Expression of neprilysin, somatostatin and the somatostatin sst5 receptor in the hippocampal formation of brains from Alzheimer's disease patients

Authors

  • Hirotaka SEKIGUCHI,

    Corresponding author
    1. Department of Psychiatry, Nagoya University Graduate School of Medicine,
    2. Okehazama Hospital, Toyoake,
      Dr Hirotaka Sekiguchi MD, Department of Psychiatry, Nagoya University Graduate School of Medicine, Nagoya, Aichi 464-0031, Japan. Email: s-guchi@amber.plala.or.jp
    Search for more papers by this author
  • Chikako HABUCHI,

    1. Department of Psychiatry, Nagoya University Graduate School of Medicine,
    2. Department of Psychiatry, Aichi Prefectural Shiroyama Hospital, Nagoya, Aichi, and
    Search for more papers by this author
  • Shuji IRITANI,

    1. Department of Psychiatry, Nagoya University Graduate School of Medicine,
    Search for more papers by this author
  • Tetsuaki ARAI,

    1. Department of Psychogeriatrics, Tokyo Institute of Psychiatry, Tokyo, Japan
    Search for more papers by this author
  • Norio OZAKI

    1. Department of Psychiatry, Nagoya University Graduate School of Medicine,
    Search for more papers by this author

Dr Hirotaka Sekiguchi MD, Department of Psychiatry, Nagoya University Graduate School of Medicine, Nagoya, Aichi 464-0031, Japan. Email: s-guchi@amber.plala.or.jp

Abstract

Background:  In Alzheimer's disease (AD), the accumulation of amyloid β (Aβ) in the brain is thought to be the primary pathogenic agent in the AD cascade. The following have been proposed as potential therapeutic strategies in AD: (i) protease inhibitors, including β secretase and γ secretase; (ii) Aβ vaccination; and (iii) inhibitors of Aβ agglutination. However, as yet there are no studies demonstrating successful suppression of Aβ accumulation in AD brains. Neprilysin (NEP), a neutral endopeptidase, is a major Aβ-degrading enzyme that is activated by somatostatin (SST). It is thought that NEP may be a therapeutic agent against AD, but the role of SST in AD brains has not been sufficiently elucidated to date. Thus, in the present study, we compared the expression of SST, the sst5 receptor, and NEP in the hippocampal formation in brains from both AD patients and normal controls using immunohistochemical techniques.

Methods:  Twelve human brains (six control brains and six AD brains) were used in the present study. The diagnosis of AD was made according to the Braak stage. Control brains were selected from cases with no cognitive impairment clinically that were classified as being at Braak neurofibrillary tangle (NFT) Stage II. The AD brains were selected from cases classified as greater than Braak NFT Stage IV.

Results:  In the present study, SST and sst5 receptor-like immunoreactivity was significantly reduced in AD brains compared with normal brains. Although NEP-like immunoreactivity was also significantly reduced in AD brains compared with normal brains, in the CA4 region NEP was preserved in the hippocampal formation of AD brains.

Conclusion:  These results suggest that the origin of the Aβ accumulated may be correlated with the reduction of the SST neuronal network in AD brains. Activating intrinsic NEP through the SST neuronal system may contribute to a reduction in the risk of AD. Further investigations into the role SST receptors may provide new pharmacotherapeutic strategies for the treatment of AD.

INTRODUCTION

Neprilysin (NEP) is one of various targets for the treatment of Alzheimer's disease (AD), because this peptidase is the primary proteolytic neuropeptidase for amyloid beta (Aβ) protein.1,2 Neprilysin is a membrane-associated peptidase whose active site faces the lumen or extracellular side of membranes.3 Neprilysin is essentially synthesized only in neurons in the soma and is transported axonally to presynaptic terminals.4 Therefore, presynaptic terminals and nearby intracellular (lumen side) locations are thought to be the sites of Aβdegradation by NEP.5 Regulation of NEP could be expected to lead to a treatment for AD.6

