Kevetter, 1996; Horn et al., 1995), calbindin and parvalbumin are di¡erentially localized in the four principal nuclei (medial, lateral, superior, and inferior) of the brainstem vestibular complex in the Nj control Gunn rat. Di¡erent calcium-binding proteins furthermore are associated with individual axons in the vestibular nerve (Kevetter, 1996) and presumably not only are related to di¡erent peripheral targets in the semicircular canals and the saccule and utricle (Dechesne et al., 1988; Dememes et al., 1993), but also re£ect their di¡erent discharge characteristics (regular versus irregular) that are transferred by second-order neurons to the ascending kinetic vestibulo-ocular and descending tonic vestibulospinal systems (Boyle et al., 1992). The di¡erent patterns of localization of calbindin and parvalbumin in
Fig. 4. Immunohistochemical localization of parvalbumin in the oculomotor nucleus in Nj (A) and jj (B) Gunn rats. Although motoneurons in the jj animal retain immunoreactivity, the surrounding neuropil exhibits a marked decrease in immunoreactive axons and synaptic endings. Neurons in the interstitial nucleus of Cajal in Nj animals (C) exhibit intense parvalbumin immunoreactivity that is substantially reduced in the jj animal (D). Calibration: A and B, 100 Wm; C and D, 400 Wm.
HEARES 3990 1-10-02
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
the vestibular nuclei in the Nj Gunn rat at 18 days of age also appear to re£ect a speci¢c relationship to these dichotomous projections. In Nj Gunn rats, calbindin is localized predominantly in the lateral vestibular nucleus and is associated with preterminal axons and synaptic boutons that are distributed in a high density on the somata of Deiters’ neurons. Most, if not all, of the calbindin-immunoreactive input to the lateral vestibular nucleus is derived from the GABAergic inhibitory Purkinje cells in the vermis of the cerebellum (Baurle and Grusser-Cornehls, 1994). The loss of calbindin-immunoreactive terminals in the jj animals appears to re£ect an indirect cascade e¡ect due to the loss of Purkinje cells in the cerebellum rather than a direct e¡ect of bilirubin toxicity in the lateral vestibular nucleus. The pattern of changes in the lateral vestibular nucleus, the cerebellum, and the inferior olivary nucleus in the jaundiced Gunn rat is remarkably similar to the Leaner mutant mouse, in which the genetically determined loss of Purkinje cells in alternating sagittal bands of the cerebellar cortex is associated with a decrease in GABA-immunoreactive terminals in the lateral vestibular nucleus (Grusser-Cornehls et al., 1995) and loss of calbindin-immunoreactive staining in distinct subdivisions of the inferior olivary nucleus (Heckroth and Abbott, 1994). In the Leaner mouse, the loss of the Purkinje cell inhibition results in an increase in the activity of vestibular nucleus neurons that contributes to severe motor de¢cits (GrusserCornehls et al., 1995). Similarly, a loss of Purkinje cell inputs to Deiters’ neurons in the jj Gunn rat is likely to produce an increase in the tonic activity of the lateral vestibulospinal tract and axial extensor muscles that is manifested as postural de¢cits, such as opisthotonus, that are associated with bilirubin toxicity. In contrast to calbindin, in the Nj Gunn rats parvalbumin is localized predominantly in both neurons and synaptic boutons in the superior, medial, and inferior vestibular nuclei, which contain second-order neurons that have ascending connections to the extraocular motor nuclei for the vestibulo-ocular re£ex and descending connections to the cervical spinal cord for the vestibulocollic re£ex (McElligott and Spencer, 1999). In the jj animals, the changes in the expression of parvalbumin involve predominantly, if not exclusively, synaptic inputs to these vestibular nuclei, whereas the neurons retain their immunoreactivity. Although one source of parvalbumin-immunoreactive input to the vestibular nuclei is derived from the vestibular nerve, primary vestibular axons exhibit a normal intensity of immunoreactivity in the jaundiced animals. Consequently, any de¢cits in vestibulo-ocular function resulting from bilirubin toxicity must be indirect and involve other brainstem and/or cerebellar inputs. Furthermore, despite the presence of parvalbumin-immunoreactive neurons in
89
the vestibular nuclei in the jj animals, immunoreactive staining of synaptic terminals, but not of the motoneurons, in the oculomotor and trochlear nuclei is reduced. This reduction in terminal immunoreactive labeling is likely attributable to the dramatic reduction in the expression of parvalbumin in the rostral interstitial nucleus of the medial longitudinal fasciculus and the interstitial nucleus of Cajal, which contain premotor neurons that are related to the control of vertical gaze (Horn et al., 1995; Spencer and Wang, 1996). In conclusion, bilirubin toxicity in the jaundiced Gunn rat produces changes in the expression of the calcium-binding proteins, calbindin-D28k and parvalbumin, in the vestibular nuclei that are related to deficits in posture and gaze. In contrast to the auditory system, in which sensorineural hearing loss and auditory processing de¢cits are attributable to direct e¡ects on neurons in the brainstem auditory pathway, vestibular dysfunction is likely attributable to a cascade e¡ect involving other brainstem and/or cerebellar structures that are related to the control of posture and gaze through connections with neurons in the vestibular nuclei.
