Stereological Evidence of Non-Selective Hippocampal Neurodegeneration, Growth Factors Depletion, and Behavioral Decit Following Short-Term Bilateral Adrenalectomy in Wistar Rats

In the current study we investigated the impact of short-term adrenalectomy on hippocampal neurons of Wistar rats. In addition, the underlying mechanism(s) of degeneration in these neurons was investigated by measuring the levels of insulin-like growth factor-1 and β-nerve growth factor. Moreover, we investigated the cognitive behavior in these rats. The stereological counting in the hippocampus revealed a signicant neuronal death in the dentate gyrus and CA3, but not in the CA2 and CA1, area 7 and 14 days post adrenalectomy. The ultrastructural examinations revealed degenerated and degenerating neurons only in the dentate as well as CA4, and CA3 areas over the course of 3, 7 and 14 days. The levels of IGF-1 were signicantly decreased in the hippocampus of ADX rats 12 h post adrenalectomy and lasted over the course of two weeks. However, β-NGF was signicantly reduced in ADX rats only at 14 days postoperatively. Using passive avoidance test we found a cognitive decit in the ADX compared to the sham operated rats over time. In conclusion, both granule and pyramidal cells were degenerated in the hippocampus following short-term adrenalectomy. The early depletion of IGF-1 might play a role in the hippocampal neuronal degeneration. Consequently, the loss of the hippocampal neurons after adrenalectomy leads to cognitive decits.


Introduction
There is a growing evidence that prolonged exposure to glucocorticoids induces pyramidal cells loss in the hippocampus of guinea pigs 1 , rats 2 and vervet monkey 3 . Elevated concentration of glucocorticoids provide a clinical background to the pathogenesis of Parkinson and Alzheimer disease 4 . Furthermore, the administration of high levels of glucocorticoids believed to exacerbate pyramidal cells death in variety of neurodegeneration models 2,5 . It is interesting to note that the only neuronal population that is vulnerable to such high levels of glucocorticoids is the pyramidal cells of the hippocampus.
In contrast, in the late eighties Sloviter et al. 6 showed high dependence of granule cells in the hippocampus of Long Evans rats on glucocorticoids for their survival. They discovered that the withdrawal of glucocorticoids by adrenalectomy triggers a massive granule cells loss and spare the degeneration of pyramidal cells Sloviter et al. 6 . Several other groups replicated the ndings 7-10 .
However, Sapolsky et al. 11 were the rst to show that in addition to the degeneration of hippocampal granule cell, a signi cant decrease in the number of CA4 pyramidal cells has taken place following the adrenalectomy. Our previous study 12 supported and extended the nding of Sapolsky et al. 11 . We found, in our previous studies, using cell counting and electron microscopy, long-term adrenalectomy of Wistar rats caused pyramidal cells death not only in the CA4 but in the different areas of the Cornu Ammonis CA3, CA2, and CA1 beside a drastic and extensive degeneration in the granule cells of the dentate gyrus 12,13 . Furthermore, Martinez-Carlos et al. 14 demonstrated apical dendritic atrophy in CA3 region of the hippocampus following adrenalectomy.
The insulin-like growth factors (IGFs) are polypeptides with a sequence similar to insulin. IGF-I and IGF-II receptors are abundantly expressed in the brain where IGF signaling mediates neuronal growth, development, nervous system maintenance, myelination and synapse formation 15 . These roles are attributed to their ability to cross the blood-brain barrier and their endocrine roles in the brain 16 . A number of studies have observed the involvements of IGFs in the brain cognitive functions 17 .
Beta-Nerve growth factor (β-NGF) is also a polypeptide found in the central nervous system (CNS) in which the hippocampus contains one of the highest levels of such hormone 18 . β-NGF is an essential neurotrophic factor in the developing sympathetic and sensory nervous systems 19 . It exerts a number of different effects on neurons, such as neurogenesis 20 , neuronal plasticity 21 development 22 , , and differentiation 23 . Neurons that fail to obtain su cient β-NGF die by apoptosis 24 . It has been shown that adrenalectomy in young rats caused a drastic decrease of the β-NGF levels in the hippocampus 25 . Glucocorticoids and these growth factors play important roles in the brain, nonetheless, their interactions remain unclear although some studies have suggested a relation between them 26 . Hence, the hormonal activities of glucocorticoids were investigated whether they have an effect on the levels of β-NGF in the hippocampus. It was found that in developing rats, bilateral adrenalectomy considerably diminished β-NGF levels in the hippocampus and in the distribution of its receptor 25 .
The hippocampus is known for its major role in learning and memory, several investigations indicated a decline in memory performance tasks in ADX rats which highlights the possibility of the involvement of neuronal damage in the poor cognitive performance 12,[27][28][29] . Therefore, we aimed in the current study to investigate cell death in the hippocampus after short-term adrenalectomy and explore whether both granule and pyramidal cells are vulnerable to the absence of glucocorticoids. In addition, we aimed to explore the impact of short-term adrenalectomy on the levels of two main growth factors IGF-1 and β-NGF. Moreover, we examined whether the neuronal death of different populations has an impact on the behavior of the animals.