In the search for various effector candidates to regulate NEP activity, it was reported that somatostatin (SST) upregulated NEP activity in cortical neurons.7 Somatostatin has attracted considerable attention because it regulates a variety of physiological functions in the central nervous system,8,9 gastrointestinal tract, pancreas, kidney, and other organs. Somatostatin acts as neurotransmitter and/or neuromodulator,10 and is associated with learning performance, memory, and cognitive functions.11 Expression of SST in the brain has been shown to decline with aging in various mammals, including primates.12 In a previous study, Davies and Katzman13 reported that one of the most consistent neurochemical deficits in AD is a reduction in cortical SST concentrations. In addition, a genetic deficiency of SST altered hippocampal NEP activity/localization and increased the quantity of a hydrophobic 42 mer form of Aβ, namely Aβ42.7 Because NEP is an intrinsic proteolytic peptidase of Aβ, fewer side-effects are expected with its use compared with vaccination therapy. Therefore, eludicating the physiological role of SST could enable the production of a new pharmacotherapy for the treatment of AD.5,14,15

As reported for SST, a reduction in SST receptors was reported in the cerebral cortex in AD.16 The biological actions of SST are mediated by a family of G-protein-coupled receptors with five known subtypes, namely sst1−5,17 which act as cell surface proteins.18

However, the expression of SST receptors in elderly humans has not been investigated sufficiently. It has been suggested that there is a significant decrease in the number of sst4- and sst5-immunoreactive neurons in the frontal cortex of AD brain compared with normal controls.18 The SST receptor subtypes are not expressed by astrocytes in the cortex of control brains, but sst4 receptors are observed in astrocytes in AD brains, although sst5 receptors are not.18 The SST receptors may be involved in glial-driven inflammatory reactions in the AD brain, especially sst4.19 Although there was a recent report investigating sst5 receptor expression in the cortex of human brains,18 there is little information available regarding sst5 receptor expression in the hippocampal formation.

Thus, the aim of the present study was to clarify the distribution of SST, sst5 receptors, and NEP in the hippocampal formation of AD brains immunohistochemically, comparing the results with those obtained in normal aging brains. The Braak classification was used to determine the stage of AD.20

METHODS

Materials

Twelve human brains (six control brains and six AD brains) were used in the present study. The diagnosis of AD was made according to Braak stage.20 Control brains were selected from cases with no cognitive impairment clinically and were all classified as being less than Braak neurofibrillary tangle (NFT) Stage II. The AD brains were selected from cases that were classified as being greater than Braak NFT Stage IV. Several neuropathologists evaluated this stage using Gallyas–methenamine silver staining. The profiles of the 12 specimens are given in Table 1. The NFT grading of the Braak classification was used to stage the specimens because the number of NFTs in AD brains is better correlated with cognitive function than the number of amyloid plaques.21,22

Table 1.  Characteristics of the 12 specimens (six control brains and six Alzheimer's disease brains) used for analysis
 Age (years)/ genderPMI (h)Braak stageClinical diagnosisCause of death
Amyloid depositsNFT
  1. Control brains were selected from cases with no cognitive impairment and classified as less than Braak neurofibrillary tangle (NFT) Stage II. Alzheimer's disease (AD) brains were selected from cases that were greater than Braak NFT Stage IV.

  2. PMI, postmortem interval; A, Stage A of Braak stage; C, Stage C of Braak stage; MRSA, methicillin-resistant Staphylococcus aureus.

Control group
 167/M4IChest cancerCancer
 271/M29ISuperior sagittal sinus thrombosisAcute circulatory failure
 372/M21IISchizophreniaPneumonia
 476/F67AIIBipolar disorderPneumonia, sepsis
Hodgkin's disease
 580/F23AIIDelusional disorderHeart failure
Dilated cardiomyopathy
 681/M34CIIPolymyositisRespiratory failure
Alzheimer's disease group
 168/F9CVIAD, gastric cancerSepsis
 275/M12CVIADPneumonia
 380/F7CVADPneumonia
 486/M3CVAD, renal failureSpontaneous death
Abdominal aortic aneurysm
 589/M2.5CVADPneumonia
 694/F13CIVADMRSA pneumonia

The hippocampal formations were selected and fixed in 4% paraformaldehyde and cut into 30 μm sections using a cryostat. The sections were rinsed and stored in phosphate-buffered saline (0.9% NaCl in 0.1 mol/L phosphate buffer, pH 7.4) for immunohistochemical staining.

