Presented at the 6^th International Conference on Neuroprotective Agents
Hilton Head, September, 2002

The Ultra Low Weight Glycosaminoglycan, C3, Protects Against Loss of Dendritic Branching and Spines of Pyramidal Cells Following Cortical Damage in the Rat

Ronald F. Mervis¹, Adam Bachstetter¹, Bertalan Dudas², Umberto Cornelli², and Israel Hanin².

¹NeuroStructural Research Laboratories, Columbus, Ohio 43212, USA
²Department of Pharmacology and Experimental Therapeutics,
Loyola University Medical Center, Maywood, IL 60153, USA

Introduction and Methods
The group of glycosaminoglycans (or GAGs) is composed of heparin, heparan sulfate, dermatan sulfate, and chondroitin sulfate, and their analogs, including a novel GAG, C3. C3 is an ultra-low molecular weight analog of heparin produced by fractionation of heparin via g-irradiation, followed by gel filtration. It is comprised mainly of 6-10 oligosaccharides, with a molecular weight of 2.1 kD, and a USP value of 12 U/mg. C3 can cross the blood brain barrier, and in previous studies has been shown to have neurotrophic properties resulting in an increase in dendritic branching and spines in young adult rats. Other investigations have indicated that it may beneficial in neurodegenerative disorders such as Alzheimer’s disease.

As part of a larger series of studies investigating C3, one group of adult rats had previously been subjected to saline injections into their lateral ventricles (ICV), with the needle cannulae inserted bilaterally through the frontal cortices (saline group, N = 5). Another group of rats, which also had an ICV saline injection, had been given C3 orally (once daily, by gavage, 25mg/kg) beginning one week prior to the ICV saline and continuing for one week following saline administration (C3 group, N = 4) . Total treatment time for C3 was thus 14 days.

Initial observations of Golgi-stained neurons from the region adjacent to the needle track showed damaged cortical neurons as assessed by dendritic atrophy and spine loss. In view of findings from earlier studies suggesting the neurotrophic potential of C3, we decided to assess whether oral C3 treatment may have a neuroprotective role against the cortical stab wound-like damage caused by the needle inserted into the ventricle.

Dendritic branching and spines were quantitatively evaluated from Golgi stained tissue. As dendritic material of the cortical neuron comprises over 95% of its volume and the vast majority of synapses occur on spines, these dendritic parameters mirror the integrity and sophistication of cortical circuitry as well as the general health of the neurons.

Following 14 days of C3 treatment or saline, the rats were euthanized, their brains removed and immersion fixed in 4% paraformaldehyde for subsequent Golgi staining. 3mm coronal blocks of frontal cortex encompassing the needle cannula track were stained using the Rapid Golgi method and all slides were coded for neuronal analysis by an observer blind to treatment. In addition to the layer V pyramidal neurons selected adjacent to the needle track from the C3 treated or untreated animals, a third group of layer V pyramids were evaluated from saline treated subjects (N = 5): randomly selected neurons located in the same region of frontal cortex – but distal from the needle track. (and hence, undamaged). These neurons represented a baseline control by which to compare with the damaged neurons.

For assessment of dendritic branching, camera lucida drawings were made of the basilar tree of randomly selected, well-stained layer V pyramidal cells (n = 5 per subject). The basilar trees were evaluated for extent and distribution of their arbors by the Sholl method of concentric circles. This also provides an estimate of total dendritic length. Dendritic spines were counted along 30 micron segments from approximately 4 terminal tips of the basilar dendritic branches of each neuron (5 neurons per subject).

