SYNAPSE 10:79-82 (1992)

Short Communication

Reduced Tyrosine Hydroxylase Immunoreactivity in LOCUS Coeruleus of Suicide Victims ANAT BIEGON AND SHEILA FIELDUST Lawrence Berkeley Laboratory, Berkeley, California 94720 (A.B.); Weizmann Institute of Science, Rehouot, Israel (S.F.)

KEY WORDS

Norepinephrine (noradrenaline), Human brain, Immunohistochemistry

ABSTRACT Antibodies against tyrosine hydroxylase (TH, the rate-limiting enzyme in norepinephrine synthesis) and dopamine p-hydroxylase (DBH, the last enzyme in the synthesis) were used for immunohistochemical staining of human brain locus coeruleus sections, obtained postmortem from suicide victims and matched controls. Stain density over individual cells was quantified by a computerized, video-camera-based image analysis system. Mean stain density for TH was significantly lower (by about 30%)in the locus coeruleus of suicide victims. There was no difference between suicides and controls in DBH immunoreactivity or in the number of TH immunoreactive cells. Reduced TH availability, either genetically or environmentally determined, may contribute to the noradrenergic insufficiency postulated to occur in depression and the increased p-adrenergic receptor concentrations observed in prefrontal cortex of suicide victims. The noradrenergic system is thought to be involved in the pathophysiology of depression and in the response to stress (e.g., Cooper, Bloom, and Roth, 1986 edition). Specifically, the catecholamine theory of depression postulates a noradrenergic deficit in the brains of afflicted individuals (Schildkraut, 1965). The nucleus locus coeruleus (LC) is the major source of noradrenergic innervation of the mammalian brain. The rate limiting enzyme in norepinephrine (noradrenaline, NA) synthesis is tyrosine hydroxylase (TH). Therefore, a deficiency in NA could arise from insufficient levels of TH in the LC. The present study was designed to test this hypothesis in the human brain, by comparing the TH content in LC sections obtained from suicide victims and matched controls postmortem using semiquantitative immunohistochemistry. Dopamine p-hydroxylase (DBH)is the last enzyme in NA synthesis. It is not a rate-limiting enzyme, but it is present in all noradrenergic cells (Cooper et al., 1986). Antibodies raised against DBH were employed in the validation of the technique as well as an additional control for the specificity of group differences for TH rather than any other cell constituents that may react with one or more of the reagents employed in the immunohistochemical procedure. 0 1992 WILEY-LISS, INC

Human brains were obtained at autopsy from the medical examiners' offices in New York and Jaffa, Israel as previously described (Biegon and Israeli, 1988; Gross-Isserof et al., 1990). The present study included brains from 12 male subjects, six suicide victims, and six pair-matched controls. All were drug and pathology free at the time of death. Subjects were paired for age, postmortem delay, and plane of sectioning. The subjects were also reasonably well matched for causes of death. Controls died of multiple trauma (accidental fall from heights) or gunshots (homicide). The causes of death among the suicide victims were multiple trauma, gunshot (self inflicted), and one case of hanging. Three of the six suicide victims had a confirmed history of depression. The other three had no psychiatric records. Subjects with a psychiatric history other than depression, e.g., schizophrenia, alcoholism, and substance abuse, were excluded. Thin (40-km) sections of the frozen brains were obtained from the level of the LC using a whole-body Bright cryotome as previously described (Gross-Isserof et al., 1990). Slide-mounted sections Received September 19,1991; accepted in revised form September 29,1991 Address reprint requests t o Dr. Anat Eiegon, Mail stop 55-121, Lawrence Berkeley Laboratory, 1 Cyclotron Road, Berkeley, CA 94720.

