Neuroscience Letters, 113 (1990) 17-22 17 Elsevier Scientific Publishers Ireland Ltd.

NSL 06846

Cerebral cortical calbindin D28 K and parvalbumin neurones in Down's syndrome

K. Kobayashi 1, P.C. Emson 1, C.Q. Mountjoy 2, S.N. Thornton 1, D.E.M. Lawson I and D.M.A. Mann 3

I M RC Group, AFRC Institute of Animal Physiology and Genetics Research, Babraham, Cambridge ( U.K. ), 2Department of Psychiatry, New Addenbrookes Hospital, Cambridge ( U.K.) and 3Department of

Pathology, University of Manchester, Manchester ( U.K.)

(Received 19 April 1989; Revised version received 14 January 1990; Accepted 18 January 1990)

Key words: Calbindin D2sK; Parvalbumin; Co-localization; Down's syndrome

Anti-calbindin D2g (CaBP) and anti-parvalbumin (PVA) antibodies were used to study the number and size of neurones containing these two calcium binding proteins in post-mortem brains from 7 neurologi- cally normal controls and from 4 elderly patients with clinically diagnosed Down's syndrome (DS) and whose brains contained numerous senile plaques and neurofibrillary tangles. The possible co-existence of these two calcium binding proteins in human cerebral cortex was also examined. In the controls, CaBP immunoreactive neurones were mainly non-pyramidal neurones although some pyramidal neurones were also CaBP immunoreactive. All the PVA immunoreactive neurones were non-pyramidal cells. CaBP and PVA did not apparently co-exist with each other in cortical neurones. When compared with the neurologi- cally normal controls, the number and size of CaBP and PVA immunoreactive neurones were significantly reduced in the cortex of patients with DS. These findings show that CaBP and PVA containing cortical neurones are affected in elderly persons with DS.

Clinicopathological studies of Down's syndrome (DS) indicate that the character- istic neuropathological changes observed in Alzheimer-type dementia (ATD), neuro- fibrillary tangles and senile plaques together with the symptoms of dementia appear in many DS-affected individuals who survive beyond 40 years of age and come to autopsy [4, 15]. Recently, the techniques of molecular biology and protein chemistry have begun to be applied to the problem of ATD. The structures of some of the com- ponents of the senile plaques have been determined [1, 5, 8, 12, 14]. These studies have stimulated increasing interest in the biochemistry and pathology of the brains of patients with DS, which is due to trisomy 21. In order to determine whether the changes observed in elderly DS cases match those observed in ATD, as a genetic link- age to familial Alzheimer's disease on chromosome 21 has been established [13].

We have previously reported a loss of calbindin D28K (CaBP) and parvalbumin (PVA) containing neurones in ATD [2, 6, 7]. These results suggest that the reduced

Correspondence: K. Kobayashi, MRC Group, AFRC Institute of Animal Physiology and Genetics Re- search, Babraham, Cambridge CB4 2AT, U.K.

0304-3940/90/$ 03.50 © 1990 Elsevier Scientific Publishers Ireland Ltd.

18

expression of these two calcium binding proteins may reflect a widespread distur- bance of calcium homeostasis in ATD perhaps predisposing certain vulnerable neu- rones to damage. In this study we have examined whether cortical CaBP and PVA immunoreactive neurones are also damaged in elderly patients with DS. In addition, since the possible co-existence of these two proteins, calbindin and parvalbumin in cortical neurones has not previously been examined we have studied whether these two proteins CaBP and PVA co-exist in human cortical neurones using double-stain- ing techniques. One important advantage of these two markers, for neuropathologi- cai studies, is that they are both remarkably stable post-mortem, so that it is possible to use antibodies directed against them together with automated cell counting techni-

Fig. 1. Low power photomicrographs to show the distribution of calbindin D~z (A) and parvalbumin (C) immunoreactive neurones in the frontal cortex of a neurologically normal case, compared with sections processed identically from an individual with Down's syndrome to visualize calbindin D2sK (B) and parval- bumin (D) immunoreactivity. Note the dramatic loss of calbindin (B) and parvalbumin (D) containing neurones in the Down's case. Bar: 200 urn.

19

ques to investigate whether CaBP and PVA containing neurones are damaged in DS

[2, 11]. Brains were obtained at post-mortem from 7 control patients (age at death: mean

