, Volume 172, Issue 4, pp 365–374 | Cite as

Modulation of nerve growth factor and choline acetyltransferase expression in rat hippocampus after chronic exposure to haloperidol, risperidone, and olanzapine

  • Vinay Parikh
  • Alvin V. Terry
  • Mohammad M. Khan
  • Sahebarao P. Mahadik
Original Investigation



Recently, we reported that compared to haloperidol, chronic exposure to either the risperidone (RISP) or olanzapine (OLZ) resulted in superior effects on spatial learning performance as well as the cholinergic neurons, although the mechanism for these effects was not addressed.


The objective of this study was to investigate one plausible mechanism whereby RISP and OLZ exert superior effects on cholinergic neurons, i.e. positive effects on nerve growth factor (NGF), which is known to regulate the brain cholinergic activity as well as cognitive function. Therefore, the effects of chronic exposure to HAL, RISP, or OLZ on the expression of NGF and choline acetyltransferase (ChAT) in the hippocampus (i.e. a brain area well known to be involved in cognitive function and known to receive major cholinergic projections from the medial septum) were compared.


Rats were treated with HAL (1 or 2 mg/kg per day), RISP (1.25 or 2.5 mg/kg per day), or OLZ (5 or 10 mg/kg per day) for 45 days in drinking water. NGF and ChAT were measured by immunohistochemistry and NGF protein was also measured by an enzyme-linked ImmunoSorbent assay.


Compared to controls, HAL exposure resulted in a marked reduction in NGF immunoreactivity in the hippocampal dentate gyrus (DG), CA1 and CA3 areas. In contrast, RISP did not alter, while OLZ significantly increased levels of NGF. These changes in NGF levels corresponded well with changes in ChAT immunoreactivity in the hippocampus and the medial septum.


These preclinical data, combined with previously published behavioral results, support the premise that OLZ and RISP, in contrast to HAL, preserve cholinergic pathways and cognitive function via superior effects on NGF expression and are thus therapeutically superior for extended use.


Antipsychotics Cognition Nerve growth factor Choline acetyltransferase Haloperidol Risperidone Olanzapine Schizophrenia Rat 


Previously, we found that chronic exposure to haloperidol (HAL) impaired spatial learning performance in rats, which was associated with a reduction in central cholinergic markers (Terry et al. 2002, 2003). Furthermore, we found that 45 days of pretreatment with clozapine (CLOZ) or post-treatment with olanzapine (OLZ), respectively, prevented or restored the loss of cholinergic markers induced by 45 days of HAL exposure, indicating that these drugs may have neuroprotective properties and the ability to trigger neuroplasticity (Mahadik et al. 2001; Parikh et al. 2003a). This is important, since cholinergic activity in the brain is essential to cognitive procuresses in animals and humans (Harder et al. 1998; Baxter et al. 1999). In contrast to typical agents like HAL, atypical antipsychotics are associated with reduced (or a lack of) extrapyramidal symptoms (EPS) in schizophrenic patients (Buckley 2001). Furthermore, atypical antipsychotics are more effective in reducing negative symptoms and improving cognitive performance (Sharma and Mockler 1998; Cuesta et al. 2001). These findings are significant, particularly since a wide range of cognitive deficits (i.e. attention, learning, memory, and executive functions) is generally found in these patients, even in the early stages of their illness (Tollefson 1996). Since impaired septohippocampal cholinergic function has been implicated in memory deficits (Gallagher and Nicolle 1993; Nilson and Gage 1993), understanding the possible molecular mechanisms underlying the effects of antipsychotics on septohippocampal cholinergic neuronal activity may provide clues for the further development of optimal treatment modalities.

Nerve growth factor (NGF), a member of the neurotrophin family (Levi-Montalcini 1987), plays an important role in the growth and survival of central neurons, as well as that of sensory and sympathetic cholinergic neurons, throughout life (Rylett and Williams 1994). In the CNS, NGF is mainly produced in the hippocampus and cortex (Korsching et al. 1985) and is retrogradely transported to NGF-responsive cholinergic neuronal cell bodies located primarily in the septum and basal forebrain (Thoenen et al. 1987). NGF administration also has been shown to prevent cholinergic neuronal cell death associated with injury (Kromer 1987), and to exert neurotrophic actions on cholinergic neurons of the basal forebrain (Hefti 1994). These and other studies have led to the recognition of NGF as a very potent trophic factor, particularly for the protection and neuroplasticity of cholinergic neurons projecting to the hippocampus and cerebral cortex, regions of the brain most often are associated with cognitive performance in both animals and humans (Garofalo et al. 1992; Lo 1995; Thoenen 1995).

Interestingly, a few recent studies demonstrated that treatment with HAL in rats reduced NGF levels in the brain (Alleva et al. 1996; Angelucci et al. 2000) leading to obvious questions concerning the effects of this neuroleptic on cholinergic activity in the brain as well as cognitive function. In order to investigate the possible molecular mechanisms underlying the differential effects of the antipsychotics (highlighted above) on the plasticity of cholinergic neurons, we compared the effect of 45-day chronic exposure of HAL to the effect of 45-day exposure to RISP and OLZ on the expression of NGF and ChAT in the rat hippocampus. NGF levels in the hippocampus were measured using a quantitative immunohistochemical method and an enzyme-linked ImmunoSorbent assay (ELISA) procedure. Since cholinergic projections to the hippocampus originate in the medial septum (McKinney et al. 1983), ChAT immunoreactivity was determined by quantitative immunohistochemistry of medial septum and hippocampal projections.