80

A. BIEGON AND S. FIELDUST

Fig. 1. Immunohistochemical staining of TH and DBH containing cells of the human locus coeruleus, postmortem. a. Sagittal section, unmagnified, showing TH staining in locus coeruleus (small arrow) and substantia nigra (large arrow). b. Magnified view of individual cells stained for DBH, illustrating the range of stain densities measured in individual subjects. Both illustrations represent a normal (control) brain. Antibodies to human TH and DBH were purchased from Eugene Tech International (NJ).The stain was developed by the Vectastain ABC kit purchased from Vector Laboratories (CA). A slightly modified version of the manufacturers recommended proce-

dures for frozen tissue was employed. Briefly, sections were fixed for 30 min in 4% paraformaldehyde (in PBS and 5% sucrose); washed 3 x 5’ in PBS, incubated in 0.5% Triton X 100, followed by an incubation in 10%normal horse serum. After 3 x 5’rinses in TBS containing 0.028 tween, the sections were incubated overnight a t 4°C with the primary antibody diluted (1:400) in TBS-tween + 1%goat serum. The following day, the sections were rinsed and incubated with the biotinylated antibody, followed by the vectastain reagent. Stain was developed for 4 min (DBH) or 7 min (TH) in a nickel-enhanced hydrogen peroxide-diaminobenzidine solution.

were postfixed with formalin and stained for TH and DBH using commercially available antibodies raised against human TH and DBH (ETI, NJ), the ABC kit and minor modifications of the manufacturers recommended procedures (Fig. 1). The procedure for slide mounted sections was initially tested on r a t brain sections to ascertain the specificity and correct neuroanatomical distribution of the stain. As expected, highly labeled cells were seen in both substantia nigra and LC of the rat with the TH antibody, while the DBH antibody produced dense staining only in the LC. Consecutive sections from 6 suicide victims and 6 pair-matched controls were used for TH and DBH immunostaining. Sections from suicides and pair-matched controls were processed simultaneously in the same reagents. Each antibody was used on duplicate sections (four pairs sagittal, two pairs coronal), and blanks were produced by running consecutive sections through the staining procedure without the primary antibody. Additional consecutive sections from each brain were stained histologically with cresyl violet. The stained, coverslipped sections were then subjected to comprehensive feature and density analysis with the aid of a video-camera based computerized image analysis system. A customized densitometric software package previously used for quantitative autoradiography and histochemistry (e.g., Biegon et al., 1988; Biegon and Wolff, 1986)was employed for semiquantitative measurement of TH immunoreactivity. Densitometry was performed at a X 60 magnification. Sagittal sections were sampled from the anterior to the posterior end of the LC, and all

the cell bodies visible in each sample field (6-15 cells/ field; 10-12 fielddsection, 1 mm apart) were outlined by a computer mouse to produce a measurement of the mean optical density over the cell body. Coronal sections were analysed bilaterally. Mean staining density/ subject (N = 6 per group) was then calculated based on at least 100 cells per subject, or a total of 1,200 cells. In the absence of standards (e.g., Biegon and Wolff, 1986) containing known amounts of pure antigen, the quantification relies on keeping all staining factors constant, keeping substrates in excess, and simultaneous processing of control and experimental sections of uniform thickness. Under these conditions, the quantity of reaction product is believed to be determined by the amount of TH protein present (e.g., Pearson et al., 1979). Densitometric analysis revealed lower TH stain intensity in suicides in every one of the pairs examined (Table I), resulting in a statistically significant difference regardless of the statistical test employed ( P < 0.025 by Student’s t-test for paired or independent values; two tailed). The mean group difference in stain intensity was 33%. Employing a n identical procedure on consecutive sections, substituting only the primary antibody, we did not observe differences between the two groups in DBH stain intensity(contro1s: 28.4 _t 6.5; suicides: 25.7 L 5.7; mean * sd of 6 subjectslgroup; nonsignificant). These results suggest a selective deficit in TH in the LC cells of suicide victims. However, whether this implies a presynaptic noradrenergic deficit may depend on the number of TH containing cells in the LC. Con-

TYROSINE HYDROXYLASE IN SUICIDE VICTIMS TABLE I. Tyrosine hydroxylase immunoreactiuity in locus coeruleus cells of suicide victims and pair-matched controls1