74.7 + 9.3 years SD) who had no evidence of neurological or psychiatric disorder and from 4 patients (age at death: mean 57.8 ___ 1 years SD) clinically diagnosed DS whose brains contained neurones with senile plaques and neurofibrillary tangles. There was no statistically significant difference between the control and the DS group from time of death to autopsy (control: 29+ 11.9 h; DS: 51 + 27.3 h). The collection, storage and dissection ofpost-mortem human brain has been described previously [3]. Blocks were taken from frontal (Brodmann area 9) and temporal (Brodmann area 21) gyri of formalin-fixed brains [10]. Prior to sectioning, the tissue samples were transferred to 30~ sucrose (w/v) in 0.1 M phosphate buffer and kept at 4°C until required. Serial 20 / tm sections were cut from blocks on a freezing microtome and processed for visualization of CaBP or PVA immunoreactivity using the ABC method [9], CaBP or PVA antiserum was used at a dilution of 1:1000 and 1:500, respectively. For the examination of the co-localization of CaBP and PVA a double-staining immuno- fluorescence method was used. As primary antibodies a rabbit anti-CaBP antiserum and a sheep anti-PVA antiserum were used at dilutions of 1:1000 and 1:500, respecti- vely. We used rhodamine (RITC)-conjugated goat anti-rabbit IgG for CaBP detec- tion and fluorescein isothiocyanate (FITC)-conjugated donkey anti-sheep IgG as a second antibody for PVA. RITC and FITC fluorescence were visualized using the appropriate specific filter blocks on a Leitz Dialux 22 fluorescent microscope.

Counts of stained cells and measurements of cell body size were made with the aid of a Quantimet 720 computerized image analyser as previously reported [10].

Fig. 2. Higher power photomicrographs of the distribution of calbindin D2sK immunoreactive neurones in layers II-III of the frontal cortex of a normal control (A) and a Down's case (B) and parvalbumin immunoreactive neurones in layer IV of normal control case (C) and a Down's case (D). Bar: 50/an. Note again the loss of calbindin (B) and parvalbumin immunoreactive neurones (D) in the Down's case.

20

Four sections were selected randomly from serial sections from each case and the grey level of the Quantimet was set to detect CaBP- or PVA-stained cellular material and the counts and measurements were made in 4 vertical columns. Details of the reproducibility of this image analysis system have been described elsewhere [10].

In normal human brain CaBP and PVA immunoreactive neurones were observed in all layers except layer I of the cerebral cortex (Figs. 1A and C). The CaBP-positive cells were concentrated in layers II and III (Figs. IA and 2A). They were mainly mul- tipolar neurones, although bipolar, bitufted and some pyramidal neurones were also visualized (Fig. 2A). All of the PVA immunoreactive neurones were non-pyramidal and mainly multipolar cells, with a relatively greater abundance in layer IV (Figs. 1C and 2C). Using the immunofluorescence double-staining method, we could not find any evidence that these two markers co-existed within the same cortical neurones (Fig. 3), although the co-existence of PVA and CaBP immunoreactivity was readily demonstrated in Purkinje cells of the cerebellum (data not shown). The large number of orange autofluorescent lipofuscin granules in the cells of these elderly control cases made photography of fluorescent CaBP- or PVA-positive nerve cells difficult, although the positive cells (green or red) could be clearly distinguished from the or-

Fig. 3. Lack of co-existence of calbindin D28K and parvalbumin immunoreactivity in human c o ~ neu- rones. Double-colour immunofluorescent procedure using rhodamine to visualize calbindin D2sK immuno- reactive neurones (A) and (C) and fluorescein isothiocyanate to visualize parvalbumin immunoreactive neurones (B and D). Note that the calbindin D2g~ immunoreactive neurones (A) do not contain parvalbu- min immunoreactivity (B) and similarly the parvalbumin immunoreactive neurones (D) do not contain calbindin D2sK immunoreactivity (C). Note that the fluorescent material in (B) is due to lipofuschin gra- nules and is non-specific. The cells arrowed do not contain parvalbumin immunoreactivity.

21

ange autofluorescent under the fluorescence microscope. Neuropathologically, in the DS cases we could see degenerating neuronal profiles and the presence of degenerat- ing dendrites of CaBP and PVA immunoreactive neurones (Figs. 1 B, D and 2B, D). In particular, the CaBP-positive pyramidal cells of layer III were dramatically de- pleted in DS (Fig. 2B). Compared to control brains, the number of CaBP immuno- reactive neurones was significantly reduced in the frontal and temporal cortex in DS (Fig. 4A) with average reduction being 38% in the frontal cortex and 28% in the tem- poral cortex. The number of PVA immunoreactive neurones was also significantly reduced in DS in both cortical regions (Fig. 4B), with a reduction of 27% in the fron- tal cortex and 22% in the temporal cortex. In DS there was also a significant reduction in the cell body area of both CaBP- and PVA-positive neurones, in both the frontal and the temporal cortex (Fig. 4).

A

150

100

50

Cell count~

Frontal Temporal Frontal

O 150,

lOO-

.~ o o 2 ©

.-&-

100

50

O

C O C D C D (P<0.001) (P<0,02) (P<0.05)

~;ell size (urn 2)

Temporal ©

©

100 ~ o o o ~

50

C D (P<0.05)

B

150.

100

50.

Cell counts Cell size (urn 2)

Frontal Temporal Frontal Temporal

150 ©

lOO

0 7 o A

50

100-

50-

100.

@

5O

C D C D C D C D (P<0.02) (P<0.01) (P<0.01) (P<0.01)

Fig. 4. Total number and mean cell body size of ealbindin D2s K (A) and parvalbumin (B) immunoreactive ncurones. C, control group; D, DS group; Bar mean value of each group.