Materials And Methods
All animal handling procedures and experimental protocols conducted in accordance with relevant guidelines and regulations and approved by the regional ethical committee of the College of Medicine and Health Sciences, UAE University, United Arab Emirates (RECA/01/05). In addition, study was carried out in compliance with the ARRIVE guidelines.

2-1-Animals and adrenalectomy
Eight-weeks old males Wistar rats 170-220 g were obtained from College of Medicine and Health Sciences animal facility (Al Ain, United Arab Emirates) and used in the current study. Under pentobarbital (35 mg/kg body weight, Ilium-Troy Laboratories, New South Wales, Australia) anesthesia, rats were subjected either to adrenalectomy or to sham operation (laparotomy) as described by Sloviter et al. 6 . Shaving the back of the animals was done by using an electric shaving machine (Wahl, Illinois, USA). The rat was placed on the surgical table ventrally and in order to perform an aseptic surgery, the back of the animal was cleaned with 70% ethanol (Sigma-Aldrich, Missouri, USA). Bilateral incisions were made into the skin and the dorsal muscle to access the peritoneal cavity. The size of the muscle incision was just big enough to expose the adrenal gland on the top of the kidney and some free space to remove it without leaving any intact residuals of the gland. The peri-adrenal fat was grasped using blunt forceps and the adrenal gland was exteriorized, with a tipped-blunt scissors a cut was made at the connective tissue between the kidney and the adrenal gland. The incision was sutured, and the animal was returned to the cage. The same procedure was applied to the sham operated animals except the removal of the adrenal gland.
All animals were placed in plexiglas cages, four rats per group, maintained in a temperature (22°C) and 70±5% of humidity. Rats were housed in a 12:12 h normal light-dark cycle and received ad libitum access to food and water throughout the experiment. ADX rats were provided with 0.9% saline instead of drinking water in order to maintain electrolyte balance and prevent the deleterious effects of sodium chloride insu ciency.

2-2-Determination of plasma corticosterone levels
The level of corticosterone (CORT) in the serum was used to assess the e cacy of adrenalectomy.
Approximately 1 ml blood samples were taken directly from the heart at the time of sacri ce. The sera samples were stored at -80 °C until CORT levels were measured by Enzyme Immunoassay (EIA) Kit (Life sciences, Lausen, Switzerland) according to the manufacturer's protocol. The procedure uses a polyclonal antibody to CORT to bind in a competitive manner. A standard curve in the range from 32 to 20,000 pg/ml was constructed to calculate CORT concentrations in the samples.