Results and Conclusions
Compared to the (undamaged) layer V pyramids distal from the needle track, e.g., baseline controls, neurons adjacent to the needle track from the untreated (saline) subjects showed a significant loss of both dendritic material (-15%, Sholl analysis: p < 0.001, Wilcoxon) and dendritic spines (–21.4%, p < 0.0001, Mann-Whitney). This indicated that the needle cortical stab wound resulted in significant damage to these neurons. By contrast, in comparison to the damaged neurons from the saline-treated subjects, the neurons from rats given C3 orally (for 7 days prior to, and 7 days after, the cortical stab wound) had significantly more dendritic material (p < 0.001, Wilcoxon) and 26% more spines (p < 0.0001, Mann-Whitney). Combining branching and spine data together (a so-called Circuitry Index reflecting estimated total synaptic connections), the cortical stab wound-induced neurotrauma resulted in a 33% reduction in circuitry in nearby neurons compared to neurons distal from the wound (p = 0.0079, Mann-Whitney). The C3 treatment protected neurons adjacent to the stab wound from this damage, the Circuitry Index now being the same as the baseline neuronal values (p = 0.9, N.S.).

In conclusion, C3 treatment has clearly exhibited a strong neuroprotective role maintaining normal cortical circuitry for neurons that would otherwise show significant neurotrauma-related injury. Further studies will be required to determine if C3 achieved this result primarily by means of minimizing initial damage and/or by enhancing neuroplastic recovery of the neurons from the cortical stab wound injury.

(Supported in part by NIA STTR Grant # 1-R41-AG15740-2 awarded to RFM and IH)

Presented at the meeting of the Society for Neuroscience,
Orlando, FL, November, 2002

Rapid Nonhuman Primate Model for NeuroAIDS: Evidence for Neuronal Damage.

R.F. Mervis¹, A. Bachstetter¹, H. Singleton¹, S. Henriksen², L. Madden², E. Roberts², M. Burudi², C. Marcondes², M. Taffe², and H. Fox².

¹NeuroStructural Research Labs, Columbus, OH 43212;
² Dept. Neuropharmacology, The Scripps Research Institute, La Jolla, CA 92037

CNS dysfunction induced by HIV infection, known as neuroAIDS, continues to be a significant problem in the AIDS pandemic. We have obtained a reproducible model of neuroAIDS in rhesus macaques, utilizing a CD8-depletion regimen at the time of SIV inoculation. Most of the monkeys show cognitive, motor, neurophysiological, and neuropathological abnormalities within four months after infection. To more fully characterize the neuronal changes accompanying the CNS disease, we utilized Golgi-impregnation staining to evaluate pathological alterations in dendritic branching and spines from multiple brain regions. Randomly selected pyramidal neurons located in layers II/III and V in the Anterior Cingulate Gyrus (ACG), the Middle Frontal Gyrus (MFG), the Superior Temporal Gyrus (STG) and in CA1 hippocampus were evaluated (3 subjects/group; 6 neurons/region/subject). From coded slides, camera lucida drawings were made of the basilar pyramidal trees. Sholl analysis was used to quantify the extent and distribution of neuronal dendritic arbors. Dendritic spines were counted along 30 micra lengths of terminal dendritic tip segments. In the SIV-infected monkeys the presence of damaged neurons was often in spotty patches and was characterized by dendritic varicosities, spine loss, and dendritic branch atrophy. Most areas showed significant loss of branching and/or spines with an average reduction of neural circuitry throughout each of these brain regions of almost 17%. These findings validate this reproducible nonhuman primate model for neuroAIDS. (Supported by NIH grants MH59468, MH62261, MH61692 and DA12444)

Presented at the meeting of the Society for Neuroscience,
Orlando, FL, November, 2002

IGF1 IS ESSNTIAL FOR NORMAL DENDRITIC GROWTH

Clara M Cheng¹, Ronald F Mervis², Shui-Lin Niu³, and Carolyn A Bondy^1.