Immunohistochemistry

Sections were rinsed twice in 0.1 mol/L Tris-Cl buffered saline (TBS; pH 7.4, 0.9% NaCl) containing 0.3% Triton X-100 (TX) and 2% normal goat serum (NGS) for 15 min each time at room temperature. Then, sections were incubated with anti-SST polyclonal antibody (diluted 1/500 in 2% NGS−0.3% TX−0.1 mol/L TBS solution; Chemicon, Temecula, CA, USA), anti-sst5 receptor polyclonal antibody (diluted 1/500 in the same solution; Chemicon), and anti-NEP polyclonal antibody (diluted 1/450 in the same solution; Chemicon) for 2 h at room temperature. Sections were then incubated in medium containing biotinylated anti-universal (rabbit/mouse) IgG (diluted 1/100 in NGS–TX–TBS; Vecstain, Burlingame, CA, USA) for 45 min at room temperature, before being finally incubated with an avidin biotin peroxidase complex (ABC method) for 45 min. After each incubation step, sections were rinsed twice in TBS solution for 10 min each time and reacted with 0.05%, 3,3′-diaminobenzidine dissolved in 0.05 mol/L Tris-HCl buffer (pH 7.6) for 2 or 3 min, and mounted onto gelatin-coated slides. Specimens were observed under a light microscope.

Control studies were prepared exactly the same way except that prior to immunocytochemical labeling, the diluted anti-SST, anti-sst5 receptor, and anti-NEP antibodies were preabsorbed with an adequate concentration of SST (AnaSpec, Fremont, CA, USA), sst5 receptor (Advanced Targeting Systems, San Diego, CA, USA), and NEP (Alpha Diagnostic International, San Antonio, TX, USA), respectively.

Quantitative analysis

Sections from six different control and AD brains that had been processed for SST, sst5 receptor, and NEP immunohistochemistry were subjected to quantitative analysis to determine the total number of positive neurons. First, the anatomical architecture was clarified precisely in the limbic region of the each of the specimens based on hematoxylin–eosin (HE)- and Klüver–Barrera (KB)-stained species. Based on the result of this anatomical localization, the neural count was determined in regions of interest (ROI). Neurons were considered to be SST, sst5 receptor, and NEP immunoreactive if the staining of their cell bodies was distinctly higher than background. Three randomly selected regions (each 100 × 100 μm) were evaluated in the hippocampal formation, namely the subiculum, CA1 and CA4 regions, and the number of immunopositive GABAergic cells was determined. Then, we compared the expression of SST, sst5 receptors, and NEP between control and AD brains.

To assess whether any changes observed were caused by AD pathology or to selective changes in neurons containing the peptide and/or peptidase, we also counted the number of neurons in the same area after Nissl staining and determined the reduction in immunopositive neurons in the AD group compared with the normal control group.

Results are presented as the mean ± SD. The significance of differences was assessed by the Mann–Whitney U-test with P < 0.05 considered statistically significant.

RESULTS

The SST, sst5 receptor, and NEP-like immunoreactivities were observed throughout the entire hippocampal formation (Fig. 1). The neuronal staining pattern of anti-SST, anti-sst5 receptor, and anti-NEP antibodies was the same in both control and AD brains (Fig. 2).

Figure 1.

General overview of somatostatin sst5 receptor immunoreactivity in the hippocampal formation of a brain from the control group. Higher-resolution images are shown for immunopositive neurons in the CA1, CA4, and subiculum regions. Expression of the sst5 receptor was observed from the subiculum to the dentate gyrus. Bars, 100 µm.

Figure 2.

Photomicrographs illustrating immunopositive pyramidal cells in hippocampal formations from control (a–c) and Alzheimer's disease (d–f) brains visualized using anti-neprilysin (NEP) staining. (a,d) The subiculum; (b,e) area CA1; (c,f) area CA4. Photomicrographs of the absorption test are inserted in the lower right-hand corners of each image. Bars, 20 µm.

Quantitative analysis was undertaken of immunopositive neurons in the subiculum, CA1, and CA4 areas of the hippocampal formation in control and AD brains. It was found that in the subiculum, CA1, and CA4, the number of SST-immunopositive neurons decreased in AD brains compared with normal brains (Fig. 3). Similarly, there was a decrease in the number of sst5 receptor-immunopositive neurons in the subiculum, CA1, and CA4 regions of AD brains compared with control brains (Fig. 4). Conversely, although there was a significant reduction in the number of NEP-immunopositive neurons in the subiculum and CA1 of AD brains, there was no significant difference in NEP-immunopositive neurons in the area CA4 between AD and control brains (Fig. 5).