2-3-Stereological analysis
Stereological quanti cation of pyramidal and granule cells was performed on the hippocampus as described earlier 30 . The total number of animals used is 50 animals distributed as follows: 4 h (Sham = 5, ADX = 5), 24 h (Sham = 5, ADX = 5), 3 days (Sham = 5, ADX = 5), 7 days (Sham =5, ADX = 5) and 14 days (Sham = 5, ADX = 5). Brie y, three series of tissue sections were processed for cresyl fast violet (Nissl) staining to label neuronal cells. Stereological analysis was performed with a stereology workstation (Computer Assisted Stereological Toolbox (CAST) -grid, Olympus, Denmark). Slides to count were determined in systematic random sampling as described earlier 30 . The total number of pyramidal and granule cells was estimated using the optical dissector method. Counts included all neuronal cells distributed in a systematic-random fashion within an unbiased counting frame throughout the Hippocampus.

2-4-Transmission electron microscopy
At each designated time point (3,7 and 14 days), animals (n= 18) were deeply anesthetized with intraperitoneal administration of sodium pentobarbital (35 mg/kg body weight). Transcardial perfusion xation was performed through the ascending aorta, with 50 ml of 0.1 M phosphate-buffered saline followed by 300 ml of freshly prepared solution of Karnovsky'S xative, PH 7.2 at 3 days (n=6), 7 days (n=6) and 14 days (n=6) after the surgery. For electron microscopy samples were immersed in McDowell and Trump xative for 3 h at 25 °C as described previously 31 . Tissues were rinsed with phosphate buffer saline and xed with 1% osmium tetroxide for 1 h. Samples were washed with distilled water, dehydrated in graded ethanol and propylene oxide. Finally the tissue was embedded in Agar-100 epoxy resin at 65 °C for 24 h. Blocks were trimmed and semi thin and ultra-thin sections were cut with Reichert Ultracuts (ultramicrotome) approximately -3.7 mm relative to bregma according to Paxinos and Watson. 32 . The ultrathin sections (95 nm) mounted on 200 mesh Cu grids were contrasted with uranyl acetate followed by lead citrate double stain. The grids were examined and photographed under a Philips CM10 transmission electron microscope 31 .
IGF-1, a trophic factor for neurons, has also been shown to be an important regulator of cell metabolism, differentiation, and survival. The Quantikine® ELISA from R&D Systems® assay was used for the quantitative determination of rat IGF-1 in tissue homogenates of the hippocampi as described in the manufacturer's protocol. A monoclonal antibody speci c for rat IGF-I had been pre-coated onto a microplate. Any rat IGF-I present in the sample was bound by the immobilized antibody. An enzyme-linked polyclonal antibody speci c for rat IGF-I was added to the wells. Following the formation of bound antibody-enzyme reagent, a substrate solution yielded a blue product that turned yellow with the addition of the Stop Solution. The intensity of the color measured at 450nm was in proportion to the amount of rat IGF-I bound in the initial step.
β-NGF is a trophic factor for neurons and is involved in the maintenance of the sympathetic and sensory nervous systems. A DuoSet® ELISA from R&D Systems® assay was used for the quantitative measurement of rat β-NGF in hippocampal tissue homogenate as described in the manufacturer's protocol. We used a speci c antibody for rat β-NGF, goat anti-rat β-NGF, coated on a 96-well plate. The β-NGF present in a sample was bound to the wells by the immobilized antibody. Later, biotinylated goat anti-rat β-NGF antibody was added followed by HRP conjugated streptavidin. The addition of the substrate solution (1:1tetramethylbenzidine and H 2 O 2 ) allowed color development in proportion to the amount of β-NGF bound. Finally, the Stop Solution (2 N sulfuric acid; H 2 SO 4 ) changed the color from blue to yellow. the intensity of the color was measured at 450 nm was in proportion to the concentration of β-NGF in the hippocampal homogenate.