¹Developmental Endocrinology Branch, National Institute Child Health & Human Development, NIH, Bethesda, MD 20892
²NeuroStructural Research Laboratories, Columbus, Ohio 43212
³Laboratory of Membrane Biophysics and Biochemistry, National Institute Alcohol Abuse and Alcoholism, NIH, Rockville, MD

  IGF1 gene deletion results in reduced brain growth and mental retardation. We have previously shown that brain glucose utilization is profoundly impaired during early postnatal development in the IGF1 null mouse. Given that IGF1 is normally selectively expressed by projection neurons that grow the most extensive dendritic arbors, we hypothesized that absence of IGF1 would result in attenuation of dendritic growth and synaptogenesis. Sholl analysis of the basilar trees of Golgi stained fronto-parietal cortical layer II-III pyramids in IGF1-/- and wild-type (WT) 50 day-old mice shows that IGF1-/- neurons have significant deficits in both dendritic length and complexity. IGF1-/- neurons have fewer dendritic intersections, particularly at increasing distances away from the soma (P=0.0002) and fewer branch points at each branch order measured indicating reduced branching complexity. IGF1-/- dendrites are ~ 25% shorter than WT (P<0.001). Dendritic spines are reduced by 16% (P = 0.002) and synaptotagmin levels are reduced by 30% (P = 0.01) in the IGF1-/- brain, suggesting decreased synaptogenesis. Similar findings were obtained in cortical layer V and in the piriform cortex. Cholesterol and phospholipid are each reduced by ~10% in IGF1-/- brains (P<0.01 and P<0.05 respectively), suggesting impaired lipogenesis in the IGF1-/- brain. These data show that IGF1 is essential for normal dendritic growth and synaptogenesis, probably through its insulin-like, anabolic effects on glucose utilization and lipogenesis.

Presented at the meeting of the Society for Neuroscience, Orlando, FL, November, 2002

Neuropathology of the Niemann Pick A (ASMKO) Mouse

S.L. Macauley¹, L. Shihabuddin*¹, E.H. Schuchman², R.F. Mervis³, T. Taksir¹, and G.R. Stewart¹

¹Genzyme Corporation, Framingham, MA 01701;
²Mount Sinai School of Medicine, New York, NY, 10029;
³NeuroStructural Research Laboratories, Columbus, OH, 43212

    This study describes the nature, distribution and progression of neuropathology in the Niemann Pick Type A (NP-A) mouse that was created by targeted disruption of the gene encoding acid sphingomyelinase (ASM). ASMKO (-/-) and normal littermate (+/+) animals from 4 to 28 weeks of age were prepared by fixation perfusion and in most cases the brains was processed for paraffin histology. In H& E stained specimens, lysosomal accumulation was readily apparent in the cytoplasm of ASMKO animals, particularly in larger neurons throughout the brain and spinal cord. Sensory neurons within dorsal root ganglion were also severely affected with massive lysosomal accumulation giving the cells a foamy appearance classically associated with visceral NP-A pathology. Mild lysosomal accumulation was present as early as six weeks and progressively worsened with age. Brain regions most severely affected included pyramidal cells in the cortex (layers II, III, V) and hippocampus (CA1), most subnuclei of the thalamus, and magnocellular brainstem structures such as the trigeminal nucleus. Despite advanced disease in older animals, the only region observed to undergo neurodegeneration was an almost complete depletion of Purkinje cells from the cerebellum (see poster by Stewart et al). Silver impregnation staining did reveal evidence for more widespread degenerating structures (see poster by Switzer). However, the use of retrograde tracers demonstrated that diseased neurons were able to efficiently transport dyes and that distant brain structures retained normal, intact connections (e.g. striatum to nigra). Further structural analysis using Golgi staining revealed relatively preserved fine neuronal morphology, although there was evidence for stripping of dendritic spines as well as meganeurite formation within some CA1 and cortical pyramidal neurons.

Presented at the meeting of the Society for Neuroscience,
Orlando, Fl, November, 2002

TETRACHLOROAZOBENZENE (TCAB)-INDUCED DEVELOPMENTAL NEUROTOXICITY IN CORTEX AND CEREBELLUM IN THE RAT:
A Quantitative Morphometric Golgi Analysis.

