Neuronal nitric oxide synthase (nNOS) mRNA expression and NADPH-diaphorase staining in the frontal cortex, visual cortex and hippocampus of control and Alzheimer's disease brains

MOLECULAR BRAIN RESEARCH ELSEVIER

Molecular Brain Research 41 (1996) 36-49

Research report

Neuronal nitric oxide synthase (nNOS) mRNA expression and NADPH-diaphorase staining in the frontal cortex, visual cortex and hippocampus of control and Alzheimer’s disease brains Paula J. Norris a’*, Richard L.M. Faull b, Piers C. Emson a a MRC Molecular Neuroscience Group, Department of Neurobiology, The Babraham Institute, Babraham, Cambridge, CB24AT,

UK

b Department of Anatomy, School of Medicine, University of Auckland, Priuate Bag 92019, Auckland, New Zealand Accepted 13 February 1996

Abstract Neuronal nitric oxide synthase (nNOS) mRNA levels and NADPH diaphorase (NADPH-d) staining were compared in the frontal cortex, visual cortex and hippocampus (dentate gyms and CA subfields of Ammon’s horn) of five Alzheimer’s disease (AD) and six control brains. The cellular abundance of nNOS mRNA was quantified by in-situ hybridisation using 35S-labelled antisense oligonucleotides complementary to the human nNOS sequence. Although the mean level of nNOS expression was decreased in all three regions in AD cases as compared to controls, it did not reach significance. Neurones positively labelled for nNOS mRNA and neurones positive for NADPH-d histochemistry displayed similar distributions in control and AD cases. In AD brains the density of neurones having detectable levels of nNOS mRNA was significantly decreased in the white matter underlying the frontal cortex (P < 0.05) but not in the frontal cortex gray matter; no change was observed in the gray or white matter of the visual cortex in AD. The number of cells expressing detectable levels of nNOS mRNA in the hippocampus was also significantly decreased (P < 0.05) in AD. The density of NADPH-d-positive cells was not significantly decreased in the gray or white matter of the frontal or visual cortices in AD compared to controls; however, the number of NADPH-d-positive cells was significantly decreased in the hippocampus (P < 0.01). These data indicate that although the cellular abundance of nNOS mRNA is not significantly decreased in these three regions in AD, there is a significant decrease in the number of cells expressing detectable levels of nNOS mRNA in the white matter underlying the frontal cortex and in the dentate gyms and CA subfields of the hippocampus in AD. Furthermore,there was also a significant decrease in the number of NADPH-d-positive cells in the dentate gyros and CA subfields of the hippocampus in AD as compared to controls. These results suggest specific populations of nNOS/NADPH-d cells in the white matter underlying the frontal cortex and in the hippocampus are vulnerable in AD. The implications of these findings are discussed. Keywords: Neuronal nitric oxide synthase; NADPH-diaphorase histochemistry; Hybridisation, in-situ; Alzheimer’s disease; Excitotoxicity; Hippocampus; Cerebral cortex

1. Introduction Alzheimers disease (AD) is characterised by the presence of neurofibrillary tangles, senile plaques and neuronal degeneration. The parts of the brain that are most vulnerable include the hippocampus, parahippocampal gyms, entorhinal cortex and the arnygdaloid nuclei. The parahippocampal gyms receives connections originating in the association cortex of the temporal, parietal and frontal

* Corresponding author. MRC, Laboratory of Moleculw Biology, Neurobiology Division, Medical Research Council, Hills Road, Cambridge CB2 2QH, UK. Fax: (44) (1223) 402-310. 0169-328X/96/$15 .00 Published by Elsevier Science B.V. PII SO 169- 328 X(96 )OO064-2

lobes and these areas of cortex are also severely affected in AD. The neuronal isoform of the enzyme nitric oxide synthase (nNOS) and NADPH-diaphorase (NADPH-d, a histochemical marker for NOS-containing cells) [24,16,33] have been shown to co-localize with somatostatin (SOM) immunoreactivity and neuropeptide Y (NPY) immunoreactivity in several forebrain regions in the rat and human brain including the cerebral cortex, neostriatum and amygdala [45,46,28,43,19,39]. In the cerebral cortex nNOS/NADPH-d, NPY and SOM are reported to have greater than 90% co-localisation in non-pyramidal neurones [27] and they are estimated to constitute approximately 2% of all neurones in the cerebral cortex [44]. The

P,J. Norri.ret al./Molecular Brain Research 41 (1996) 36-49

37

activation of the enzyme nNOS is dependent on calcium concentration and a number of other co-factors [6]; once activated nNOS converts the substrate L-arginine to Lcitrulline and nitric oxide (NO). NO has been implicated in excitotoxicity. NO released by nNOS-containing neurones has been shown to mediate the neurotoxicity induced by NMDA in cultured cerebral cortical neurones [17]. Although neurones containing nNOS/NADPH-d are reported to be relatively resistant to degeneration in Huntington’s disease (HD) [20], in AD [25], ischaemia and neurotoxin-induced insults [3,42,12], NADPH-d-containing neurones have been shown to undergo dystrophic changes in the cerebral cortex and the hippocampal formation in patients with AD [27,25,43,31]. In a recent study on HD we found that the abundance of nNOS mRNA is decreased in the striatum, particularly in the dorsal regions of the striatum which usually show more advanced pathology [34]. The reported sparing of NADPH-d neurones in AD and the potential involvement of NO in the pathology of this disease led us to investigate whether nNOS gene expression is altered in the frontal cortex, and hippocampus

New Zealand Neurological Foundation Brain Bank. Control brains were collected from patients with no history of neurological disease. AD brains were collected from patients with a clinical history of AD. In each AD case the clinical diagnosis of AD was confirmed by neuropathological analysis (Dr J.H. Xuereb), all AD brains showed extensive neurodegeneration and contained neurofibriktry tangles and neuritic plaques. Upon receipt of the brains, the brainstems were removed and the hemispheres separated by cutting through the corpus callosum. Blocks containing the frontal cortex (Brodmann areas 9, 10 and 11) and visual cortex (Brodmann areas 17 and 18) and the hippocampus were cut from one hemisphere and frozen rapidly on dry ice and stored at –80”C. Subsequently, cryostat sections of the frontal and visual cortices and hippocampus were cut at a thickness of 16 p,m and thaw mounted on to gelatin/chrome-alum-coated glass slides. These slides were kept at –70°C until processed for NADPH-d histochemistry or in-situ hybridisation. The details of the tissue used in this study are contained in Table 1.