2-6-Step-through passive avoidance test
In order to examine the impact of aforementioned biochemical and cellular alterations in the hippocampus on cognitive function the Passive Avoidance (PA) task was used as described earlier 33,34 . Fifty ve animals were used as follow: 3 days (Sham = 6, ADX = 9), 7 days (Sham = 8, ADX = 10) and 14 days (Sham = 7, ADX = 16). All animals were handled daily for at least 3 days before the experiments. The PA task was conducted using a two-compartment standard shuttle box (51x25x24 cm) (Harvard Apparatus, Massachusetts, USA). The two compartments were of equal size and had a stainless-steel bar oor connected by built in sliding door (7x7 cm). The right-hand compartment (shock compartment) was painted black to obtain a dark chamber, while the left-hand compartment was illuminated by a bulb (24V; 5W) installed on the top Plexiglas cover. PA training was conducted in a single session (day 1).
The rats were placed in the lighted compartment (with no access to the dark compartment) and were allowed to explore for 60s. During the exploration period in the PA apparatus, when 60s expired, the sliding door was automatically opened, and the rat was allowed to cross over into the dark compartment. Once the rat had entered the dark compartment with all four paws, the sliding door was automatically closed and a weak electrical current (constant current, scrambled, duration 3s, 0.3 mA) was delivered through the grid oor. Latency to cross into the dark compartment (training latency) was recorded. If a rat failed to move into the dark compartment within 300 s (cut off latency), the door was reopened and the rat was gently moved into the dark compartment, where it received the foot shock. After exposure to the foot shock, the rat was allowed to stay for 30s in the dark compartment before it was removed from the PA apparatus and returned to its home cage.
Retention was tested 24 h after training (day 2) 33,34 . The animal was again placed in the lighted (safe) compartment with access to the dark compartment for a period of 300s. The latency to enter the dark compartment with all four paws was automatically measured (retention latency). If the rat failed to enter the dark compartment within 300 s, it was removed and assigned a maximum test latency score of 300s.

2-7-Statistical analysis:
All data are reported as the mean ± standard error of the mean and the analysis was considered signi cantly different if P ≤ 0.05. Results were analyzed by two-tail Student t test SPSS version 25 (IBM, USA).
The CORT levels were signi cantly (P<0.01) reduced in the serum of 4 hr ADX rats (0.16 ± 0.16ng/ml) compared to sham operated rats (17.07 ± 4.81ng/ml). While at later time-points, circulating CORT levels were undetectable by the kit at concentrations lower than 0.103 ng/ml. A signi cant effect of postsurgical time was present for sham operated rats. Mean CORT levels were 64. 17

3-2-Stereological analysis
Our stereological results showed the effect of the removal of the adrenal gland is deleterious to the dentate gyrus of the hippocampus of the ADX compared to sham operated rats starting from the third day postoperatively. As shown in Fig.2 the rst three days following adrenalectomy the number of granule cells in the DG was affected but it did not reach a statistical signi cance. However, after 7 days postoperatively the number of the granule cells decreased signi cantly (P<0.001), and such effect continued to be seen 14 days later in the ADX rats compared to the sham operated rats (P<0.001) .
Our results showed the effect of adrenalectomy on the hippocampus is not selective. The removal of the adrenal glands leads to the reduction in the levels of glucocorticoids to undetectable levels that caused ultimately the death of different types of the neuronal cells throughout the hippocampus. The number of pyramidal cells in the CA1 was not affected signi cantly over the course of time (4h, 24h, 3 days, 7 days and 14 days) in both ADX and sham operated rats (Fig.3).
Similarly, to what we have seen with the pyramidal cells of CA1 the removal of the adrenal gland did not affect the neuronal population of the CA2 area (Fig.4).
Our stereology results showed that the removal of the adrenal gland affect directly the survival of the CA3 pyramidal cells in the hippocampus. We observed signi cant loss (P<0.001) of CA3 neurons 7 days and 14 days post adrenalectomy in the ADX rats compared to sham operated rats (Fig.5).