A. Bachstetter¹, R.F. Mervis¹, H. Singleton¹, G. Wolfe², J. Bishop³, and G. J. Harry³,

¹NeuroStructural Research Labs, Columbus, OH 43212;
²Therimmune Research Corp., Gaithersberg, MD 20878;
³NIH, National Institute for Environmental Health Sciences,
Research Triangle Park, NC 27709

   3,3’,4,4’-tetrachloroazobenzene (TCAB) is a dioxin-like organochlorine compound formed as a byproduct in the manufacture of herbicides. It impairs thyroid function, and thus controls critical aspects of brain development. The potential neurotoxic consequences of TCAB exposure need to be determined. TCAB was administered by gavage to SD female rats starting two weeks before pregnancy until the pups were weaned. TCAB dose levels were 0.1 mg/kg/day (low dose) or 3.0 mg/kg/day (high dose). Controls received corn oil. Male pups were sacrificed when 22 days old. Neocortex was stained using the Rapid Golgi method and cerebella, using the Golgi-Cox variant. Using coded slides, from each subject 6 randomly selected layer V pyramids and 6 Purkinje cells (PCs) were evaluated. Analysis of dendritic arbor of the V pyramids by Sholl analysis showed that control arbor > low dose TCAB > hi dose TCAB. Significant dendritic spine loss was seen only in the high dose TCAB. The Neuronal Circuitry Index (branching/spine index) showed that low dose TCAB reduced pyramidal cell circuitry by 18%, and high dose TCAB exposure, by 32%. In cerebellum, PC dendritic area was significantly reduced only in the high dose TCAB group. However, both low dose and high dose groups had small ectopic dysmorphic PC-like neurons located in the molecular layer. The results suggest that the impaired development of neocortical and cerebellar neurons may be mediated by thyroid dysfunction due to the TCAB exposure.

Presented at the meeting of the Society for Neuroscience,
Orlando, FL, November, 2002

DEVELOPMENTAL EXPOSURE TO HEXACHLOROBENZENE (HCB)
IN THE RAT RESULTS IN ECTOPIC AND DYSMORPHIC
CEREBELLAR PURKINJE CELLS

A. Bachstetter¹, R.F. Mervis¹, H. Singleton¹, G. Wolfe², J. Bishop³, and G. J. Harry³.

¹NeuroStructural Research Labs, Columbus, OH 43212;
²Therimmune Research Corp., Gaithersberg, MD 20878;
³NIH, National Institute of Environmental Health Sciences, Research Triangle Park, NC 27709

Hexachlorobenzene (HCB) is a dioxin-like organochlorine compound that binds to the aryl hydrocarbon receptor (Ah). HCB may impair thyroid metabolism. HCB is widespread in the environment and accumulates in biological systems. It is transplacentally transferred from mother to fetus during pregnancy and through breast milk. Neurotoxic consequences of HCB exposure on the developing brain are, however, poorly documented. HCB (2.5 mg/kg/day) was administered by gavage to Sprague-Dawley female rats starting two weeks before pregnancy until the pups were weaned. Controls received corn oil. Pups were sacrificed when 22 days-old and cerebella were stained using the Golgi-Cox method. There were seven male subjects per group. Using coded slides, from each subject seven randomly selected Purkinje cells (PCs) were evaluated for soma size and areal extent of their dendritic arbor. There were no differences between groups for PCs that had correctly migrated and were located in the Purkinje cell layer. However, there were also a significant number of neurons with Purkinje cell-like characteristics that had improperly migrated into the molecular cell layer. These highly dysmorphic PCs had dendritic areas 90% smaller than normal, and somas which were 75% smaller than normal. Since thyroid function influences critical aspects of cerebellar development, the abnormal migration of PCs and their dysmorphic development may reflect the neurotoxic impact of HCB exposure on thyroid function.