Figure 3.

Quantitative analysis of somatostatin (SST)-immunopositive neurons in hippocampal formations from control (□) and Alzheimer's disease (inline image) brains. Data are the mean ± SEM. **P < 0.01 compared with control brain. Sub, subiculum.

Figure 4.

Quantitative analysis of somatostatin sst5 receptor-immunopositive neurons in hippocampal formations from control (□) and Alzheimer's disease (inline image) brains. Data are the mean ± SEM. *P < 0.05, **P < 0.01 compared with control brain. Sub, subiculum.

Figure 5.

Quantitative analysis of neprilysin (NEP)-immunopositive neurons in hippocampal formations from control (□) and Alzheimer's disease (inline image) brains. Note, there was no significant change in the the number of immunopositive neurons in the area CA4 between the two groups. Data are the mean ± SEM. **P < 0.01 compared with control brain.

Neuronal reduction

Based on observations made in Nissl-stained specimens, there was a reduction in the average number of neurons in brains from the AD group compared with those from the control group (Fig. 6). Although the average number of neurons decreased in each of the subiculum, CA1, and CA4 areas in AD brains compared with control, the reduction in the number of immunopositive neurons in each of these areas was much greater, except in the case of NEP-immunoreactive neurons in the CA4 area (see above). This indicates that the reduction in the number of immunopositive neurons is caused not only by the AD pathology, but also by selective peptide and/or peptidase action.

Figure 6.

Changes in the number of neurons (inline image), as determined by Nissl staining, and in the number of somatostatin (SST; ◆)-, sst5 receptor (●)-, and neprilysin (NEP; ▴)-immunoreactive neurons observed in the Alzheimer's disease (AD) and control groups in the (a) subiculum, (b) CA1, and (c) CA4. Reductions in the number of neurons in the AD group are given as a percentage of the number of neurons in the control group. General neuronal loss was detected in the AD group compared with the control group, as indicated by Nissl staining. These reductions were 87.6, 75.1, and 96.1% of control levels in the subiculum (NS), CA1 (< 0.01), and CA4 (NS), respectively. A more significant reduction was observed for each type of immunoreactive neuron (except for NEP-immunoreactive neurons in the CA4 region: reduced to 91.7% of control levels; NS): SST was reduced by 50.8, 48.2, and 62.1% in the subiculum (< 0.01), CA1 (< 0.001), and CA4 (< 0.01), respectively; sst5 receptors were was reduced by 67.2, 60.8, and 71.4% in the subiculum (< 0.05), CA1 (< 0.001), and CA4 (< 0.01), respectively; and NEP-like immunoreactive neurons were reduced by 57.3 and 63.9% in the subiculum (< 0.01) and CA1 (< 0.01), respectively.