(regions which are particularlyvulnerableto pathological changesin AD) relative to an area that is less damaged, the

2.2. Tissue pH

visual cortex, in AD; we also examined the distribution of cells positive for nNOS gene expression in these regions in control and AD brains. For comparative purposes consecutive sections from each region have also been processed for NADPH-d histochemistry, a marker for NOS-containing cells.

Tissue pH determination was carried out using the method of Kingsbury et al. [26]. Briefly, blocks of tissue were cut and homogenized, from frozen, in 10 vol. of distilled water at neutral pH. pH values were measure in duplicate at room temperature using a standard electrode. 2.3, NADPH diaphorase histochemistry

2. Materials and methods 2.1. Tissue collection Postmortem human brains used in this study were obtained from the Cambridge Brain Bank laboratory and the

This was performed as before with some modifications [32]. Frozen sections were quickly warmed up to room temperature using a hair drier, then fixed with 490 paraformaldehyde in 0.1 M phosphate buffer (PB, pH 7.4) for 10 min. Fixed sections were rinsed 3 times in 0.1 M PB and then incubated at 37°C for 50 rnin in 0.1 M PB

Table 1 Source of brain tissues Case no. Control cases H73 HIS WI H78 H79 H81 AD cases FD142 FD144 FD145 FD154 FD161

Age/sex

Postmortem delay(h)

pH/tissue

Cause of death

66/M 67/F 75/M 48/F 75/M 55/M

7.5 8.5 3.5 11.5 11 10.5

ND ND 6.3/CB 6.3/CB 6.7/P 6.8/cB

myocardiat infarction myocardial infarction myocardial infarction myocardiat infarction myocardial infarction myocardial infarction

40/M 84/M 69/M 69/F 69/M

7 5 4 9 13

6.2/CB 6.7/CB 6.6/cB 6.2/FC 6.7/CB

ND ND ND senile dementia bronchopneumonia

ND, not determined;CB, cerebella cortex;FC, frontalcortex;P, pens.

38

P.J, Norris et al. /Molecular Brain Research 41 (1996) 36-49

containing 0.370 Triton X-1OO,0.1 mg/ml Nitroblue tetrazolium and 1.0 mg\ml ~-NADPH. Following this, the sections were rinsed in 0.1 M PB rinsed in distilled water and then coverslipped with glycerin gel. 2.4. Preparation of probes For the detection of nNOS mRNA, three antisense deoxyoligonucleotide sequences (BB149, BB150, BB152, made by British Biotechnology, UK) which correspond to the bases 1181–1225, 3876–3930 and 3986–4030 of the human nNOS mRNA were used, as described previously [34]. The radiolabelled probes were purified by gel filtration on Sephadex G-50. For nNOS in-situ hybridisation the three probes were mixed and used together to increase the sensitivity of the technique. 2.5. In-situ hybridisation In-situ hybridisation was performed as described earlier [32] with some slight modifications. Frozen sections were fixed with 4% paraforrnaldehyde in 0.1 M PB for 30 min at room temperature, rinsed twice with 0.1 M PBS pretreated with O.1~0diethylpyrocarbonate (DEPC). Sections were dehydrated, air dried and then hybridised for 17 h at 37°C in the following buffer; 4 X standard saline citrate (SSC), 50% formamide, 1 X Denhardt’s solution, 500 wg/ml sheared salmon sperm DNA, 0.25 mM dithiothreitol, and 10Yodextran sulphate. The 35S-labelled probes were used at a concentration of approximately 4 fmol/@. In control experiments a 100-fold excess of unlabeled probe was added to the hybridisation buffer. Hybridised sections were dipped in 1 X SSC/O.00190 @mercaptoethanol to remove the hybridisation buffer, washed stringently 4 times in 1 X SSC/0.00IYo ~-mercaptoethanol pre-heated to 55°C for 30 min each and washed again in 1 x SSC/O.001% ~-mercaptoethanol for 1 h at room temperature. Sections were then rinsed briefly in distilled water, followed by 70% ethanol/300 mM ammonium acetate and finally rinsed briefly in absolute ethanol and air dried before exposing to film (Hyperfilm, B-max) for autoradiography. The sections were then processed by emulsion autoradiography, sections were exposed to emulsion for 9 weeks. 2.6. Quant#ication of autoradiographic labelling Microscopic analysis of silver grains overlying neurones was performed using a computerised image analysis system (Seescan Bioscience, UK) linked to a Leitz microscope [2]. For each control and AD case nNOS mRNAlabelled cells in the hippocampus and frontal and visual cortices were analysed. In the frontal and visual cortices an area of approximately 65 mmz was drawn on the coverslip covering the brain section; all positively labelled cells in this area (cortical gray and underlying white matter in-