3-2-Transmission electron microscopy
In our previous ultrastructural study, we reported that the absence of glucocorticoids for long period of time (5 months) induces non selective degeneration in the hippocampus where different neuronal populations were affected through exhibiting signs of abnormalities and different cell death types were observed 13 . In the current study we aimed to examine the impact of the absence of such hormones on different cells of the hippocampus at an early stage and at different time points (3, 7 and 14 days).
The examination of thin sections of the hippocampi of sham operated and ADX rats under the electron microscope three days postoperatively showed granule cells degeneration on the tip of the upper blade of the dentate gyrus where the cells exhibited condensed chromatin and irregular cell membrane and the beginning of vacuolation in the cytoplasm only in ADX rats (Fig.6 A,B,C,D). Interestingly we have not noticed cell abnormalities in the rest of the dentate gyrus of the ADX rats (Fig.6 B,D).
Concerning the different areas of the Cornu Ammonis (CA areas) three days post adrenalectomy, we found solely CA4 pyramidal cells displayed signs of cell damage following the removal of the adrenal gland of ADX rats (Fig.7 B,D) compared to sham operated rats (Fig.7 A,C). Nevertheless, pyramidal cells of CA1, CA2 and CA3 cells did not show any signs of neurodegeneration after three days of surgery in both groups (Results are not shown).
Seven days following adrenalectomy the ultrastructural examination of the hippocampus revealed a progression in neurodegeneration compared to the third day of adrenalectomy where cell death was seen all over in both blades of the dentate gyrus and more extensively on the tip of the upper blade where we observed more exacerbation of the neurodegenerative process (Fig.8A'). In addition to what we have seen on the third day after adrenalectomy more degenerative cells were observed in the CA4 seven days postoperatively (Fig.8B'). Electron microscopy examination of the neuropil in the CA3 area of the hippocampus of ADX rats revealed cell death of the pyramidal cells for the rst time seven days following adrenalectomy (Fig.8C'). No sign of degeneration was seen in the CA1 and CA2 of the adrenalectomized rats. The hippocampus of the sham operated rats did not show any abnormalities in the different neuronal populations (Fig. 8A, B, C) We observed 14 days post adrenalectomy, an extensive cell death on the tip of the upper blade of the dentate gyrus followed by less degree of cell death in lower blade. Through the examination of granule layer of the dentate gyrus, we observed differences in the gross appearance of the neuropil between the ADX and sham operated rats. In the ADX animals, the granule layer showed empty areas with multiple ''blank'' spaces that indicated already degenerated cells. In addition to the changes in the neuropil, we observed ultrastructural changes in the upper and lower blade of the dentate gyrus (Fig.9). We have observed 14 days following adrenalectomy the condensation of chromatin beside the shrinkage of the cell body revealing both apoptotic and necrotic process occurring in the dentate gyrus of the hippocampus. These observations indicate the major vulnerability of the granule cells of the hippocampus to the absence of glucocorticoids achieved by the removal of the adrenal gland (Fig.10). In contrast, in the sham operated rats the neuropil was continuous with abundant, tightly packed granule cells. As the granule cells are small neurons, the nucleus occupied most of the cell body with only a thin rim of cytoplasm surrounding it. The examination of the Cornu Ammonis of the ADX rats revealed an extensive cell death in the CA4 area where chromatin condensation and vacuolation were seen (Fig.11).
Electron microscopy examination of the CA3 area of the ADX rats, 14 days post adrenalectomy, revealed the occurrence of cell death of the pyramidal cells along the principal layer of this area where extensive chromatin condensation, and reduction of the soma volume and the wavy shrinkage of the nuclear membrane and the vacuolization of mitochondria occurs (Fig.12). Sham operated rats exhibited healthy clustered pyramidal cells with regular plasma membrane, and well de ned nucleus with consistent dispersed chromatin. More importantly, our results showed no ultrastructural abnormalities of the CA1 CA2 pyramidal cells of both groups (Results not shown).