Materials and methods


Male albino Wistar rats (225–250 g) were obtained from Harlan Sprague-Dawley, Inc (Indianapolis, Ind., USA) and housed individually in a temperature controlled room (25°C) with a 12-h/12-h light-dark cycle. Upon arrival, each animal was provided with tap water and food (Purina Rat Chow) ad libitum for 1 week. Thereafter, tap water was replaced with the drug solutions described below. All procedures employed during this study were reviewed and approved by the Medical College of Georgia Committee on Animal Use for Research (CAURE) and the Veterans Affairs Medical Center Subcommittee on Animal Use. These procedures were consistent with AAALAC guidelines as per Public Health Service (PHS) Policy on Humane Care and Use of Laboratory Animals (2000, reprint).

Drug treatments

All the drugs were prepared daily and administered in solutions that replaced each animal’s drinking water. HAL (Sigma Chemicals, St Louis, Mo., USA), RISP (Janssen Pharmaceutica, Trenton, N.J., USA), and OLZ (Eli Lilly, Indianapolis, Ind., USA) were each dissolved in a 0.1 M solution of acetic acid and then diluted (1:100) with tap water. The amount of drug intake was measured every day and adjustments were made for each animal depending on the amount of fluid it had consumed and its weight. Rats were exposed to HAL (HAL-1 and HAL-2; 1 or 2 mg/kg per day, respectively), RISP (RISP-1.25 and RISP-2.5; 1.25 or 2.5 mg/kg per day, respectively), or OLZ (OLZ-5 and OLZ-10; 5 or 10 mg/kg per day, respectively) for 45 days. Tap water containing 0.1 M acetic acid (100×1) was used for the control group to assure that an unanticipated effect caused by the vehicle was not introduced. After 45 days of treatment, blood was drawn from each test animal and the samples were sent to an independent laboratory, Analytical Psychopharmacology Laboratories of Nathan Kline Institute (Orangeburg, N.Y., USA).

NGF immunohistochemistry

After 45 days of treatment, rats (n=3–5/group) were deeply anesthetized with ketamine/xylazine and perfused with cold 0.01 M phosphate buffered saline (PBS) through the left cardiac ventricle to remove circulating blood elements. Brains were quickly removed and cryoprotected in an embedding medium. Coronal sections (20 μm in thickness) were cut at interaural 4.84 mm, bregma –4.16 mm to obtain sections of the hippocampus (dentate gyrus, DG; CA1 and CA3) (Paxinos and Watson 1998) using a cryostat microtome (Leica CM 3050S, Leica Microsystems, Inc., Chantilly, Va., USA) at –20±2°C. The procedure used for NGF immunohistochemistry was modified from that used by Lee et al. (1996). Fresh frozen sections were fixed in ice-cold acetone for 30 min and then air dried. Sections were then rinsed in 0.01 M PBS containing Tween 20 (PBST). After blocking with 10% normal goat serum for 1 h, sections were washed and incubated overnight at 4°C with rabbit anti-mouse polyclonal NGF antibody (1:100) (Chemicon International, Inc., Temecula, Calif., USA). Endogenous peroxidase was blocked for 30 min with 0.1% H2O2 and 100% methanol, then sections were washed and incubated for 2 h with biotinylated anti-rabbit IgG antibody obtained from goats (1:50), followed by incubation for 1 h with avidin-biotin-horseradish peroxidase complex. Staining was developed with 3–3’-diaminobenzidine tetrahydrochloride (DAB) in the presence of 0.02% H2O2 and nickel chloride (Vectastain kit; Vector Laboratories, Burlingame, Calif., USA).

ChAT immunohistochemistry

The method used for ChAT immunohistochemistry was modified from that used by Angelucci et al. (2000). After chronic treatment with antipsychotics, rats (n=4–5/group) were anesthetized and perfused with 100 ml of saline followed by 300 ml ice-cold 4% paraformaldehyde in 0.1 M PBS. After perfusion, all brains were post-fixed in paraformaldehyde for 2 h (with shaking) at 4°C, followed by storage in a 30% sucrose solution in 0.1 M PBS for 48 h. The tissues were embedded with OCT in liquid nitrogen and kept at –80°C till further use. Coronal sections 40 μm thick were cut at specific anatomical landmarks: (a) interaural 9.48 mm, bregma 0.48 mm to obtain sections with subfields of medial septum and (b) interaural 4.84 mm, Bregma –4.16 mm to obtain sections from hippocampus (Paxinos and Watson 1998) using a cryostat microtome. Cryoprotected fixed sections were washed 3 times in PBST, blocked with 10% normal horse serum for 1 h, and incubated with 10 μg/ml mouse monoclonal anti-ChAT antibody (Chemicon International) overnight at 4°C. The sections were washed 3 times with PBST and then incubated with 1:20 diluted rat adsorbed biotinylated horse anti-mouse IgG (Vector Laboratories), containing 1% horse serum for 2 h. After washing, the sections were incubated with avidin-HRP for 1 h and staining was developed with DAB.

Quantitative image analysis of immunohistograms

Immunohistograms from each treatment group were analyzed with a Zeiss Axioplan-2 microscope equipped with CCD camera, personal computer, and Zeiss KS-300 image analyses software by an experimenter blinded to the study code. Two sequential sections per animal from each experimental group were evaluated and the average of two values per region examined was used for analysis.

The analysis to determine the amount of NFG in the hippocampus was performed by examining acquired images of dimensions 582×455 μm2 in dentate gyrus (DG) granule cell layer and 582×228 μm2 each in CA1 and CA3 cell layer pyramidal neurons. For quantitation, both the morphologic and densitometric assessments were taken into account. Three parameters reflecting the amount of NGF in the hippocampus were assessed: (1) the number of stained cells in each field, (2) the percentage of the stained cell area within the image, and (3) the optical density (OD) of the immunostained cells. Quantitation using these parameters has been described previously (Mausset et al. 2001). All the parameters described above were measured for the cells detected as positive. The cells were considered positive if their OD values were higher than a defined threshold (OD value above which only cell bodies and not processes were detectable). The number of stained cells analyzed per field was obtained by counting individual positive cells. Three rectangles per section, with dimensions of 582×455 μm2 for DG and 582×228 μm2 for CA1 and CA3 regions, were delineated. NGF immunoreactive neurons were identified in these regions and counted (experimenter blinded to the treatment groups in each region), and a mean value of two counts was used for further analysis. The percentage of the area containing stained cells was calculated as the ratio between the stained cell area and the total area of the analyzed field. Staining intensity expressed as mean OD (MOD) was obtained by averaging OD values of all stained profiles in the analyzed field and subtracting the background OD of each section. The OD range 0–2 was divided into 256 digitized values (0–255).