cluded) were analysed. Emulsion-coated brain sections were viewed under bright-field illumination (Leitz microscope) using a 40 X objective and a blue gelatin filter to enhance the field image contrast. Initially the threshold values were determined so that only silver grains overlying methylene blue-counterstained cells were detected; these values then remained set for the duration of the analysis. Live on-line microscopic images were captured by a video camera (Sony XC-77CE), digitized and displayed on a monitor. To determine the intensity of labelling of a cell the density of silver grains was measured. With the aid of a cursor the outline of the cell cytoplasm was traced and the number of silver grains overlying the cytoplasm was then expressed as the number of pixels per cell. Under these conditions there is a good correlation between the number of pixels and the number of silver grains [38]. Background levels were assessed by measuring the density of autoradiographic grains in random locations over the tissue. Differences in grain densities between control and AD were assessed with the Mann-Whitney U statistic, a non-parametric test which compares the mean values obtained for each individual in each experimental group [40]. 2.7. Quant@cation of cell density The number of nNOS rnRNA-labelled cells and the number of NADPH-d-positive cells in each region was counted manually under a light microscope at 25 X objective using a 1 cm2 graticule. In the frontal and visual cortices the number of labelled cells in the gray and white matter within the boundary of the 65 mm2 area were counted. The area of cortical gray and the underlying white matter were subsequently measured accurately using the Seescan Image Analysis system (Seescan Bioscience) and the densities of cells/mm2 calculated. Differences in cell densities between control and AD were assessed as before with the Mann-Whitney U statistic.

3. Results 3.1. Tissue pH Tissue pH has previously been established to be an indicator of mRNA preservation in human postmortem brain [26]. There was no significantdifference in tissue pH between the control and AD cases; the mean pH value for the control cases was 6.52 (S.D. 0.26), whilst the mean pH value for the AD cases was 6.48 (S.D. 0.25). 3.2. In-situ hybridisation: specijicip controls The three nNOS specific oligonucleotides, complementary to different regions of the same target mRNA have been used separately and together on consecutive sections, the hybridisation pattern was the same with each of the

P.J. Norris et al./Molecular Brain Research 41 (1996) 36-49

39

probes (data not shown). An excess of each unlabeled nNOS or SOM oligonucleotide probe (400 fmol/pl) blocked the hybridisation signal detected by the identical 35S-labelled probes(s) (Fig. 1).

Table 2 Density of grains overlying labelled cells, detected by in-situ hybridisation in the frontal cortex, visual cortex and hippocampus of control of AD brains Brain region

Control (n= 6)

AD (n = 5)

3.3. Quantitative analysis of NOS mRNA expression

Frontal cortex Visual cortex Hippocampus

1.00+0.08\pm2 0.44 f0.05/pm2 0.20 *0.03/pm2

0.78 +0.08\~m2 0.28 ~0.12/pm2 0.12 *0.07/pm2

Table 2 shows the mean densities of silver grains overlying labelled cells in six control and five AD cases. In the AD group the mean density of grains overlying labelled cells was lower than the control group in the fi-ontal and visual cortices and hippocampus by 22%, 36% and 40%, respectively (Table 2); however, this did not reach significance. In Fig. 2 the frequencies of labelling densities are compared in control and AD brains. The graphs indicate a decrease in the total number of labelled cells in the frontal and visual cortices and hippocampus in AD; however, they do not indicate a change in the density

Values represent meant SEM.

of labelling per cell in these regions in AD. In the frontal cortex highest levels of expression were 0.6 grains/pm2 for both control and AD (Fig. 2a,b), whilst in the visual cortex the highest level of expression was lower at 0.3 grains/~m2 (Fig. 2c,d). The lowest levels of Iabelling were found in the hippocampus where the levels of expression only reached 0.1 and 0.25 grains/pm2 in control and AD, respectively (Fig. 2e,f). 3.4. Distribution of NADPH-d-positiue cells and nNOS mRNA-labelled cells in the frontal and uisual cortices In both control and AD brains NADPH-d-positive neurones were found in layers II–VI of the frontal and visual cortices and in the white matter underlying these structures (Figs. 3 and 4). In the frontal cortex of AD brains some neurones showed distortion and puning of fibre plexuses (Fig. 3c). In the frontal cortex of the control group approximately 4.3?Z0,26.OYO and 69.670 of positive neurones were found in layers 11/111,layers IV/V\VI and the subcortical white matter, respectively. In the visual cortex a similar distribution of positive cells was detected, there were 6.2%, 6.8% and 86.8% in layers 11/111,layers IV/V/VI and the subcortical white matter, respectively. Many NADPH-d-positive and nNOS mRNA-labelled cells were located in the subcortical white matter, these were mostly bipolar cells (Fig. 4d,e) although some multipolar forms were also evident (Fig. 4a,b,c). NADPH-d-positive cells were generally intensely stained although occasionally small oval lightly stained neurones were observed in layers II and III. Neuronal NOS mRNA-containing neurones had a similar distribution to NADPH-d-positive neurones and were detected in layers II–VI and in the underlying white matter in both the frontal and visual cortices of control and AD brains (Fig. 5). 3.5. Densities of nNOS-labelled neurones and NADPH-dpositive neurones in the frontal and visual cortices The density of cells expressing nNOS mRNA and having NADPH-d histochemistry were measured in consecutive sections. The number of positive cells in the frontal and visual cortices were counted and the density of cells per mm2 of gray and white matter was calculated (Table

40

P.J. Norris et al./Molecular Brain Research 41 (1996) 36-49

3). The density of NADPH-d-positive neurones counted in the frontal and visual cortices were generally higher than the density of neurones containing detectable levels of nNOS mRNA on consecutive sections. There was no significant difference in the density of NADPH-d neurones

A

.2,

20

in the frontal cortex gray matter or underlying white matter or visual cortex gray or underlying white matter in AD as compared to controls (Table 3, Fig. 6B). However, the mean density of neurones positive for nNOS mRNA was significantly reduced (P < 0.05) in the white matter under-

B z,

1

20

1

silver grains/#m2

❑

Control

HAD

silver grains/pm2

D z, 20

5 0

silver grains1~2

silver grainslpm2

50 0

silver grains/pm2

I CQO-l=tvlwt-ul :Ooooood

I

I

~

silver grains@n2

Fig. 2, Frequency histograms showing the density of silver grains labelling cells in control and AD brains. A,B: Frontal cortex. C,D: Visual cortex; E,F: hippocampus.