DISCUSSION

A neuropathological approach has been applied to AD research for some time, yielding important information.23 The aim of the present study was to investigate the distribution of SST, sst5 receptors, and NEP in the hippocampal formation of AD brains using immunohistochemical techniques. Kumar showed fluctuations in the expression of every SST receptor subtype in the cortex.18 Further to his results, in the present study we have clarified the tissue expression of the sst5 receptor, which was associated with the expression of SST and NEP, and fluctuations in NEP expression in the hippocampal formation of AD brains. This is the first study to compare the expression of these three compounds in AD and normal brains. We showed that both SST and sst5 receptor-like immunoreactivities were significantly reduced compared with levels in control brain tissue. These results are consistent with those of previous studies.13,16,18 In particular, Arai et al. reported a significant reduction in SST-containing neurons in the hippocampus compared with other areas of the cortex.24 This is consistent with our observations regarding the expression of SST. In the present study, a reduction in both SST and sst5 receptor expression was observed in AD brain compared with normal brain. In a healthy organ, generally up- or downregulation of agonist levels in vitro may result in compensatory down- or upregulation of its receptors. For example, it has been reported that serotonin receptors are increased in the postmortem brain of depression, whereas serotonin levels are decreased.25 However, in aged organs or in a degenerated organ, including AD, this homeostatic mechanism is weakened and a reduction in the agonist directly may result in receptor downregulation. In the present study, we found that NEP-like immunoreactivity was decreased in the hippocampal formation of AD brains compared with control brain tissue, except in the area CA4. The NEP-like immunoreactivity in area CA4 was not significantly decreased in AD compared with normal brain. Although a previous study reported a significant reduction of NEP mRNA and protein levels in the hippocampus,26 the present study demonstrated precise expression of NEP in the hippocampal formation for the first time. Although SST and sst5 receptor-like immunoreactivities were significantly decreased in the area CA4 in AD brains, NEP-like immunoreactivity was preserved in the same area in AD. The fact that the expression of NEP did not decrease in the area CA4 of AD hippocampal formations could be related to the staging of AD according to Braak's classification, in which NFT changes in the area CA4 appear at a late stage.20 That is, because NEP in neurons in the area CA4 is preserved until a late stage, it is likely that the appearance of Aβ in AD pathology occurs later in the area CA4 than in other areas. It has been reported that NEP is selectively decreased at the terminal zones and on the axons of the lateral perforant path and mossy fibers in aging mice.27 This may be due to the fact that the level of expression of NEP is dependent on the type of neuronal cell. The area CA4 (dentate gyrus) consists mainly of granule cells, but other areas of the hippocampal formation consist mainly of pyramidal cells. It may be that the neuronal characteristics of each functional anatomical area affect the changes in NEP levels, and so may exhibit differences in Aβ deposits and accumulation. Therefore, the histological features of the CA4 area may impact on the expression of NEP and the process of AD pathology.

It is known that there are five SST receptor subtypes, but the roles of each individual receptor remains unclear. It was reported that the sst5 receptor is expressed in neuronal cells,18 and the present study is the first to demonstrate its distribution in the hippocampal formations in AD in comparison with that in normal brain. The present results indicate that the sst5 receptor may regulate the activity of NEP. Therefore, the sst5 receptor may be a specific target for the upregulation of NEP activity.

There are two possible hypotheses regarding the impairment of cognitive function in AD. One is based on a decrease in SST neurotransmitter function. Somatostatin is involved in various mental functions, such as memory, learning etc.,28,29 and it is thought that there may be an association between some psychiatric disorders and significant reductions/increases in SST levels in the cerebrospinal fluid, as shown for depression/mania patients.30,31 The second hypothesis proposes that reductions in SST may result in impaired cognitive function through Aβ pathology decreasing the activity of NEP. A recent study reported that SST genetic variants modified the risk for AD.32 Although the mechanism underlying this effect is not known precisely, it may be considered that the function of SST itself influences the synthesis of NEP, and SST may inhibit and/or promote the accumulation of Aβ. Furthermore, it has been reported that SST-containing neurons within the hilus of the dentate gyrus exhibit antiepileptic properties,33 so changes in SST-containing neurons may be associated with the seizures observed in the terminal stages of AD.

Several therapeutic strategies have been propsed for AD, including the use of protease inhibitors, such as β- and γ-secretase, Aβ vaccinations, and inhibitors of Aβ agglutination. However, there have been many clinical difficulties associated with these therapies. For example, there are several problems associated with vaccine therapy, such as the serious side-effect of meningoencephalitis34 and complicated inoculations. However, an effective therapeutic strategy could be developed to promote the degeneration of Aβ by activating intrinsic NEP through the SST neuronal system. That is, activation of intrinsic NEP may contribute to a reduction in the risk for AD without the side-effects associated with current therapies.

In addition, if recent molecular biological techniques could be used to genetically control NEP activity (i.e. to increase it) and a sst5 receptor-selective agonist could be administered to AD patients, these strategies could represent new treatment measures for AD. From a diagnostic point of view, if it is certain that reductions in NEP cause the processes of AD pathology, measuring NEP levels in vivo may be a way to predict and detect early AD.

In addition, we need to clarify the mechanism underlying the overlapping intracellular expression of SST and NEP. To investigate NEP regulation by the SST system, further studies are required, including double staining. These studies may lead to further progress in the treatment, prevention, and early detection of AD.

Ancillary