ChAT immunoreactive nerve fibers in the hippocampus were analyzed by the method of Nakamura et al. (2002). Three rectangles per section for DG, CA1, and CA3 subfields with similar dimensions as described above were selected for analysis. All the digitized images of ChAT immunoreactive nerve fibers were converted to gray scale and the brightness, contrast, and masking was adjusted to enhance the visibility of fibers (Photoshop 5.0; Adobe Systems, San Jose, Calif., USA). Quantitative data for ChAT immunoreactive fibers are expressed as a measure of fiber pixel density that was taken as the average pixel intensity for each image.

In the medial septum, each rectangle per section with dimensions 1164×909 μm2 was delineated for ChAT analysis. Quantitation was done in a similar fashion as that described for NGF. ChAT immunoreactive neurons were counted twice in each rectangle on the same day and the average of two counts per section was considered. The number of ChAT positive cells in the medial septum represented the mean of ChAT immunoreactive neurons in two rectangles from two sections per experimental animal. The percentage of the ChAT stained cell area and MOD was measured in each rectangle of the captured image as described above.

Double immunostaining

To assess whether NGF-containing neurons received cholinergic projections in the hippocampus and whether septal cholinergic neurons contain NGF, a few sections were double stained. Frozen sections were fixed in acetone, washed in PBST, blocked with 10% normal donkey serum for 1 h and incubated overnight at 4°C with anti-mouse polyclonal NGF antibody (1:100) and monoclonal ChAT antibody. After three to four washes, secondary antibodies (Jackson ImmunoResearch Laboratories Inc., West Grove, Pa., USA) in the form of donkey anti-rabbit cy3 (1:100) and donkey anti-mouse cy2 (1:200) were applied and visualized. Controls included the reversing the order of primary antibodies, or omitting the primary or secondary antibody.

ELISA procedure for NGF protein

The animals (n=6–8 per group) were killed after 45 days of treatment, brains were removed, and the hippocampus was dissected, snap frozen in liquid nitrogen, and stored at –70°C. The frozen tissues were homogenized in an ice-cold buffer consisting of 100 mM Tris/HCl pH 7.0, containing 2% bovine serum albumin (BSA), 1 M NaCl, 4 mM EDTA, 2% Triton X 100, 0.1% sodium azide, 5 μg/ml aprotinin, 0.5 μg/ml antipain, 157 μg/ml benzamidine, 0.1 μg/ml pepstatin A and 17 μg/ml phenylmethyl-sulphonyl fluoride using polytron homogenizer (Brinkmann Instruments, N.Y., USA). The supernatants were collected by centrifugation at 15,000 g for 30 min and used for ELISA. NGF was determined using a two-site enzyme immunoassay kit (Chemicon International). The sensitivity of ELISA was 10–15 pg/ml and no significant crossreactivity was observed with other neurotrophins like BDNF, NT3, or NT4/5. Data are represented as pg/g wet weight. All assays were performed twice.

Statistical analyses

The data are expressed as mean±SEM. One-way analysis of variance (ANOVA) was used to compare the difference between control and treatment groups. When significant differences were observed, a pairwise multiple comparison (Student-Newman-Keuls) post hoc test was conducted. All statistical tests were performed with the statistical program SigmaStat version 2.03.


Plasma drug levels

Plasma levels of the antipsychotics after 45 days of treatment were as follows: HAL-2=2.1±0.25 ng/ml; RISP-2.5=20.2±3.4 ng/ml; OLZ-10=10.8±3.0 ng/ml. RISP and OLZ treated animals showed plasma levels that were in the range of the plasma levels (20–80 ng/ml and 9–15 ng/ml, respectively) at therapeutic doses for the treatment of schizophrenia (Andersson et al. 2002). Plasma levels of HAL in rats were lower compared to the therapeutic concentration of 10–20 ng/ml, which may indicate its high metabolic turnover in rats.

Chronic exposure of antipsychotics and hippocampal ngf immunoreactivity

The specificity of the immunohistochemical procedure for NGF is illustrated in Fig. 1A,B. Representative photomicrographs of NGF-stained sections from each drug group are shown for dentate gyrus (DG) granular cell layer (Fig. 1C–F), CA3, and CA1 pyramidal neurons (Fig. 2). The data from quantitative image analysis of NGF immunoreactivity is presented in Fig. 3A–C. The first set of analyses is the number of NGF stained cells (the less sensitive of all other parameters), which is affected only if there is a large change in NGF content. Forty-five days of HAL-2 exposure markedly reduced the number of NGF positive cells in CA1 (48±4 versus 66±5, P<0.05 versus control) and CA3 (32±4 versus 52±5, P<0.05 versus control) pyramidal cell layer (Fig. 3A). No such reduction was observed with HAL-1. Exposure to RISP-1.25, RISP-2.5, and OLZ-5 did not alter the cell count of NGF positive cells in CA1 and CA3 areas; however, OLZ-10 significantly increased the cell count in the CA1 region (P<0.05 versus control). None of the antipsychotics significantly affected the cell count in DG.
Fig. 1