P,J. Norris et al./Molecular Brain Research 41 (1996) 36–49

lying the frontal cortex in AD as compared to the control group (Fig. 6A); in all other regions of the frontal and visual cortices (gray and white matter) there was no alteration in the density of cells in AD (Table 3, Fig. 6A). 3.6. Distribution and nwnber of NOS and NADPH-d cells in the hippocampus For hippocampal sections the number of cells in the hippocampal formation were counted and their position mapped (Fig. 7). All the AD hippocampi showed signs of atrophy and the gyri were shrunken compared to controls (Fig. 7). In both control and AD hippocampi there were substantially higher numbers of NADPH-d-positive cells detected than cells labelled for nNOS rnRNA in consecutive sections. For example in one control case (H81), 67 NADPH-d-positive cells were detected and only two cells expressing detectable levels of nNOS mRNA were recorded (Fig. 7). Unlike the NADPH-d-positive cells in the frontal and visual cortices which were intensely reactive, the

41

reactive cells in the hippocampus were small, oval or round in shape, only lightly stained and there was little or no staining of processes (Fig. 8). The nNOS mRNAlabelled cells detected in the hippocampus were also oval or round in shape and only lightly labelled (Fig. 9) compared to cells in the frontal and visual cortices. In both control and AD brains NADPH-d-positive and nNOS mRNA-labelled neurones were localised primarily in the polymorph layer of the dentate gyms, adjacent to the granule cell layer (none of the granule cells appeared to be labelled) and a few cells were occasionally detected in more central areas of the polymorph layer (Figs. 7–9). NADPH-d- or nNOS rnRNA-positive cells were also detected in the stratum pyramidal and occasionally the stratum radiatum in the subfields CA1–CA3 of Ammon’s horn (Fig. 7). In AD the number of nNOS mRNA-positive cells was significantly decreased (P < 0.05) in the hippocampus as compared to control brains (Table 4), with only two out of five cases having any detectable labelled cells (Fig. 7). There was also a very significant decrease

Fig. 3. Photornicrographs showing cells stained by NADPH-diaphorase histochemistry in control (a,b) and AD (c,d) brains. Sections have been counterstained with Mayer’s Carrrrahrm to visurdize cell bodies. a: NADPH-d-stainedneuronein layer IV of the frontal cortex. b: NADPH-d-stained neuronesin layer VI of the frontal cortex.c: NADPH-d-stainedneuronein layer 111of the frontalcortex showing some distortion and pinning of fibre plexuses. d: NADPH-d-stained neurone in layer II of the frontal cortex. Scale bar= 50 ~m.

42

P.J. Norris et al. /Molecular Brain Research 41 (1996) 36-49

Fig. 4. Photomicrographs showing the distribution of NADPH-d-positive cells in the white matter underlying the frontal cortex in control (a,b,c) and AD (d,e) brains. a: Low-magnification view of the distribution of NADPH-d-positive cells in a control brain. In (b) and (c) the two neurones identified by arrows in (a) are shown at higher magnification. d: Low-magnification view of the distribution of NADPH-d-positive cells in an AD brain. In (e) the neurone identified by an arrow in (d) is shown at higher magnification, Sections have been counterstained with Mayer’s Carmalum to visualize cell bodies. Scale bar (a,d) = 100 ~m; (b,c,e) = 50 p.m.

(F’ < 0.01) in NADPH-d-positive cells in the hippocampus in AD as compared to controls, with only three out of five cases having any detectable labelled cells (Table 4, Fig. 7).

4. Discussion In the control brains the distribution of NADPH-d-positive neurones and the distribution of nNOS mRNA-labelled

neurones across the layers of the frontal and visual cortices and particularly the prominent distribution of cells in the subcortical white matter was consistent with the findings of others [27,43,21]. Quantitative in-situ hybridisation revealed that the level of nNOS mRNA labelling per cell (grains/pm*) was not significantly altered in the frontal or visual cortices in AD. However, the general level of labelling was lower in the visual than frontal cortex in both controls and AD cases.

Table 3 Density of nNOS mRNA-containing neurones as detected by in-situ hybridisation and density of NADPH-d-positive neurones in the frontal and visual cortices of control and AD brains Brain region

Frontal cortex gray matter white matter Visual cortex gray matter white matter

Density of neurones positively labelled for nNOS mRNA

Density of NADPH-d-positive neurones

Control

AD

Control

AD

0.40 * o.lo/mm* 0.90 ~ 0.28/mm2

0.15 + o.03/mm2 0.23 ~ 0.13/mm2 *

0.34 i o.05/mm2 2.78 + 0.77/mm2

0.34 i 0.05/mm2 2.20 ~ 0.60/mm2

0.15 + 0.05\mm2 1.10 * o.39/mm2

0.02 + o.ol\mm2 0.58 + 0.35/mm2

0.09 + 0.02/mm2 2.65 + 0.72mm2

0.10 + 0.04/mm21 2.60 + 0.65/mm2

Values represent mean + SEM. Controls, n = 6; AD, n = 5 except for T where n =4. ‘ P <0.05 compared to respective controls.

P.J. Norris et al./Molecular Brain Research 41 (1996) 36–49

43

Fig. 5. Photomicrographs showing silver grains labelling cells in sections processed for in-situ hybridisation with probes specific for nNOS mRNA, a: Low-magnification view of the distribution of nNOS mRNA-positive cells in the white matter underlying the frontal cortex in a control brain. b: One of the positive neurones in (a) is shown at higher magnification. c: Low-magnification view of the distribution of nNOS mRNA-positive cells in the white matter underlying the frontal cortex in an AD brain, d: The positive nenrone in shown in (c) is shown at higher magnification. Sections have been counterstained with methylene blue to visualize cell bodies. Scale bar (a,c) = 100 ~m; (b,d) = 50 pm,

A

B 42-

3 E

1.5-

m ~

❑ Control ■

AD

T

1-

,:,:,:,:, ,,..,.,,, .,.,., .,, ,.,.,.,, , ,.,.,,,,.

u 0,5-

0

T

;:::;:;:; .........