3-5-Growth Factors Analysis
Using ELISA, our results showed differences in IGF-1 protein levels. In the hippocampus, the IGF-1 protein levels were signi cantly decreased in the ADX rats at 12h (P<0.01) compared to the sham operated rats. The results remained consistently low at 24h (P<0.01), 3 days (P<0.05), 7 days (P<0.01), and 14 days (P<0.05) in the hippocampus of ADX rats compared to the sham operated (Fig. 13).
The evaluation of the levels of β-NGF in the hippocampus homogenates of ADX rats exhibited a signi cant decrease (P<0.05) in β-NGF levels only at 14 days postoperatively compared to sham operated rats. It is worth noting that there was no statistical difference between the groups prior to 14 days (Fig.   14).

3-6-Animal behavior test
Based on our histological and biochemical ndings we expected a negative impact on the behavior of the ADX rats and hence Passive Avoidance Test was performed. The evaluation of the latency time for both ADX and sham operated rats over the course of time (3, 7 and 14 days) indicated a signi cant increase (P<0.05) on the third day following surgery (Fig.15).
Furthermore, at day seven a signi cant increase (P<0.05) in the latency time was observed in ADX rats compared to the sham (Fig.16).
Our results showed 14 days following adrenalectomy a signi cant increase in the latency time persisted between the two groups (P<0.05) (Fig.17).