Representative photomicrographs depicting specificity of NGF immunoreactivity in rat hippocampus and effect of chronic antipsychotic treatment in dentate gyrus (DG) granular cell layer. A and B show background and high intensity NGF specific immunoreactivity in hippocampal neurons immunostained with rabbit pre-immune and anti-NGF polyclonal antibody, respectively. Arrowheads indicate DG, CA1 and CA3 sub-regions of the hippocampus. Rectangles indicate dimensions (582×455 μm2 for DG) and (582×228 μm2 for CA1 and CA3) of captured images for quantitation. CF show representative photomicrographs that illustrate the NGF stained granular neurons in the dentate gyrus after 45 days of exposure to antipsychotics. C control (vehicle); D HAL (2 mg/kg per day haloperidol); E RISP (2.5 mg/kg per day risperidone); F OLZ (10 mg/kg per day olanzapine). NGF is exclusively localized in neuronal cells. Haloperidol, but not risperidone and olanzapine, show a wide range of reduction in NGF immunoreactivity in a large number of neurons. Bar represents 80 μm

Fig. 2

Representative photomicrographs illustrating the effect of antipsychotics on NGF immunoreactivity in hippocampal CA1 and CA3 pyramidal neurons. Tissue sections were from groups CON, HAL, RISP, and OLZ as described in Fig. 1. Marked reduction in NGF immunoreactivity is seen with HAL and increase with OLZ. Bar represents 100 μm

Fig. 3

Effects of chronic exposure with antipsychotics on the number, area and mean optical density (MOD) of NGF immunoreactive cells in hippocampus DG, CA1 and CA3 areas. A Number of NGF positive cells, B % area of NGF stained cells, C MOD of NGF stained cells. Bar graphs show means±SEM (n=3–5 animals/group). Tissue sections were from animal groups CON, HAL, RISP, and OLZ as described in Fig. 1. *,**,***P<0.05, 0.01, 0.001 HAL versus CON; aP<0.05 OLZ versus CON

The second set of analyses, the percentage of the area containing stained cells, shows the distribution of NGF (Fig. 3B). HAL-2 significantly reduced the area of NGF stained cells in DG (9.1±0.8 versus 13.0±0.9, P<0.05 versus control), CA1 (5.0±0.6 versus 7.1±0.4, P<0.05 versus control), and CA3 (3.4±0.6 versus 7.0±0.6, P<0.01 versus control) areas. A significant reduction in the CA3 subfield was also observed with HAL-1 (P<0.05 versus control). No significant changes in the distribution of NGF-stained cells were observed with RISP and OLZ in DG, but OLZ-10 significantly increased the distribution in hippocampal CA1 (8.8±0.4) and CA3 (9.4±0.5) pyramidal neurons (P<0.05 for both versus control).

The last set of analyses, mean optical density (MOD) values of stained cells, shows the NGF concentrations (Fig. 3C). A marked dose-dependent reduction in MOD of stained cells was observed with chronic HAL exposure in DG (HAL-1, 69±4, P<0.01 and HAL-2, 57±3, P<0.001 versus control, 84±3); CA1 (HAL-1, 51±4, P<0.05 and HAL-2, 48±2, P<0.01 versus control, 62±3), and CA3 (HAL-1, 45±3, P<0.01 and HAL-2, 36±3, P<0.001 versus control, 59±2) areas. Neither RISP-1.25 nor RISP-2.5 affected the MOD values, whereas OLZ-5 and OLZ-10 significantly increased the MOD values in DG (OLZ-5, 96±3 and OLZ-10, 99±5; P<0.05 for both versus control), CA1 (OLZ-5, 71±3 and OLZ-10, 72±3; P<0.05 for both versus control), and CA3 (OLZ-5, 70±3 and OLZ-10, 72±5; P<0.05 for both versus control).

Figure 4 shows the NGF protein levels in hippocampus as measured by ELISA. These data support the immunohistochemical differences, i.e. a significant reduction in NGF protein with HAL compared to controls (HAL-1, 3048±197, P<0.05 and HAL-2, 2550±244, P<0.01 versus control, 4015±271), no change with both RISP-1.25 and RISP-2.5, and even a significant increase with OLZ (OLZ-5, 4809±280 and OLZ-10, 5167±308; P<0.05 for both versus control).
Fig. 4

NGF protein levels in hippocampus measured by ELISA after 45 days of antipsychotic treatment in rats. Bar graphs show means±SEM (n=6–8 animals/group). Hippocampal tissue extracts were from groups CON, HAL-1 and HAL-2 (1 and 2 mg/kg per day); RISP-1.25 and RISP-2.5, (1.25 and 2.5 mg/kg per day); OLZ-5 and OLZ-10 (5 and 10 mg/kg per day). *,**P<0.05, 0.01, HAL versus CON; aP<0.05, OLZ versus CON. Values represent pg/g wet weight