3-

T ,.,,.,.,, ,,,.,..,, ,.,,,,, ,, ,.,., ,,.,

2-

;:::::::. ;:::::::: ,,.,.,,,,

E T % + u

,:,:.:.:.

1-

......... .,.,., ,,, ;:::::::. ......... :::::::::

0

Fig. 6. Graphs showing the density of nNOS mRNA-labelled cells (A) and NADPH-d-positive cells (B) in the frontal cortex (FC) gray and white matter (win) and the visual cortex (VC) gray and white matter (win) in control and AD brains. Values represent mean+ S.E.M. “ P <0.05.

44

P.J. Norris et al./Molecular Brain Research 41 (1996) 36-49

In the cerebral cortex nNOS/NADPH-d, NPY and SOM show more than 90Y0 co-localisation in non-pyramidal cells [27]. A number of studies have reported that

SOM concentrations are depleted in the cerebral cortex of AD cases [15,22,37,9,5,14]. However, despite its frequent co-localisation with SOM, NPY levels are generally found

NOS in situ

NADPH-d

NADPH-d

NOS in situ

FD142

H81 . . .

. .

.

7

.

.

,. 3

.7 L

. ..

+

.“ . ... . mw4

.“ . ,. .,

(&# “.

‘*@ ..

cN~@AI

.“

..

“.

‘ .. . .

0.2.

: ..

.

,. & & H7S

. . Wq H75

“..

“.

e’? (&/ . .

+=’

&

CA3

“. ,.

...

..

mi6i

CA3

. ... .

.

CA3

00 .

.

Vq

.

H78 .! .X3

.. “.. ..

“ . . .. &@... ” .’

H77

, .

(’g?OG

. .. ,.. .. u...

.. T

-c....

..... ... -. T Jti

CA1

Fig. 7. Maps showing the distribution of NADPH-d-reactive cells in the human hippocampus. The case numbers are shown at the left-hand side of the diagrams; H, denotes control brain; FD, denotes an AD brain. Each dot represents one NADPH-d-reactive neurone or one nNOS mRNA-labelled neurone. CA, cornrr ammonis; DG, dentate gyms. Scale bar= 5 mm.

P..l. Norris et al. /Molecular Brain Research 41 (1996) 36-49

to be normal in AD [15,22], although they were reduced in one study [4]. Gasper et al. [23] and Unger and Lange, [43] have suggested that the apparent dissociation between SOM and NPY levels in AD is due to a subset of SOM neurones (approximately 20~0 of SOM cells, which do not have NPY irrtmunoreactivity) being particularly vulnerable in AD. This subpopulation of SOM immunoreactive cells which was not stained by NADPH-d histochemistry was found predominantly in layers II, III and V of the cerebral cortex [23]. Doumaud et al. [18] found an overall reduction in the density of SOM mRNA-labelled cells in the frontal cortex gray and white matter in AD, however, due to variation among the brains these values did not reach significance. In the present study we observed no significant alteration in the density of NADPH-d-positive neurones or nNOS rnRNA-labelled neurones in the gray matter of the frontal or visual cortices in AD, thus supporting the idea that these nNOS/NPY/SOM cells are resistant to neurodegeneration in AD.

45

Although NPY/SOM/NOS cells are apparently less affected in AD a number of studies have reported dystrophic changes in these neurones including cell shrinkage and pruning or distortion of fibre plexuses [11,27,43]. We found a significant decrease in the density of nNOS mRNA-labelled cells, but not NADPH-d-positive cells in the white matter underlying the frontal, but not visual, cortex in AD disease. This finding compares well with the neuropathological characteristics of AD which suggest the frontal cortex is more susceptible to degeneration than the visual cortex. The morphology of the NADPH-d-positive cells in the subcortical white matter was very similar to that of SOM- and substance P-imrnunoreactive neurones described by Ang and Shul [1]. These neurones were also found in the subcortical white matter and their densities were reduced in AD [1]. The similarity between the morphology of these neurones and their susceptibility in AD suggests nNOS may co-localize with substance P and/or SOM in the subcortical white matter. These neurones

Fig. 8. Photornicrographs showing neurones stained by NADPH-d histochemistry in the hippocampus of control (a,b) and AD (c,d) brains. Sections have been counterstained with Mayer’s Carmafum to visuafize cell bodies. NADPH-d-positive cells are indicated by arrows. PO,polymorph layer of the dentate gyms; gr, granule cell layer of the dentate gyms. Scale bar (a–c) = 50 ~m; (d)= 100 pm.

46

P.J. Norris et al. /Molecular Brain Research 41 (1996) 36-49

Fig. 9. Photomicrographs showing nNOS mRNA-positive nenrones (indicated by arrows) in the hippocampus of control (a–c) and AD (d) brains. Sections have been counterstained with methylene blue to visualize cell bodies. PO,polymorph layer; gr, granule cell layer of the dentate gyms. Scafe bar= 50 pm.

which are found in the white matter are known as ‘interstitial cells’ [8]. These cells are believed to be the residual of the early generated subplate neurone population; they extend their processes into the cortical gray [21] and it has been postulated that they may be essential for the maintenance of normal signaling in the adult cerebral cortex [13].

Clearly if this is the case then the loss of such interstitial neurones would have important implications for the function of normal neuronal processes and may help us to understand the deficits which arise and are associated with AD. In the hippocampus lightly stained NADPH-d-positive

Table 4 Number of NOS mRNA-containing neurones as detected by in-situ hybridisation and density of NADPH-d-positive neurones in the hippocarnpus of control and AD brains Brain region

Hippocampus

Density of mRNA positive neurones

(DG and CA subfields)

Density of NADPH-d-positive neurones

Control

AD

Control

AD

11.33 * 4.05

2.80 + 2.56 *

59.17 k 2.77

7.00 + 3.29 ‘ *

Vafues are total number of positive cells detected in the dentate gyms (DG) and CA subfields. Values represent mean + SEM. Controls, n =6; AD, n =5. * P <0.05 compared to respective controls. * * P <0.01 compared to respective controls.