22

These data clearly indicate that both large pyramidal cells and smaller interneu-

rons are lost from the neocortex in DS, These results are consistent with those we

have previously reported in A T D , which indicates that the pat tern of neurona l loss

is similar in DS and in ATD. However, whether this implies a similar aetiology re-

mains to be established.

We are grateful to the Menta l Heal th F o u n d a t i o n (U.K.) , Bayer, F .R .G. and the

Medical Research Counci l for suppor t ing this work. We would like to thank Mr.

T. Buss for photographic work and Mrs. B.A. Waters for helping to prepare the

manuscr ipt . K .K. is a visiting fellow of the Nai to F o u n d a t i o n from Tokyo Metro-

pol i tan Matsuzawa Hospital , Tokyo, Japan.

1 Abraham, C.R., Selkoe, D.J. and Pottern H., lmmunochemical identification of the serine protease inhibitor cd-antichymotrypsin in the brain amyloid deposits of Alzheimer's disease, Cell, 52 (1988) 487--501.

2 Arai, H., Emson, P.C., Mountjoy, C.Q., Carassco, L.H. and Heizmann, C.W., Loss of parvalbumin- immunoreactive neurones from cortex in Alzheimer-type dementia, Brain Res., 418 (1987) 164-169.

3 Bird, E.D. and Iversen, L.L., Huntington's chorea Postmortem measurement ofglutamic acid decar- boxylase, choline acetyltransferase and dopamine in basal ganglia, Brain, 97 (1974) 457 ~472.

4 Ellis, W.G., McCulloch, J.R. and Corley, C.L., Presenile dementia in Down's syndrome. Ultrastruc- tural identity with Alzheimer's disease, Neurology, 24 (1974) 101-106.

5 Goldgaber, D., Lerman, M.I., McBride, O.W., Saffiotti, V. and Gajdusek, D.C., Characterization and chromosomal localization of a cDNA encoding brain amyloid of Alzheimer's disease, Science, 235 (1987) 877-880.

6 Ichimiya, Y., Emson, P.C., Mountjoy, C.Q., Lawson, D.E.M. and Heizmann, C.W., Loss of calbindin- 28K immunoreactive neurones from the cortex in Alzheimer-type dementia, Brain Res., 475 (1988) 156-159.

7 Ichimiya, Y., Emson, P.C., Mountjoy, C.Q. and Lawson, D.E.M., Calbindin immunoreactive neu- rones in the nucleus basalis of Meynert in Alzheimer-type dementia, Brain Res., 499 (1989) 402MO6.

8 Kang, J., Lemaire, H.G., Unterbec, A., Salbaum, J.M., Masters, C.L., Crzeschik, K.H., Multhaup, G., Beyreuther, K. and Muller Hill, B., The precursor of Alzheimer's disease amyloid A4 protein re- sembles a cell surface receptor, Nature, 325 (1987) 733-736.

9 Kristenson, K., The Use of Axonal Transport for Studies of Neuronal Connectivity, Elsevier, New York, 1975.

10 Mountjoy, C.Q., Roth, M., Evans, N.J.R. and Evans, H.M., Cortical neuronal counts in normal elderly controls and demented patients, Neurobiol. Aging, 4 (1983) 1 -11.

11 Parmentier, M., Ghysels, M., Rypens, F., Lawson, D.E.M., Pasteels, J.L. and Pochet, R., Calbindin in vertebrate classes: immunohistochemical localization and Western blot analysis, Gen. Comp. Endocrinol., 65 (1987) 399-407.

12 Robakis, N.K., Wisniewski, H.M., Jenkins, E.C., Devine-Gage, E.A., Houck, G.E.. Yao, X.L., Ramakrishna, N., Wolfe, G., Silverman, W.P. and Brown, W.T., Chromosome 21q21 subloealization of gene encoding beta-amyloid in cerebral vessels and neuritic (senile) plaques of people with Alz- heimer disease and Down syndrome, Lancet, 1 (1987) 384-385.

13 Haines, J.L., Nee, L., Watkins, B.C., Myers, R.H., Feldman, R.G., Pollen, D., Drachman, D., Grow- don, J., Bruni, A., Foncin, J.F, Salmon, D., Frommelt, P., Amaducci, L., Sorbi, S., Piacentini S., Ste- wart, G.D., Hobbs, W.J., Conneally, P.M. and Gusella, J.F., The genetic defect causing familial Alz- heimer's disease maps on chromosome 21, Science, 235 (1987) 885-889.

14 Tanzi, R.E., Gusella, J.F., Watkins, P.C., Burns, G.A.P., St George Hyslop, P., Van Keuren, M.L., Patterson, D., Pagan, S., Kurnitt, D.M. and Neve, R.L., Amyloid B protein gene: eDNA, mRNA dis- tribution and genetic linkage near the Alzheimer locus, Science, 235 (1987) 880-884.

15 Wisniewski, K.E., Dalton, A.J., Crapper McLachlan, D.R., Wen, G.Y. and Wisniewski, H.M., Alz- heimer's disease in Down's syndrome. Clinical pathologic studies, Neurology, 35 (1985) 957-961.