Chronic exposure of antipsychotics and ChAT immunoreactivity

ChAT immunoreactivity in the medial septum and hippocampus is shown in Fig. 5 and quantitative data are included in Fig. 6. There was a marked reduction in the distribution of ChAT stained cells with HAL (HAL-1, 2.9±0.27, P<0.05 and HAL-2, 2.42±0.21, P<0.01 versus control, 3.84±0.4) in the medial septum (Fig. 6B). Similarly, HAL exposure markedly reduced the MOD of ChAT (HAL-1, 2.9±0.27, P<0.05 and HAL-2, 2.42±0.21, P<0.01 versus control, 3.84±0.4) in the septum (Fig. 6C). No significant changes were observed with either RISP-1.25 or RISP-2.5 in ChAT immunoreactive cells or MOD. However, exposure to OLZ markedly increased the area (OLZ-5, 4.7±0.5, P<0.1 and OLZ-10, 4.9±0.36, P<0.05 versus control) and MOD (OLZ-5, 4.7±0.4 and OLZ-10, 4.9±0.36, P<0.05 for both versus control) of ChAT stained cells (Fig. 6B,C). Although there were no statistically significant alterations observed in cell counts between groups, HAL-2 significantly lowered the number of ChAT positive cells (P<0.05) compared to OLZ-10 (Fig. 6A). Varicose ChAT immunoreactive fibers were distributed in the granule cell layer of DG and in the pyramidal neurons of CA1 and CA3 subfields. The density of ChAT was highest in the hippocampal CA3 pyramidal cell layer (Fig. 5). Compared to the controls, the HAL significantly decreased the fiber pixel density in DG (HAL-1, 0.06±0.005, P<0.05 and HAL-2, 0.052±0.005, P<0.01 versus control, 0.079±0.006), CA1 (HAL-1, 0.062±0.003, and HAL-2, 0.057±0.005, P<0.05 for both versus control, 0.078±0.006), and CA3 (HAL-1, 0.08±0.004, P<0.05 and HAL-2, 0.072±0.004, P<0.01 versus control, 0.1±0.006) (Fig. 6D). Such a decrease was not found in RISP and OLZ. Rather, the OLZ had significantly increased ChAT immunoreactive fiber density in DG (OLZ-5, 0.092±0.004 and OLZ-10, 0.093±0.003; P<0.05 for both versus control), CA1 (OLZ-5, 0.095±0.005 and OLZ-10, 0.096±0.006; P<0.05 for both versus control), and CA3 (OLZ-5, 0.117±0.006, P<0.05 and OLZ-10, 0.134±0.007; P<0.01 versus control).
Fig. 5

Representative photomicrographs illustrating the effect of antipsychotics on ChAT immunoreactivity in neurons of medial septum and projections to the hippocampal CA3 region. Tissue sections were from animal groups CON, HAL, RISP, and OLZ as described in Fig. 1. Marked reduction in ChAT immunoreactivity is observed in HAL, and increase in OLZ. Bar represents 100 μm (medial septum) and 120 μm (hippocampal CA3 region)

Fig. 6

Comparative effect of chronic exposure with antipsychotics on the number of stained cells (A), staining area (B) and mean optical density (MOD) (C) of ChAT stained neurons in medial septum and fiber pixel intensity of ChAT (D) in hippocampus. Bar graphs show means±SEM (n=4–5 animals/group). Tissue sections were from groups CON, HAL-1, HAL-2, RISP-1.25, RISP-2.5, OLZ-5 and OLZ-10 as described in Fig. 4. *,**P<0.05, 0.01 HAL versus CON; aP<0.05 OLZ versus CON; bP<0.01 OLZ versus CON; cP<0.05 HAL versus OLZ

Double immunostaining of NGF and ChAT

Double immunostaining of NGF and ChAT in the hippocampal CA3 area is shown in Fig. 7A–C. ChAT immunoreactivity (green, Fig. 7B) was mainly observed in the synaptic terminals and fibers distributed within the CA3 pyramidal cell bodies containing NGF (red, Fig. 7A). No colocalization of NGF and ChAT was observed within the cell body of CA3 pyramidal neurons (Fig.7C). Some of the projections show yellow staining indicative of the presence of both ChAT (green) and NGF (red) (Fig.7C), however, the low resolution prevents the conclusion of colocalization. Figure 7D–F show colocalization of NGF and ChAT in cholinergic cell bodies of the medial septum, indicating the presence of NGF in cholinergic cells.
Fig. 7

Photomicrographs illustrating the double immunostaining of NGF and ChAT in hippocampus and medial septum. AC show localization of NGF (red) in the neuronal cell bodies, ChAT (green) on the synaptic terminals and projections surrounding the hippocampal CA3 pyramidal cells and possible double staining (yellow, in hippocampal cholinergic projections), respectively. DF show colocalization (yellow, marked with arrows) of NGF (red) and ChAT (green) in the medial septal neurons. Bars represent 100 μm (AC) and 30 μm (DF)


The experiments described in this study resulted in two key findings. First, 45 days of exposure to the two doses of HAL, but not RISP or OLZ, markedly reduced NGF immunoreactivity in several regions of the rat hippocampus. Second, changes in the cholinergic marker, ChAT in the cell bodies of the medial septum, and cholinergic projections to the hippocampus, all paralleled the changes in NGF levels. Furthermore, the deleterious effects of HAL on NGF and ChAT were more pronounced in HAL-2 compared to HAL-1, suggesting a dose-dependent effect. Our findings with HAL are supported by earlier reports suggesting that chronic treatment with HAL is associated with a reduction in NGF levels and ChAT immunoreactivity in the rat brain (Mahadik et al. 1988; Alleva et al. 1996; Angelucci et al. 2000; Terry et al. 2003). These reductions in ChAT and NGF following exposure to HAL were associated with altered morphology, i.e. cellular shrinkage, reduction in cytoplasmic volume dendritic fibers as previously reported (Mahadik et al. 1988; Jeste et al. 1992). A slight reduction in the cell count of NGF and ChAT positive neurons observed with HAL in this study may not reflect cell death. Significant reductions in the levels of NGF/ChAT result in significant reduction in staining intensity, which yield lower counts of stained cells by the image analysis software. Moreover, RISP preserved and OLZ increased NGF immunoreactivity in the hippocampus and ChAT immunoreactivity in hippocampal cholinergic projections and cholinergic cell bodies of the medial septum at both doses used in the experiments. NGF immunoreactivity was observed in the neuronal cell bodies including in the nuclear region. Our study is in agreement with those of Lee et al. (1998). Furthermore, the NGF immunoreactivity in the nuclear region appeared in discrete dots in nuclear and perinuclear regions as reported (Marchisio et al. 1980).