81

Masserano et al., 19811, this study supplies the first direct evidence from the human brain for the involveAge (w) TH IR ment of LC in general and LC TH in particular in suiPair Control Suicide Control Suicide cide and possibly depression. TH could be a substrate for the genetic loading seen in affective disorders (Men1 18 17 32.4 12.9 2 30 36 28.7 27.7 dlewicz, 19881,although a family study suggesting such 3 52 43 17.8 13.1 a link (Egeland et al., 1987) was not replicated in other 4 55 43 34.7 21.3 pedigrees (Detera-Wadleigh et al., 1987). Reports of 5 61 62 27.6 23.2 6 65 74 37.4 21.7 age-related reduction in LC cell number (determined by mean 46.8 45.8 29.77 19.98* TH staining) also exist (Chan-Palayet al., 1989b),but it SD 17 20 6.91 5.87 appears from those reports that the decline in LC cell ‘TH IR, Tyrosine hydroxylase immunoreactivity, expressed as stain intensity, in number is significant only in extreme old age (70-100 optical density (OD) X 100 units, calculated for each subject from individual cell optical density results of the computerized densitometric analysis (based on at least years old), which is not represented in our sample. On 100 cells/subject). Adjacent white matter readings were subtracted from each LC the basis of one case, the suggestion has been made that reading. Sections from suicides and pair-matched controls were processed simultaneously in the same reagents. Mean postmortem delay (time between death and age related LC cell loss is even more accentuated when autopsy; f.SD) was 28.3 i 10 hr in controls and 25.7 i 13.5 hr in suicide victims. * P = 0.023, Student’s t-test for paired values. depression, too, is present. We do not see a reduced LC cell number in our relatively young group of suicides and controls, but rather a selective reduction in TH ceivably, a larger number of cells/LC may compensate immunoreactivity, i.e., content. However, these findfor lower TH content/cell. To partially address this ings are not a t all contradictory; it may very well be the question, a morphometric software package (Galai, Is- case that late-onset depression (Georgotas and McCue, rael) was used to determine the number of TH positive 1988) or the depression often seen in patients with cells per section, acquiring images at low light micro- Alzheimer’s disease (Wragg and Jeste, 1989), in which of the cell counts marked reductions in LC cell populations were reported scopic magnification ( ~ 2 5 1 Some . were verified using a simple grid in a microscope eye- (e.g.,Zweig et al., 19881,is indeed linked to a presynappiece and found to be in excellent agreement. The re- tic noradrenergic deficit resulting from loss of norsults of the morphometric analysis showed no statisti- adrenergic cells, while early-onset depression assocically significant differences between this group of ated with genetic as well as environmental factors is suicides and controls in mean number of TH immuno- linked to reduced availability of TH in the LC. Since TH reactive cells found in the LC sections (controls: is the rate-limiting enzyme, insufficient TH production 398 5 127; suicides: 357 2 110; mean t sd of six will have the same end result as loss of cells producing subjects/group). Within the age range represented in the neurotransmitter, i.e., reduced noradrenergic tone. our brain sample, there were no significant effects of A possible additional result of such a presynaptic defiage on any of the parameters measured. The general cit could be an attempt a t a postsynaptic adaptive refeatures of the LC revealed in our sample are in good sponse through upregulation of postsynaptic p-adrenagreement with recent reports on the shape, size and ergic receptors. We have indeed observed increased cell density and patterns in humans (Chan-Palay and P-adrenergic receptor concentration in the prefrontal Asan, 1989a,b;German et al., 1988; Baker et al., 1989). cortex of the same subjects included in the present The total length of the nucleus, as determined from study (Biegon and Israeli, 1988). The same phenomeconsecutive coronal sections from two individuals, is non was described by others (Mann et al., 1986). Thus, 1.2-1.4 cm. The anterior end of the nucleus is charac- we propose that lower TH levels in the LC, genetically terized by a relatively low cell density and mostly large or environmentally determined, may be the source of a cells. Cell density increases and cell size decreases noradrenergic presynaptic deficit in suicidal and dealong the rostrocaudal axis. Comparison of pigmented pressed individuals. This kind of deficit can also excells on the histologically stained sections with TH and plain the increases in P-adrenergic receptor density DBH-positive cells on consecutive sections showed the seen in frontal cortex of suicide victims by us and others former consistently outnumber the latter. TH immu- (Biegon and Israeli, 1988; Mann et al., 1986). Current noreactive cells were less than 60% of the number of experiments in our lab, examining the TH mRNA in LC pigmented cells (data not shown). This, too, is in agree- by in situ hybridization, will help elucidate the mechament with recent reports (Chan-Palay and Asan, nisms responsible for reduced TH in the LC of suicide 1989b; German et al., 1988; Baker et al., 19891, sug- victims. gesting that not all the pigmented cells in the human LC contain detectable amounts of TH. Although a central, presynaptic noradrenergic deficit REFERENCES in depression has been suspected for decades (SchildBaker,K.G., Tork, I., Hornung, J.P., andHalam, P. (1989)Thehuman kraut, 19651, and LC was often implicated in animal locus coeruleus complex: An immunohistochemical and three dimensional reconstruction study. Exp. Brain Res. 477:257-270. models of stress and depression (e.g., Korf et al., 1973;