P.J. Norris et al./Molecular Brain Research 41 (1996) 36–49

cells and weakly Iabelled nNOS mRNA cells were found in the polymorph layer of the dentate gyms. A few NADPH-d- and nNOS mRNA-positive cells were also located in the subfields CA1–CA4 of Ammon’s horn. The morphology and distribution of these cells resembled very closely the distribution of NADPH-d-positive neurones previously described in the human hippocampus [41] and in the hippocampus of the New World Monkey (Sakniri $czhreu.r) [30]. In the hippocampus the numbers of nNOS mRNAlabelled cells and NADPH-d-positive cells were significantly reduced in AD. Fewer nNOS-positive cells were detected in both the polymorph layer of the dentate gyms and the CA subflelds of Ammon’s horn. These results are in contrast to those of others [36,25] who report a similar distribution of NOS-positive cells in the hippocampus of AD and control patients using NADPH-d staining and NOS immunohistochemistry, respectively, although they did report changes in morphology of NOS neurones. In 1987 Chan-Palay [10] reported on the distribution of SOM-immunoreactive neurones in the human hippocampus and their co-existence with NPY neurones. The distribution of SOM- and NPY-immunoreactive neurones reported was remarkably similar to our observations of nNOS neurones in the human hippocampus. NPY- and SOM-immunoreactive neurones were most abundant in the hilus and subgranular zone of the dentate gyms and the stratum oriens and pyramidal of CA1. In double-labelling experiments they showed that approximately 3090 of SOM-immunoreactive neurones in the hilus of the dentate gyms are also NPY immunoreactive; there was also some co-existence in CA1 neurones. Chan-Palay [10] also found a significant loss of SOM-immunoreactive neurones and axons in the hippocampus of AD compared to those of age-matched control brains. This was particularly severe in the hilus of the denate gyms and CA1 and surviving neurones were distorted. Neuronal NOS is known to colocalize with SOM and NPY in many regions (cortex and striatum) and it seems possible from the similar distributions and loss in AD that nNOS neurones may co-localise with SOM and or NPY in the dentate gyms and perhaps some neurones in the CA subfields of Ammon’s horn. Information on the co-localisation of NADPH-d/nNOS with neuropetides in the human hippocampus is not yet available; however, Mufson et al. [30] report preliminary findings suggesting many NADPH-d-positive neurones co-localise with SOM in the hippocampus of the New World monkey (Saivtiri sciureus). Doumaud et al. [18] examined SOM mRNA-containing neurones in the hippocampus and frontal cortex of control and AD brain. In this study there was no significant change in the density of labelled cells in these regions in AD, although there was a significant decrease in expression of SOM mRNA in the hippocampus, but not the frontal or parahippocampal cortices. It is not clear why the density of NADPH-d-positive

47

neurones is generally greater than the density of nNOS mRNA-positive neurones in the regions examined in this study. Whilst in-situ hybridisation allows us to determine the cellular abundance of nNOS mRNA in a particular cell, NADPH-d histochemistry is a method by which we can visualize the nNOS protein (or at least it requires the C-terminal portion of the enzyme responsible for electron transfer). It is possible that neurones that have low levels of nNOS mRNA are below the limits of detection, although they may still have relatively robust NADPH-d staining and this may lead to a dissociation between the results of the two techniques. We have found similar dissociations between nNOS gene expression and NADPH-d staining in a recent study on HD brain tissue [34]. NO is an intercellular signaling molecule; it is highly reactive and readily diffusible [7]. The reduced (NO”) form of the molecule reacts with superoxide anion to form peroxynitrite anions [29] that decompose to form highly reactive hydroxyl free radicals and nitrogen dioxide. The brain is particularly vulnerable to free radicals as the normally occurring antioxidant defence mechanisms are relatively deficient [35]. In this study we have shown that nNOS gene expression is not significantly altered in the frontal or visual cortices or the hippocampus in AD, hence the results do not suggest an up-regulation of nNOS activity in AD although we cannot rule out a role for NO in the neuropathology of AD. The decrease in density of nNOS mRNA-labelled neurones in the white matter of the frontal cortex and the decrease in numbers of positive cells in the hippocampus suggests nNOS cells in these regions are particularly affected by the pathology of AD. It seems likely that the apparent loss of these cells in the frontal cortex (subcortical white matter) and in the hippocampus (dentate gyms and CA subfields) may contribute to the impairment of functions associated with these regions (foresight and judgement, and learning and memory, respectively) which are seen in AD.

Acknowledgements We thank the Cambridge Brain Bank Laboratory and the New Zealand Neurological Foundation Brain Bank for postmortem human brain tissue. We would also like to thank Dr. I. Charles, Wellcome Research Laboratories, UK for nNOS-specific deoxyoligonucleotides, Mr. David Brown for advise on statistical analysis and Mr. Ian King for emulsion autoradiography. This work was supported by the Medical Research Council, UK.

References [1] Ang, L.-C. and Shul, D.D., Peptidergic neurons of subcortical white matter in aging and Alzheimer’s disease. Brain Res, 674 (1995) 329–335.