The differences in NGF immunoreactivity in the hippocampal DG, CA1, and CA3 cell layers with antipsychotics indicate that these drugs differentially alter the expression of NGF protein. Similar alterations in ChAT immunoreactivity were observed in cholinergic fibers in hippocampal areas, DG, CA1, and CA3. In addition, these changes were also reflected in cholinergic cell bodies of the medial septum, which is notable in view of the fact that cholinergic medial septum neurons have abundant afferents to the hippocampal formation (Woolf et al. 1984; Amaral and Kurz 1985). Changes in ChAT immunoreactivity in the medial septum and the hippocampus, and co-localization of NGF and ChAT, support the suggestion that altered NGF in the hippocampus may be related to altered ChAT immunoreactivity (i.e. reduced with HAL and increased with OLZ), since NGF is retrogradely transported from the hippocampus and the cortex to basal forebrain cholinergic neurons (Thoenen et al. 1987).

Differential effects of antipsychotics on NGF expression may provide one of the biochemical bases of both the biological as well as differential behavioral effects reported in animals (Mahadik et al. 1988; Kinon and Lieberman 1996; Terry et al. 2002, 2003) as well as in patients with psychotic disorders (Green and Braff 2001). The mechanisms associated with differential effects of these antipsychotics on the expression of brain NGF may be complex. HAL clearly differs from RISP and OLZ in its neurotransmitter receptor reactivity profile (Bymaster et al. 1996; Kapur et al. 1999; Seeman 2002). HAL is known to produce potent D2 antagonistic activity, whereas atypical antipsychotics like RISP and OLZ block both the D2 and 5HT2 receptors (Meltzer 1995). These actions may start downstream events that may involve a number of additional neurotransmitter systems (e.g. GABA, glutamate, acetylcholine), which may have disparate effects on NGF. In the context of central cholinergic activity, it was recently reported that CLOZ and OLZ, but not the typical antipsychotics, markedly increased acetylcholine release in the hippocampus and cortex (Ichikawa et al. 2002; Shirazi-Southall et al. 2002). Cholinergic activity does regulate NGF levels (Alberch et al. 1991; Rossner et al. 1997).

The findings of this study may have significant implications for the pathophysiology, treatment and outcome of schizophrenia. There is general agreement that cognitive improvement is critical to improve the quality of life of schizophrenic patients (Tollefson 1996). Hippocampal involvement occurs in chronic as well as in early psychotic patients (Weinberger 1999). Atypical antipsychotics are found to improve cognitive performance in both groups of patients (Sharma and Mockler 1998; Cuesta et al. 2001). In contrast, typical antipsychotics either do not improve (Lee et al. 1999) or even worsen cognitive performance (Spohn and Strauss 1989). Reduced cholinergic activity in post-mortem brain has been reported to correlate with pre-mortem cognitive performance in schizophrenics (Powchik et al. 1998). Lower plasma NGF levels were reported in neuroleptic free schizophrenic patients (Bersani et al. 1999) as well as in HAL-treated chronic schizophrenic patients (Aloe et al. 1997). Recently, we reported that plasma NGF levels were significantly lower in never-medicated first-episode psychotic patients compared with those of normal age-matched controls. Moreover, NGF levels were substantially higher in chronic schizophrenic patients treated with atypical versus typical antipsychotics (Parikh et al. 2003b).

In summary, we previously reported that exposure to HAL, but not to RISP, clozapine, or OLZ, for 45 and 90 days was associated with a decline in ChAT immunoreactivities in various areas of the rat brain; however, spatial memory was affected by HAL only after 90 days of exposure (Terry et al. 2002, 2003). Data from the present study indicate that chronic exposure to HAL reduced the expression of NGF and ChAT but RISP or OLZ alone preserved or increased NGF protein in hippocampal neurons, which is reflected in the levels of ChAT. Hence, it may be postulated that at least one underlying mechanism for the differential effects of antipsychotics on the plasticity of cholinergic neurons may involve modulatory effects on NGF. Collectively, these preclinical data support the premise that at least some atypical neuroleptics via superior effects on NGF expression protect cholinergic pathways and cognitive function and are thus theoretically superior to HAL for extended use in human patients.



The authors acknowledge Ms. Karen Ship for proofing this manuscript. This study was financially supported in part by Janssen Pharmaceutica Research Foundation, Eli Lilly and Company, and The National Institute of Mental Health (MH 066233 to A.V.T.).