Discussion
In the current study using stereological techniques and ultrastructural examination we investigated the impact of short-term adrenalectomy on different hippocampal neuronal populations in Wistar rats. In addition, the underlying mechanism(s) of degeneration in these neurons was investigated by measuring the levels of two important growth factors in the hippocampus, Insulin-like growth factor-1 (IGF-1) and βnerve growth factor (β-NGF). Moreover, we examined whether the biochemical and histological changes in the hippocampus, after short-term adrenalectomy, have an impact on the behavior of Wistar rats.
Our results showed, half an hour following the bilateral removal of the adrenal gland, a drastic depletion in the levels of corticosterone in the ADX compared to the sham operated rats. Such decrease persists over the course of time of the experiment (0.5, 2, 4, 12 hours, and 1, 3, 7, 14 days). The adrenal gland is known as the major producer of corticosterone in the body and the signi cant depletion that was observed in the ADX rats is attributed to its removal 6,35-38 .
The impact of long-term removal of the adrenal gland on neurodegeneration of hippocampal neurons has been a subject of extensive research. However, there is a controversy as to whether long-term removal selectively affects the granule cells or not. Long-term withdrawal of the glucocorticoids by adrenalectomy was wrongly suggested as a selective model of granule cells' degeneration 6,28,39,40 . However, Sapolsky et al. 11 and Adem et al. 12 showed indeed the pyramidal cells are also vulnerable to such hormonal manipulation. We therefore aimed to investigate the effect of short-term adrenalectomy on different neuronal populations of the hippocampus.
We have previously shown, using Fluoro-Jade B (FJB) staining, few dentate granule cells death on day three which increased progressively on day 7 and 14 post adrenalectomy 41  Few studies have investigated the relationship between glucocorticoids and IGF-1. It has been demonstrated that postnatal injection of dexamethasone causes a signi cant decrease in the levels of IGF-1 accompanied by cell death in the hippocampus 42 . In line with the later, it has been shown that in Alzheimer's disease the cortisol levels are signi cantly elevated with low levels of IGF-1 43 . In contrast, in the current study after short-term bilateral adrenalectomy we observed a signi cant reduction in IGF-1 levels in the hippocampus at 12 hours, 1 day, 3 days, 7 days, and 14 days. Our present results support our previous ndings 44 where we observed an alteration of IGF-1 receptor and its messenger ribonucleic acid (mRNA) in the hippocampus after long-term adrenalectomy using in vitro receptor autoradiography and in situ hybridization immunohistochemistry, respectively. Signi cantly decreased levels of IGF-1 receptor and its mRNA were noted in the DG and CA1-CA4 regions of the hippocampus after long-term adrenalectomy suggesting that the level and expression of IGF-1 receptors in the hippocampus is in uenced by adrenal hormones. In the present study the signi cant reduction of IGF-1 seen at 12 hrs after adrenalectomy takes place long before the death of the neurons which is observed on day three post adrenalectomy. Our results suggest that bilateral adrenalectomy increases susceptibility of hippocampal neurons to degeneration through early decrease in IGF-1 levels. These ndings indicate that IGF-1 exerts potent neurotrophic and neuroprotective/ antiapoptotic activities in the hippocampus as suggested previously [45][46][47] .
In contrast to IGF-1, the levels of β-NGF were signi cantly reduced in the hippocampus only 14 days post adrenalectomy. Our ndings are in line with those of Aloe 25 who found a signi cant decrease in the levels of β-NGF in young adult rat hippocampus 12 days after adrenalectomy. Taken together these ndings indicate that β-NGF might not be involved in the early neurodegeneration seen in the hippocampus after bilateral adrenalectomy. In addition to IGF-1 other trophic factors, like β-NGF and others, which have impact on hippocampal neurons are affected later after bilateral adrenalectomy and may exacerbate the neurodegenerative process in the hippocampus.
The hippocampus is well known for its role in cognition 48,49 , most importantly in learning and memory 50 . As it was aforementioned due to the high concentration of glucocorticoid receptors in the hippocampus, it is considered the most targeted area of the brain by adrenal hormones 51,52 . Several studies have shown the withdrawal of these hormones causes degeneration of different neuronal cell populations in the hippocampus. The impact of the neuronal damage caused by long-term absence of adrenal hormones on the behavior of rats was investigated by different investigators. We showed 19 weeks post adrenalectomy, a signi cant increase in the latency in the Morris maze task and a signi cant lower rearing scores by the ADX rats compared to the sham operated 53 . A signi cant decrease in ambulatory and rearing activities was observed following chronic adrenalectomy, while grooming and defecation scores were not altered 54 . Moreover, Conrad and Roy 55 showed that after long-term adrenalectomy, the ADX rats exhibited di culties in acquiring a new spatial memory. Furthermore, Spanswick et al. 29 showed the inability to reverse such de cit in the spatial memory of long-term ADX rats by chronic treatment (6 weeks) with corticosterone or by an alternative neurogenic compound, Fluoxetine. The later nding indicates the crucial role of hippocampal formation in building memories and acquiring new information. In the current study, using the passive avoidance test we investigated the effect of short-term adrenalectomy on cognitive functions. Our results of the passive avoidance test showed that the ADX rats failed to retain their cognitive capacity compared to the sham operated where we have seen in days 3, 7 and 14 a signi cant increase in the latency of the ADX rats revealing a cognitive decline in these rats. These ndings are in line with those of Oitzl and de Kloet 56 who found that adrenalectomy impaired the performance of rats in the maze 3 days after the surgery. Moreover, ADX rats were also signi cantly less active than controls on days 3 and 4 of testing as a result of a greater drop in activity over the four days of testing 57 . Taken together our present results and those of others 53,55,56 are in line with our previous results of long-term adrenalectomy and suggest that cognitive behavior impairment might start as early as day 3 following glucocorticoid withdrawal.
In conclusion, our results showed hippocampal granule and pyramidal neurons degeneration after shortterm adrenalectomy. The signi cant reduction of IGF-1 in the hippocampus seen at 12 hrs after adrenalectomy takes place long before the death of the neurons which is observed on day three post adrenalectomy. Our results suggest that bilateral adrenalectomy increases susceptibility of hippocampal neurons to degeneration through early decrease in IGF-1 levels. These ndings indicate that IGF-1 exerts potent neurotrophic and neuroprotective/antiapoptotic activities in the hippocampus. Consequently, the loss of the hippocampal neurons after adrenalectomy leads to cognitive behavioral de cits.