82

A. RlEGON AND S. FIELDUST

Biegon, A,, and Israeli, M. (1988) Regionally selective increases in adrenergic receptors in the brains of suicide victims. Brain Res., 442: 199-203. Biegon, A., and Wolff, M. (1986)Quantitative histochemistry of acetylcholinesterasr in rat and human brain postmortem. J. Neurosci. Methods, 16:39-45. Chan-Palay V., and Asan, E. (1989a) Alterations in catecholamine neurons of the locus coeruleus in senile dementia of the Alzheimer’s type and in Parkinson’s disease with and without dementia and depression. J. Comp. Neurol., 287:373-392. Chan-Palay, V. and Asan, E. (1989b) Quantitation of catecholamine neurons in the locus coeruleus in human brains of normal young and older adults and in depression. J . Comp. Neurol., 287:357-372. Cooper, J.R., Bloom, F.E., and Roth, R.H. (1986) The Biochemical basis ofNeuropharmacology.University Press, Oxford, pp. 304-308. Detera-Wadleigh, S.D., Berrettini, W.H., Goldin, L.R., Boorman, D., Anderson, S., and Gershon, E.S. (1987) Close linkage of C-HarveyRas-I and the insulin gene to affective disorder is ruled out in three North American pedigrees. Nature, 325:806-808. Egeland, J.A., Gerhard, D.S., Pauls, D.L., Sussex, J.N., I d d , K.K., Allen, C.R., Hostetter, A.M., and Houseman, D.E. (1987). Bipolar affective disorders linked to DNA markers on Chromosome 11. Nature, 325:783-787. Georgotas, A., and McCue, R. (1988) Depressive and manic states of late life. In: Depression and Mania. A. Georgotas and R. Cancro, eds. Elsevier, New York, pp. 592404. German, D.C., Walker, B.S., Manaye, K., Smith, W.K., Woodward, D.J., and North, A.J. (1988) The human locus coeruleus: Computer reconstruction of cellular distribution. J. Neurosci.. 8:1776-1788.

Gross-Isserof, R., Salama, D., and Biegon, A. (19901Autoradiographic analysis of 3H-ketanserin binding in the human brain postmortem: Effects of suicide. Brain Res., 507:20&215. Korf, J., Aghajanian, G., and Roth, R. (1973) Increased turnover of norepinephrine in the rat cerebral cortex during stress: Role of the locus coeruleus. Neuropharmacol., 12:933-938. Mann, J.J., Stanley, M., McBride, P.A., and McEwen, B.S. (1986) Increased serotonin2 and P-adrenergic receptor binding in the frontal cortices of suicide victims. Arch. Gen. Psychiatry, 43:954-959. Masserano, J.M., Takimoto, G.S., and Weiner, N. (1981) Electroconvulsive shock increases tyrosine hydroxylase activity in the brain and adrenal gland of the rat. Science, 214:662-665. Mendlewitz, J. (1988) Genetics of depression and mania. In: Depression and Mania. A. Georgotas and R. Cancro, eds. Elsevier, New York, pp, 197-212. Pearson, J., Brandeis, L., and Goldstein, M. (1979)Tyrosine hydroxylase immunoreactivity in familial dysautonomia. Science, 206:7172. Schildkraut, J.J. 119651The catecholamine hypothesis of affective disorders: A review of supporting evidence. Am. J. Psychiatry, 122509-522. Wragg, R.E., and Jeste, D.V. (1989) Overview of depression and psychosis in Alzheimer’s disease. Am. J. Psychiatry, 146:577-587. Zweig, R.M., Ross, C.A., Hedreen, J.C., Steele, C., Cardillo, J.E., Whitehouse, P.J., Folstein, M.F., and Price, D.L. (1988) The neuropathology of aminergic nuclei in Alzheimer’s disease. Ann. Neurol., 242334242,