  1. Alberch J, Carman-Krzan M, Fabrazzo M, Wise BC (1991) Chronic treatment with scopolamine and physostigmine changes nerve growth factor (NGF) receptor density and NGF content in rat brain. Brain Res 542:233–240CrossRefPubMedGoogle Scholar
  2. Alleva E, Della Seta D, Cirulli F, Aloe L (1996) Haloperidol treatment decreases nerve growth factor levels in the hypothalamus of adult mice. Prog Psychopharmacol Biol Psychiatry 20:483–489CrossRefGoogle Scholar
  3. Aloe L, Iannitelli A, Bersani G, Alleva E, Angelucci F, Maselli P, Manni L (1997) Haloperidol administration in humans lowers plasma nerve growth factor level: evidence that sedation induced opposite effects to arousal. Neuropsychobiology 36:65–68PubMedGoogle Scholar
  4. Amaral DG, Kurz J (1985) An analysis of the origins of the cholinergic and noncholinergic septal projections to the hippocampal formation of the rat. J Comp Neurol 240:37–59PubMedGoogle Scholar
  5. Andersson C, Hamer RM, Lawler CP, Mailman RB, Lieberman JA (2002) Striatal volume changes in the rat following long-term administration of typical and atypical antipsychotic drugs. Neuropsychopharmacology 27:143–151CrossRefPubMedGoogle Scholar
  6. Angelucci F, Aloe L, Gruber SHM, Fiore M, Mathe AK (2000) Chronic antipsychotics treatment selectively alters nerve growth factor and neuropeptide Y immunoreactivity and the distribution of choline acetyltransferase in rat brain regions. Int J Neuropsychopharmacol 3:13–25Google Scholar
  7. Baxter MG, Frick KM, Price DL, Breckler SJ, Markowska AL, Gorman LK (1999) Pre-synaptic markers of cholinergic function in the rat brain: relationship with age and cognitive status. Neuroscience 89:771–779CrossRefPubMedGoogle Scholar
  8. Bersani G, Iannitelli A, Maselli P, Pancheri P, Aloe L, Angelucci F, Alleva E (1999) Low nerve growth factor plasma levels in schizophrenic patients: a preliminary study. Schizophr Res 37:197–203CrossRefPubMedGoogle Scholar
  9. Buckley PF (2001) Broad therapeutic uses of atypical antipsychotic medications. Biol Psychiatry 50:912–924CrossRefPubMedGoogle Scholar
  10. Bymaster FP, Hemrick-Luecke SK, Perry KW, Fuller RW (1996) Neurochemical evidence for antagonism by olanzapine of dopamine, serotonin, α1-adrenergic and muscarinic receptors in vivo in rats. Psychopharmacology 124:87–94Google Scholar
  11. Cuesta MJ, Peralta V, Zarzuela A (2001) Effects of olanzapine and other antipsychotics on cognitive function in chronic schizophrenia: a longitudinal study. Schizophr Res 48:17–28CrossRefPubMedGoogle Scholar
  12. Gallagher M, Nicolle MM (1993) Animal models of normal aging: relationship between cognitive decline and markers of hippocampal circuitary. Behav Brain Res 57:155–162PubMedGoogle Scholar
  13. Garofalo L, Ribeiro-da-Silva A, Cuello AC (1992) Nerve growth factor-induced synaptogenesis and hypertrophy of cortical cholinergic terminals. Proc Natl Acad Sci USA 89:2639–2643PubMedGoogle Scholar
  14. Green MF, Braff DL (2001) Translating the basic and clinical cognitive neuroscience of schizophrenia to drug development and clinical trials of antipsychotic medications. Biol Psychiatry 49:374–384CrossRefPubMedGoogle Scholar
  15. Harder JA, Baker JF, Ridley RM (1998) The role of the central cholinergic projections in cognition: implications of the effects of scopolamine on discrimination learn by monkeys. Brain Res Bull 45:319–326CrossRefPubMedGoogle Scholar
  16. Hefti F (1994) Neurotrophic factor therapy for nervous system degenerative diseases. J Neurobiol 25:1418–1435PubMedGoogle Scholar
  17. Ichikawa J, Dai J, O’laughlin IA, Fowler WL, Meltzer HY (2002) Atypical, but not typical, antipsychotics drugs increase cortical acetylcholine release without an effect in the nucleus accumbens or striatum. Neuropsychopharmacology 26:325–339CrossRefPubMedGoogle Scholar
  18. Jeste DV, Lohr JB, Manley M (1992) Study of neuropathologic changes in the striatum following 4, 8 and 12 months of treatment with fluphenazine in rats. Psychopharmacology 106:154–160Google Scholar
  19. Kapur S, Zipursky RB, Remington G (1999) Clinical and theoretical implications of 5-HT and D2 receptor occupancy of clozapine, risperidone and olanzapine in schizophrenia. Am J Psychiatry 156:286–293PubMedGoogle Scholar
  20. Kinon BJ, Lieberman JA (1996) Mechanisms of action of atypical antipsychotics drugs: a critical analysis. Psychopharmacology 124:2–34Google Scholar
  21. Korsching S, Auburger G, Heumann R, Scott J, Thoenen H (1985) Levels of nerve growth factor and its mRNA in the central nervous system of the rat correlate with cholinergic enervation. EMBO J 4:1389–1393PubMedGoogle Scholar
  22. Kromer LF (1987) Nerve growth factor treatment after brain injury prevents neuronal death. Science 235:214–216PubMedGoogle Scholar
  23. Lee MA, Jayathilake K, Meltzer HY (1999) A comparison of the effect of clozapine with typical neuroleptics on cognitive function in neuroleptic-responsive schizophrenia. Schizophr Res 37:1–11CrossRefPubMedGoogle Scholar
  24. Lee T-H, Kato H, Pan L-H, Ryu JH, Kogure K, Itoyama Y (1998) Localization of nerve growth factor, TrkA and P75 immunoreactivity in the hippocampal formation and basal forebrain of adult rats. Neuroscience 83:335–349CrossRefPubMedGoogle Scholar
  25. Lee T-H, Kato HK, Kogure K, Itoyama Y (1996) Temporal profile of nerve growth factor-like immunoreactivity after transient focal cerebral ischemia in rats. Brain Res 713:199–210CrossRefPubMedGoogle Scholar
  26. Levi-Montalcini R (1987) The nerve growth factor 35 years later. Science 237:1154–1162PubMedGoogle Scholar