48

P.J. Norris et al./Molecular Brain Research 41 (1996) 36–49

[2] Augood, S.J., Herbison, A.E. and Emson, P.C., Localization of GAT-1 GABA transporter mRNA in rat striatum: cellular coexpression with GAD67 mRNA, GAD67 immunoreactivity, and parvalbumin mRNA. J. Neurosci., 15 (1995) 865–874, [3] Beal, M.F., Kowall, N.W,, Swartz, K.J., Ferrante, R.J. and Martin, J.B,, Differential sparing of somatostatin-neuropeptide Y and cholinergic neurons following striatal excitotoxin lesions. Synapse, 3 (1989) 38-47. [4] Beal, M.F., Mazurek, M.F., Chattha, G.K., Svensden, C.N,, Bird, E.D. and Martin, J.B., Neuropeptide Y immunoreactivity is reduced in cerebral cortex in Alzheimer’s disease. Ann. Neurol., 20 (1986) 282-288. [5] Beal, M.F., Mazurek, M.F., Svensden, C.N., Bird, E.D. and Martin, J.B., Widespread reduction of somatostatin-like immunoreactivity in the cerebral cortex in Alzheimer’s disease. Ann. Neurol., 20 (1986) 489-486. [6] Bredt, D.S. and Snyder, S.H., Isolation of nitric oxide synthase, a calmodtdin-requiring enzyme. Proc. Natl. Acad. .Wi. USA, 87 (1990) 682-685. [7] Bredt, D.S. and Snyder, S,H., Nitric oxide, a novel neuronal messenger. Neuron, 8 (1992) 3-11, [8] Cajal, S.R., Histologic du Systeme Neerueux de l’Homme et des Vertebres. Malonine, Paris, (1911 (reprinted Madrid, 1952)). [9] Candy, J.M,, Cascoigne, A.D., Biggins, J.A., Smith, Al., Perry, R.H., McDermott, J,R. and Edwardson, J.A., Somatostatin immunoreactivity in cortical and some sub-cortical regions in Alzheimer’s disease, J. Neurol. Sci., 71 (1985) 315-323. [10] Chan-Palay, V., Somatostatin immunoreactive neurons in the human hippocampus and cortex shown by imrmrnogold/silver intensification on vibratome sections: coexistence with neuropeptide Y neurons, and effects in Alzheimer-type dementia. J. Corrrp,Neurol., 260 (1987) 201-223. [11] Chan-Pafay, V., Lang, W., Allen, Y.S., Haesler, U. and Polak, J.M., Cortical neurons immunoreactive with antisera against neuropeptide Y are altered in Alzheimer’s type dementia. J. Comp. Neurol., 238 (1985) 390-400. [12] Choi, D.W., Nitric oxide: foe or friend to the injured brain?. Proc. Natl. Acad. Sci. USA, 90 (1993) 9741–9743. [13] Chun, J.J,M. and Shatz, C.J., Interstitial cells of the adult neocortical white matter are the remnant of the early generated subplate neuron population, J. Cornp,Neurol., 282 (1989) 555-569, [14] Davies, P., Katzman, R. and Terry, R.D,, Reduced somatostatin-like immunoreactivity in cerebral cortex from cases of Alzheimer’s disease and Alzheimer senile dementia. Nature, 288 (1980) 279-280. [15] Dawbarn, D., Rossor, M.N., Mountjoy, C.Q., Roth, M, and Emson, PC. (1986) Decreased somatostatin immunoreactivity but not neuropeptide Y immunoreactivity in cerebral cortex in senile dementia of Alzheimer type. Neurosci. Lett., 70, 154–159. [16] Dawson, T.M., Bredt, D.S., Fotuhi, M., Hwang, P.M. and Snyder, S.H., Nitric oxide synthase and neuronal NADPH diaphorase are identical in brain and peripheral tissues. Proc. Natl. Acad, Sci. USA, 88 (1991b) 7797-7801. [17] Dawson, V.L., Dawson, T,M., London, E.D,, Bredt, D,S, and Snyder, S,H., Nitric oxide mediates glutamate nenrotoxicity in primary cortical culture. Proc. Natl. Acad. Sci. USA, 88 (1991) 6368–6371, [18] Doumaud, P., Cervera-Pierot, P,, Hirsch, E., Javoy-Agig, F,, Kordon, C.L., Agid, Y. and Epelbaum, J., Somatostatin messenger RNA-containing neurons in Alzheimer’s disease: an in situ hybridization study in hippocampus, parahippocampal cortex and frontal cortex. Neuroscience, 61 (1994) 755-764. [19] Emson, PC,, Augood, S.J., Senaris, R., Guevara Guzman, R., Kishimoto, J., Kadowaki, K,, Norris, P,J. and Kendrick, K.M., Chemical signaling in striatal interneurones. In: G.W. Arbuthnot and P.C. Emson (Eds.), Chemical Signaling in che Basal Ganglia. Prog. in Brain Research, 1993, pp. 155–166. [20] Ferrante, R,J., Kowall, N.W,, Beal, M.F,, Richardson, E.P. and