Presented at the meeting of the Endocrinology Society
San Francisco, CA, June, 2002

Endogenous Brain IGF1 is Essential for Normal Dendritic Growth.

Clara M Cheng, Ronald F Mervis, Shui-Lin Niu, and Carolyn A Bondy

Developmental Endocrinology Branch, NICHD, NIH, Bethesda, MD 20892
NeuroStructural Research Laboratories, Columbus, Ohio 43212
Laboratory of Membrane Biophysics and Biochemistry, NIAAA, NIH

IGF1 deficiency results in reduced postnatal brain development in both human and animal models. Our previous studies on IGF1 gene deleted mouse brain (IGF1-/-) have shown that despite the reduction in brain size, the IGF1-/- brain anatomy and cell numbers are normal for the most part. Given that the dendritic neuropil occupies a very large percentage of the brain parenchyma, the attenuation in IGF1-/-brain size may result from the deficit in the growth of dendritic processes in the absence of endogenous brain IGF1. To test this hypothesis, we used Golgi histological staining and biochemical analysis to evaluate the dendritic growth and complexity of cortical neurons in the IGF1-/- and wild-type (WT) brains of 50 day-old mice. Morphometric analysis of fronto-parietal cortical pyramidal layer II-III neurons shows that the soma size of IGF1-/- neuron is reduced by 10% (P<0.001), and that the cell density is significantly higher by 25% compared to that of WT (P=0.005). Analysis of camera lucida drawing of the basilar trees of the Golgi stained II/III pyramids shows that IGF1-/- neurons have less dendritic material than WT’s as reflected by deficits in both dendritic length and complexity. Sholl analysis of the basilar trees reveals that IGF1-/- neurons have fewer dendritic intersections with the measuring concentric circles, particularly at the increasing distances away from the soma (P=0.0002). Branch point analysis shows that IGF1-/- neurons have fewer branch points at each branch order measured indicating less branching complexity. IGF1-/- neurons have an average of 25% shorter dendrites than WT’s (P<0.001). Supporting this finding, brain total lipid, the major substance for constructing dendrites, is decreased by 29% in the IGF1-/- brain homogenates (P=0.059). Cholesterol and phospholipid, the two primary components of the brain lipid, are reduced by 9% in IGF1-/- brains individually (P<0.01 for cholesterol and P<0.05 for phospholipid). Further investigation of the enzymes involved in neuronal morphogenesis demonstrates that the cdc42, a member belonging to the Rho GTPase family, is reduced in IGF1-/- brains by 30% compared to WT’s (P<0.005). Rac1 and RhoA, the other two members in the Rho family, show no change in IGF1-/- brains. Our results suggest that the normal growth of dendritic processes is significantly impaired in the absence of endogenous brain IGF1 and that the cdc 42 is one of the major modulators for the action IGF1 in regulating brain dendritic development.

Presented at the forthcoming meeting of the International
Society for the Study of Fatty Acids & Lipids, Montreal, May, 2002

ALTERATIONS IN BEHAVIOR AND NEUROANATOMY IN DHA DEFICIENT RATS

Norman Salem, Jr.¹, Ronald F. Mervis², Aneeq Ahmad¹, Toru Moriguchi¹, Janice Nicklay¹, Rebecca Greiner¹, Sunyoung Lim¹, Adam Bachstetter², and Burton Slotnick^3.

¹Section of Nutritional Neuroscience, Lab of Membrane Biochemistry & Biophysics, NIAAA, NIH;
²NeuroStructural Research Laboratories, Columbus, Ohio;
³Department of Psychology, American University, Washington, D.C.