  27. Lo DC (1995) Neurotrophic factors and synaptic plasticity. Neuron 15:979–981PubMedGoogle Scholar
  28. Mahadik SP, Laev H, Korenoversusky A, Karpiak SE (1988) Haloperidol alters rat CNS cholinergic system: enzymatic and morphological analyses. Biol Psychiatry 24:199–217PubMedGoogle Scholar
  29. Mahadik SP, Evans DR, Terry A, Hill WD (2001) Neuroprotective action of atypical antipsychotics in schizophrenia: improved cognitive performance and underlying mechanism of action. Schizophr Res 49:94Google Scholar
  30. Marchisio PC, Naldini L, Calissano P (1980) Intracellular distribution of nerve growth factor in rat pheochromocytoma PC12 cells: evidence for a perinuclear and intranuclear location. Proc Natl Acad Sci USA 77:1656–1660PubMedGoogle Scholar
  31. Mausset A-L, Seze R, Montpeyroux F, Privat A (2001) Effects of radiofrequency exposure on the GABAergic system in the rat cerebellum: clues from semi-quantitative immunohistochemistry. Brain Res 912:33–46CrossRefPubMedGoogle Scholar
  32. McKinney M, Coyle JT, Hedreen JC (1983) Topographic analysis of the innervation of the rat neocortex and hippocampus by the basal forebrain cholinergic system. J Comp Neurol 217:103–121PubMedGoogle Scholar
  33. Meltzer HY (1995) Atypical antipsychotics drugs. In: Bloom FE, Kupfer DJ (eds) Psychopharmacology: fourth generation of progress. Raven Press New York, pp 1277–1286Google Scholar
  34. Nakamura N, Fujita H, Kawata M (2002) Effects of the gonadectomy on immunoreactivity for choline acetyltransferase in the cortex, hippocampus, and basal forebrain of adult male rats. Neuroscience 109:473–485CrossRefPubMedGoogle Scholar
  35. Nilsson OG, Gage FH (1993) Anticholinergic sensitivity in the aging rat septohippocampal system as assessed in a spatial memory task. Neurobiol Aging 14:487–497CrossRefPubMedGoogle Scholar
  36. Parikh V, Khan MM, Salat P, Kalla A, Mahadik SP (2003a) Effects of risperidone, olanzapine and clozapine versus. haloperidol on nerve growth factor and cholinergic activity in rat brain. Biol Psychiatry 53:26SGoogle Scholar
  37. Parikh V, Evans DR, Khan MM, Mahadik SP (2003b) Nerve growth factor levels in never medicated first episode psychotic patients and medicated chronic schizophrenic patients. Schizophr Res 60:117–123CrossRefPubMedGoogle Scholar
  38. Paxinos G, Watson CR (1998) The rat atlas in stereotaxic coordinates, 4th edn. Academic Press, San Diego, Calif.Google Scholar
  39. Powchik P, Davidson M, Haroutunian V, Gabriel SM, Purohit DP, Perl DP, Harvey PD, Davis KL (1998) Postmortem studies in schizophrenia. Schizophr Bull 24:325–341PubMedGoogle Scholar
  40. Rossner S, Wortwein G, Gu Z, Yu J, Schliebs R, Bigl V, Perez-Polo JR (1997) Cholinergic control of nerve growth factor in adult rats: evidence from cortical cholinergic deafferentation and chronic drug treatment. J Neurochem 69:947–953PubMedGoogle Scholar
  41. Rylett RJ, Williams LR (1994) Role of neurotrophins in cholinergic-neuron function in the adult and aged CNS. Trends Neurosci 17:486–490CrossRefPubMedGoogle Scholar
  42. Seeman P (2002) Atypical antipsychotics: mechanism of action. Can J Psychiatry 47:27–38Google Scholar
  43. Sharma T, Mockler D (1998) The cognitive efficacy of atypical antipsychotics in schizophrenia. J Clin Psychopharmacol 18:12S–19SGoogle Scholar
  44. Shirazi-Southall S, Rodriguez DE, Nomikos GG (2002) Effects of typical and atypical antipsychotics and receptor selective compounds on acetylcholine efflux in the hippocampus of the rat. Neuropsychopharmacology 26:583–594Google Scholar
  45. Spohn HE, Strauss ME (1989) Relation of neuroleptic and anticholinergic medication to cognitive functions in schizophrenia. J Abnorm Psychol 98:367–380PubMedGoogle Scholar
  46. Terry AV, Hill WD, Parikh V, Evans DR, Waller JL, Mahadik SP (2002) Differential effects of chronic haloperidol and olanzapine exposure on brain cholinergic markers and spatial learning in rats. Psychopharmacology 164:360–368CrossRefPubMedGoogle Scholar
  47. Terry AV, Hill WD, Parikh V, Waller JL, Evans DR, Mahadik SP (2003) Differential effects of haloperidol, risperidone and clozapine exposure on cholinergic markers and spatial learning performance in rats. Neuropsychopharmacology 28:300–309CrossRefPubMedGoogle Scholar
  48. Thoenen H (1995) Neurotrophins and neuronal plasticity. Science 270:593–598PubMedGoogle Scholar
  49. Thoenen H, Bandtlow C, Heumann R (1987) The physiological function of nerve growth factor in the central nervous system: comparison with the periphery. Rev Physiol Biochem Pharmacol 109:145–178PubMedGoogle Scholar
  50. Tollefson GD (1996) Cognitive functions in schizophrenic patients. J Clin Psychiatry 57 (Suppl II):31–39Google Scholar
  51. Weinberger DR (1999) Cell biology of the hippocampal formation in schizophrenia. Biol Psychiatry 45:395–402CrossRefPubMedGoogle Scholar
  52. Woolf NJ, Eckenstein F, Butcher LL (1984) Cholinergic system in the rat brain: I. Projections to the limbic telencephalon. Brain Res Bull 13:751–784CrossRefPubMedGoogle Scholar

Copyright information

© Springer-Verlag 2003

Authors and Affiliations

  • Vinay Parikh
    • 1
    • 2
  • Alvin V. Terry
    • 3
  • Mohammad M. Khan
    • 1
  • Sahebarao P. Mahadik
    • 1
    • 2
  1. 1.Department of Psychiatry and Health BehaviorMedical College of GeorgiaAugustaUSA
  2. 2.Medical Research Service (242), Veterans AffairsMedical CenterAugustaUSA
  3. 3.Program in Clinical and Experimental TherapeuticsUniversity of Georgia College of Pharmacy (Augusta Campus)AugustaUSA

Personalised recommendations