Acknowledgements Supported by a Grant from the National Institutes of Health NIDCD R01-DC00369. The technical assistance of Barbara Mann and Michael Nguyen is greatly appreciated.
References Ahdab-Barmada, M., 1987. Kernicterus in the premature neonate. J. Perinatol. 7, 149^152. Ahlfors, C.E., Shapiro, S.M., 2001. Auditory brainstem response and unbound bilirubin in jaundiced (jj) Gunn rat pups. Biol. Neonat. 80, 158^162. Airaksinen, M.S., Eilers, J., Garaschuk, O., Thoenen, H., Konnerth, A., Meyer, M., 1997. Ataxia and altered dendritic calcium signaling in mice carrying a targeted null mutation of the calbindin D28k gene. Proc. Natl. Acad. Sci. USA 94, 1488^1493. Baurle, J., Grusser-Cornehls, U., 1994. Calbindin D-28k in the lateral vestibular nucleus of mutant mice as a tool to reveal Purkinje cell plasticity. Neurosci. Lett. 167, 85^88. Blanc, W.A., Johnson, L., 1959. Studies on kernicterus; relationship with sulfonamide intoxication, report on kernicterus in rats with glucuronyl transferase de¢ciency and review of pathogenesis. J. Neuropathol. Exp. Neurol. 18, 165^189. Boyle, R., Goldberg, J.M., Highstein, S.M., 1992. Inputs from regularly and irregularly discharging vestibular nerve a¡erents to secondary neurons in squirrel monkey vestibular nuclei. III. Correlation with vestibulospinal and vestibuloocular output pathways. J. Neurophysiol. 68, 471^484. Celio, M.R., 1990. Calbindin D-28k and parvalbumin in the rat nervous system. Neuroscience 35, 375^475. Churn, S.B., DeLorenzo, R.J., Shapiro, S.M., 1995. Bilirubin induces
HEARES 3990 1-10-02
90
W.T. Shaia et al. / Hearing Research 173 (2002) 82^90
a calcium-dependent inhibition of multifunctional Ca2+/calmodulin-dependent kinase II activity in vitro. Pediatr. Res. 38, 949^954. Connolly, A.M., Volpe, J.J., 1990. Clinical features of bilirubin encephalopathy. Clin. Perinatol. 17, 371^379. Dechesne, C.J., Thomasset, M., Brehier, A., Sans, A., 1988. Calbindin (CaBP 28 kDa) localization in the peripheral vestibular system of various vertebrates. Hear. Res. 33, 273^278. Dememes, D., Eybalin, M., Renard, N., 1993. Cellular distribution of parvalbumin immunoreactivity in the peripheral vestibular system of three rodents. Cell Tissue Res. 274, 487^492. Diamond, I., Schmid, R., 1966. Experimental bilirubin encephalopathy: the mode of entry of bilirubin bilirubin-14 C into the CNS. J. Clin. Invest. 45, 678^689. Enderlin, S., Norman, A.W., Celio, M.R., 1987. Ontogeny of the calcium binding protein calbindin D-28k in the rat nervous system. Anat. Embryol. Berl. 177, 15^28. Friauf, E., 1993. Transient appearance of calbindin-D28k-positive neurons in the superior olivary complex of developing rats. J. Comp. Neurol. 334, 59^74. Friauf, E., 1994. Distribution of calcium-binding protein calbindinD28k in the auditory system of adult and developing rats. J. Comp. Neurol. 349, 193^211. Gourley, G.R., 1997. Bilirubin metabolism and kernicterus. Adv. Pediatr. 44, 173^229. Grusser-Cornehls, U., Luy, M., Baurle, J., 1995. Electrophysiology and GABA-immunocytochemistry in the vestibular nuclei of normal (C57BL/6J) and Leaner mutant mice. Brain Res. 703, 51^62. Heckroth, J.A., Abbott, L.C., 1994. Purkinje cell loss from alternating sagittal zones in the cerebellum of Leaner mutant mice. Brain Res. 658, 93^104. Horn, A.K., Buttner Ennever, J.A., Suzuki, Y., Henn, V., 1995. Histological identi¢cation of premotor neurons for horizontal saccades in monkey and man by parvalbumin immunostaining. J. Comp. Neurol. 359, 350^363. Johnson, L., Garcia, M.L., Figueroa, E., Sarmiento, F., 1961. Kernicterus in rats lacking glucuronyl transferase. Am. J. Dis. Child. 101, 322^349. Johnson, L., Sarmiento, F., Blanc, W.A., Day, R., 1959. Kernicterus in rats with an inherited de¢ciency of glucuronyl transferase. Am. J. Dis. Child. 97, 591^608. Kandler, K., Friauf, E., 1993. Pre- and postnatal development of e¡erent connections of the cochlear nucleus in the rat. J. Comp. Neurol. 328, 161^184. Kandler, K., Friauf, E., 1995. Development of electrical membrane properties and discharge characteristics of superior olivary complex neurons in fetal and postnatal rats. Eur. J. Neurosci. 7, 1773^ 1790. Kevetter, G.A., 1996. Pattern of selected calcium-binding proteins in the vestibular nuclear complex of two rodent species. J. Comp. Neurol. 365, 575^584. Kotak, V.C., Sanes, D.H., 1995. Synaptically evoked prolonged depolarizations in the developing auditory system. J. Neurophysiol. 74, 1611^1620. Kotak, V.C., Sanes, D.H., 1996. Developmental in£uence of glycinergic transmission: regulation of NMDA receptor-mediated EPSPs. J. Neurosci. 16, 1836^1843. Larroche, J.C., 1968. In: Vinken, P., Bruyn, G.W. (Eds.), Handbook of Clinical Neurology, Vol. 6. North-Holland, Amsterdam, pp. 491^515. Lohmann, C., Friauf, E., 1996. Distribution of the calcium-binding proteins parvalbumin and calretinin in the auditory brainstem of adult and developing rats. J. Comp. Neurol. 367, 90^109.