If docosahexaenoic acid (DHA) is essential for optimal nervous system function, then diets that lead to a loss in brain DHA would be expected to result in functional deficits. Brain function is usually assessed using behavioral tasks of varying complexity. In this work, rats were given diets with only safflower oil as their source of essential fats (n-3 Def group) or safflower oil plus sources of alpha-linolenate and DHA (n-3 Adq group) for two generations. Male pups (behavior) and female pups (anatomy) were weaned to the same diets that their mothers consumed and then tested at various ages. In spatial tasks, animals in the n-3 Def group acquired the Morris water maze more slowly and performed poorly on the retention trial. In a go, no-go olfactory discrimination task, n-3 Def rats again acquired the task more slowly and failed to acquire a learning set. Anatomical studies indicated smaller neuronal areas of cell bodies in the hippocampus as well as other brain areas. Golgi staining and Sholl analysis indicated that granule cells in the dentate gyrus had significantly less dendritic arborization in the n-3 Def animals at both 21 and 68 days of age than the n-3 Adq group. A similar loss of dendritic arborization was also observed in cortical pyramidal neurons in 21-day old n-3 Def rats, however, this difference was no longer found in 68 d old animals apparently due to greater dendritic pruning in the n-3 Adq brains. Persistent alterations in hippocampal circuitry in the n-3 Def animals may be the neurostructural basis for the observed behavioral deficits.

Presented at the Seventh International Geneva/Springfield
Symposium on Advances in Alzheimer Therapy, April, 2002

GLYCOSAMINOGLYCANS AS A POTENTIAL THERAPY FOR ALZHEIMER’S DISEASE

I. Hanin, B. Dudas, Q. Ma, M. Rose, J. Fareed, S.Lorens, M. Hejna,
J. Lee, M. Walzer, ^*R. Mervis, and U. Cornelli,

Loyola University Chicago Stritch School of Medicine, Maywood, Illinois, and ^*NeuroStructural Research Laboratories, Columbus, Ohio, USA.

Over the past decade a mixture of glycosaminoglycans (GAGs) has been shown to improve cognition in aged rats and to reverse cognitive and neurochemical deficits in patients with Alzheimer’s disease (AD). This mixture of GAGs (called GAP or GAG-polysulfate), which is composed of heparin (10-20%), heparan sulfate (40-60%), dermatan sulfate (20-35%) and chondroitin sulfate (2-8%), has earned a use-patent in several European countries. The effects mentioned above may be attributable to the ability of GAGs to compete with endogenous proteoglycans (PG), such as heparan sulfate PG, and to prevent their tendency to: a) induce amyloid formation via an increase in aggregation rate of b amyloid (Ab) into b-pleated fibrils and b) increase polymerization of tau into paired helical proteins, which make up the neurofibrillary tangles of AD. During the past three years we have been working to identify the active component(s) of the GAP mixture, by fractionating each of the components of GAP into smaller molecular weight fractions, and testing their respective protective activity on several animal models of brain neuron degeneration.

C3 is an ultra-low molecular weight analog of heparin produced by fractionation of heparin via g-irradiation, followed by gel filtration. It is comprised mainly of 4-10 oligosaccharides, with an apparent molecular weight of 2.1 kD, and a USP value of 12 U/mg. The fractionation procedure has been standardized; thus we can manufacture a reproducible product in large quantities (kgs). This substance has been found to be the most active of several products obtained via fractionation of the above-mentioned components of GAP).

To date we have observed, using C3, that: a) it can cross the blood brain barrier (BBB); b) it prevents tau-2 immunoreactivity and reactive astrocytosis in both the AF64A, and the Ab(25-35) treated rat; c) its protective effect is obtained following either oral or subcutaneous administration of C3; d) it has neurotrophic effects in the brain, in vivo, as measured by Golgi impregnation technology; e) it is devoid of significant anticoagulant activity; and f) it is safe at high doses.

Thus, C3 is a promising candidate for further development and clinical testing in patients with AD and other central neurodegenerative disease states. It will soon be employed in Phase I clinical studies.