Martin, J.B,, Selective sparing of a class of striatal neurons in Huntington’s disease. Science, 230 (1985) 561-563. [21] Fischer, H.C, and Kuljis, R.O., Multiple types of nitrogen monoxide synthase-\NADPH diaphorase-containing neurons in the human cerebral neocortex. Brain Res., 654 (1994) 105–1 17. [22] Gabriel, S.M., Bierer, L.M., Harotunianm, V., Purohit, D.P., Perl, D.P. and Davis, K.L., Widespread deficits in somatostatin but not neuropeptide Y concentrations in Alzheimer’s disease cerebral cortex. Neurosci. Lett., 155 (1993) 116–120. [23] Gaspar, P., Duyckaerts, C., Febvret, A., Benoit, R., Beck, B. and Berger, B., Subpopulations of somatostatin 28-immunoreactive neurons display different vulnerability in senile dementia of the Alzheimer type. Brain Res., 490 (1989) 1-13. [24] Hope, B.T., Michael, G. J,, Knigge K. M. and Vincent S. R., Neuronal NADPH diaphorase is a nitric oxide synthase. Proc, Natn, Acad, Sci. USA, 88 (1991) 2811-2814. [25] Hyman, B,T., Marszloff, K., Wenniger, J.J., Dawson, T.M., Bredt, D.S. and Snyder, S.H., Relative sparing of nitric oxide synthase-containing neurons in the hippocampal formation in Alzheimer’s disease. Ann. Neurol., 32 (1992) 818–820, [26] Kingsbury, A.E,, Foster, O.J.F., Nisbet, A.P., Cairns, N., Bray, L,, Eve, D.J., Lees, A.J. and Marsden, CD., Tissue pH as an indicator of mRNA preservation in human postmortem brain. Mol. Brain Res., 28 (1995) 311-318. [27] Kowall, N,W. and Beal, M.F., Cortical somatostatin, neuropeptide Y, and NADPH diaphorase neurons: normal anatomy and alterations in Alzheimer’s disease, Ann, Neurol., 23 (1988) 105–1 14. [28] Kowall, N.W., Ferrante, R,J., Beal, M,F,, Richardson, E.P., Sofroniew, M,V., Cuello, A.C. and Martin, J.B., Neuropeptide Y, somatostatin, and reduced nicotinamide adenine dinucleotide phosphate diaphorase in the human striatum: a combined immunocytochemical and enzyme histochemical study. Neuroscience, 20 (1987) 817-828, [29] Lipton, S.A., Choi, Y.-B,, Pan, Z.-H., Lei, S.Z., Chen, H,-S.V., Sucher, NJ., Loscalzo, J., Singel, D.J. and Stamler, J.S., A redoxbased mechanism for the neuroprotective and neurodestroctive effects of nitric oxide and related nitroso-compounds. Nature, 364 (1993) 626-632. [30] Mufson, E.J., Brady, D.R. and Carey, R.G., Reduced nicotinamide adenine dinucleotide phosphate-diaphorase (NADPH-d) histochemistry in the hippocampal formation of the new world monkey (Saimiri sciureus). Brain Res., 516 (1990) 237-247. [31] Nakarmrra, S,, Kawamata, T., Akiguchi, I., Karneyama, M, and Kimura, H., Reduced nicotinamide adenine dinucleotide phosphate diaphorase histochemistry in neocortex and hippocampus in patients with Alzheimer type dementia and aged controls. Clin. Neurol., 27 (1987) 1059-1063. [32] Norns, P.J., Hmdwick, J.P. and Emson, P.C., Localization of NADPH cytochrome P450 oxidoreductase in rat brain by immunohistochernistry and in situ hybridization and a comparison with the distribution of nenronal NADPH-diaphorase staning. Neuroscience, 61 (1994) 331-350. [33] Norris, P.J., Hardwick, J,P. and Emson, P.C., Studies on the localisation and expression of nitric oxide synthase using histochemical techniques. Histochem. J., 27 (1995) 745–756. [34] Norris, P.J., Waldvogel, H,, Faull, R.L.M. and Emson, P.C., Decreased nitric oxide synthase mRNA expression and immnuoreactivity in the striatum of pathologically graded cases of Huntington’s disease. Neuroscience, 72 (1996) 1037-1047. [35] OkmOW,C.W., A radical hypothesis for neurodegeneration. Trends Neurosci., 16 (1993) 439-444. [36] Rebeck, G,W,, Marzloff, K. and Hyman, B.T., The pattern of NADPH-diaphorase staining, a marker of nitric oxide synthase activity, is altered in the perforant pathway terminal zone in Alzheimer’s disease. Neurosci. Lett., 152 (1993) 165-168, [37] Rosser, MN., Emson, PC., Mountjoy, C.Q., Roth, M. and Iversen,

P.J. Norris et al./Molecular Brain Research 41 (1996) 36-49 L., Reduced amounts of immunoreactive somatostatin in the temporal cortex in senile dementia of Alzheimer’s type. Neurosci. Lett., 20 (1980) 373-377. [38] Salin, P., Mercugliano, M. and Chesselet, M.-F., Differential effects of chronictreatmentwith haloperidoland clozapineon the level of preprosomatostatinmRNA in the striatum, nucleus accumbens, and frontal cortex of the rat. Cell. Mol. Neurobiol., 10 (1990) 127-144. [39] Selden, N., Geula, C., Hersh, L. and Mesulam, M.-M., Human striatum: chemoarchitecture of the caudate nucleus, putamen and ventral striatum in health and Alzheimers disease, Neuroscience, 60 (1994) 621-636. [40] Siegel, S., Nonparametric Statistics for the Behavioral Sciences, McGraw-Hill, New York, 1956. [41] Sobreviela, T. and Mufson, E.J., Reduced nicotinamide adenine dinucleotide phosphate-diaphorase/nitric oxide synthase profiles in the human hippocampal formation and the perirhinal cortex. J. Comp. Neurol., 358 (1995) 440-464. [42] Uemnra, Y., Kowall, N.W. and Beal, M.F., Selective sparing of

49

NADPH-diaphorase-somatostatin-nenropeptide Y neurons in ischemic gerbil striatum. Ann. Neurol., 27 (1990) 620–625. [43] Unger, J.W. and Lange, W., NADPH-diaphorase-positive cell populations in the human amygdala and temporal cortex: neuroanatomy, peptidergic characteristics and aspects of aging and Alzheimer’s disease. Acta Neuropathol., 83 (1992) 636-646. [44] Vincent, S.R. and Hope, B.T., Neurons that say NO. Trends Neurosci., 15 (1992) 108–113. [45] Vincent, S.R. and Johansson, O., Stria&d neurons containing both somatostatin and avismpancreatic polypeptide (APP)-like immunoreactivities and NADPH diaphorase activity: a light and electron microscopicstudy. J. Corrrp.Neurol., 217 (1983) 264-270. [46] Vincent, S.R., Johansson, O., Hokfelt, T., Skirboll, L., Elde, R.P., Terenius, L., Kimmel, J. and Goldstein, M., NADPH diaphorase: a selective histochemical marker for striatal neurons containing both somatostatin and avian pancreatic polypeptide (APP)-like immunoreactivities. J. Comp. Neurol., 217 (1983) 252–263.