Malamud, N., 1961. In: Swinyard, C.A. (Ed.), Kernicterus and its Importance in Cerebral Pasly. Charles C. Thomas, Spring¢eld, IL, pp. 230^246. Martich-Kriss, V., Kollias, S.S., Ball, W.S., Jr., 1995. MR ¢ndings in kernicterus. Am. J. Neuroradiol. 16, 819^821. McBurney, R.N., Neering, I.R., 1987. Neuronal calcium homeostasis. Trends Neurosci. 10, 164^169. McElligott, J.G., Spencer, R.F., 1999. In: Anderson, J.H., Beitz, A.J. (Eds.), Neurochemistry of the Vestibular System. CRC Press, Miami, FL, in press. Rose, A.L., Wisniewski, H., 1979. Acute bilirubin encephalopathy induced with sulfadimethoxine in Gunn rats. J. Neuropathol. Exp. Neurol. 38, 152^165. Schutta, H.S., Johnson, L., 1969. Clinical signs and morphologic abnormalities in Gunn rats treated with sulfadimethoxine. J. Pediatr. 75, 1070^1079. Shapiro, S.M., 1988. Acute brainstem auditory evoked potential abnormalities in jaundiced Gunn rats given sulfonamide. Pediatr. Res. 23, 306^310. Shapiro, S.M., 1991. Binaural e¡ects in brainstem auditory evoked potentials of jaundiced Gunn rats. Hear. Res. 53, 41^48. Shapiro, S.M., 1993. Reversible brainstem auditory evoked potential abnormalities in jaundiced Gunn rats given sulfonamide. Pediatr. Res. 34, 629^633. Shapiro, S.M., 1994. Brainstem auditory evoked potentials in an experimental model of bilirubin neurotoxicity. Clin. Pediatr. (Phil.) 33, 460^467. Shapiro, S.M., Conlee, J.W., 1991. Brainstem auditory evoked potentials correlate with morphological changes in Gunn rat pups. Hear. Res. 57, 16^22. Shapiro, S.M., Hecox, K.E., 1988. Development of brainstem auditory evoked potentials in heterozygous and homozygous jaundiced Gunn rats. Brain Res. 469, 147^157. Shapiro, S.M., Hecox, K.E., 1989. Brain stem auditory evoked potentials in jaundiced Gunn rats. Ann. Otol. Rhinol. Laryngol. 98, 308^317. Silverman, W.A., Andersen, D.H., Blanc, W.A., Crozier, D.N., 1956. A di¡erence in mortality rate and incidence of kernicterus among premature infants allotted to two prophylactic antibacterial regimens. Pediatrics 18, 614^625. Solbach, S., Celio, M.R., 1991. Ontogeny of the calcium binding protein parvalbumin in the rat nervous system. Anat. Embryol. Berl. 184, 103^124. Spencer, R.F., Baker, R., 1990. Immunohistochemical localization of parvalbumin in cat brainstem oculomotor system. Soc. Neurosci. Abstr. 16, 1083. Spencer, R.F., Shaia, W.T., Gleason, A.T., Sismanis, A., Shapiro, S.M., 2002. Changes in calcium-binding protein expression in the auditory brainstem nuclei of the jaundiced Gunn rat. Hear. Res. 171, 129^141. Spencer, R.F., Wang, S.F., 1996. Immunohistochemical localization of neurotransmitters utilized by neurons in the rostral interstitial nucleus of the medial longitudinal fasciculus (riMLF) that project to the oculomotor and trochlear nuclei in the cat. J. Comp. Neurol. 366, 134^148. Strebel, L., Odell, G.B., 1971. Bilirubin uridine disphosphoglucuronyltransferase in rat liver microsomes: genetic variation and maturation. Pediatr. Res. 5, 548^559. Uziel, A., Marot, M., Pujol, R., 1983. The Gunn rat: an experimental model for central deafness. Acta Otolaryngol. 95, 651^656.
HEARES 3990 1-10-02