(Supported in part by NIA STTR Grant # 1-R41-AG15740)

Presented at the meeting of the Society of Toxicology,
Nashville, April, 2002

LONG-LASTING NEUROSTRUCTURAL CONSEQUENCES IN THE RAT HIPPOCAMPUS BY DEVELOPMENTAL EXPOSURE TO A MIXTURE OF POLYCHLORINATED BIPHENYLS (PCBs).

R F Mervis¹, A D Bachstetter¹, G J Harry², H A Tilson^3, ,and P R S Kodavanti³.

¹NeuroStructural Research Labs, Columbus, OH, USA;
²NIH, NIEHS, RTP, NC, USA, and
³Neurotox Div., USEPA, NHEERL/ORD, Research Triangle Park, NC, USA.

The objective of the study was to assess the effects of developmental exposure to a commercial mixture of PCBs (Aroclor 1254) on neuronal dendritic morphology of hippocampal CA1 pyramidal neurons in postnatal day (PND) 22 and PND 60 male Long-Evans rats. Rat pups were born to mothers who were exposed to Aroclor 1254 (AccuStandard Inc., Lot # 124-191; 0 and 6 mg/kg/day) from gestational day 6 through PND 21. Thus, pups were exposed to PCBs in utero and through weaning. Male rats (N = 5-6 per group) were sacrificed on PND 22 and PND 60. Brains were formalin-fixed for Rapid Golgi staining of tissue blocks; coded slides of hippocampus were prepared. For branching and spine analysis, 6 CA1 pyramids were randomly selected from each brain. Camera lucida drawings of the basilar dendritic tree were analyzed using the Sholl method of concentric circles. For spine analysis, counts were made along internal and terminal tip segments of 6-7 neurons from each brain. Results of the branching analysis for the PND 22 PCB-exposed rats showed that, compared to controls, there was significantly less dendritic branching in the outer 2/3rds of the dendritic tree (p = 0.002, Wilcoxon test). Spine analysis also showed a reduction in spines on the terminal tips segments of 22 day old PCB-exposed rats (p=0.005, T-test). These results suggest that perinatal exposure to Aroclor 1254 resulted in morphometric changes in hippocampus. By PND 60, spine density on terminal tip segments had returned to normal levels. However, branching analysis now showed that compared to controls there was an excessive amount of dendritic material in the distal 2/3rds of the tree (p=0.001, Wilcoxon test). This suggested a possible structural "hyperplasticity" in neurons damaged by PCB exposure during the developmental period with a residual long-term dysmorphic impact on hippocampal circuitry. (This abstract does not necessarily reflect USEPA policy).

Presented at the 2^nd Annual Neurobiology of Aging Conference
San Diego, November, 2001

Global Dendritic Neuroplasticity in the Young Adult Rat Brain FollowingAdministration of C3, an Ultra Low Molecular Weight Glycosaminoglycan:  Implications for Treatment of Alzheimer’s Disease.

R.F. Mervis¹, H. Waters¹, A. Bachstetter¹, H. Singleton^1,2,
B. Dudas³, U. Cornelli³, J.M. Lee³, and I. Hanin³.

¹NeuroStructural Research Laboratories, Columbus, OH 43212;
²Capital University, Columbus, OH
³Loyola University Medical Center, Maywood, IL 60153

     Glycosaminoglycans (GAGs) are linear polysaccharides consisting of a hexosamine, dissacchride units, and sulfate substituents. Proteoglycans (PG), such as heparin sulfate PG, have been shown to facilitate the occurrence of classical neuropathological hallmarks of Alzheimer’s disease (AD) such as amyloid formation and increased polymerization of tau protein into neurofibrillary tangles. GAGs compete with endogenous PGs to inhibit these occurrences. Therefore, it was thought that treatment with GAGs might attenuate the neuropathological changes associated with AD and would be a useful treatment strategy. We now show that C3 also exerts widespread neuroplastic effects on dendritic branching in young adult rats.
     F344 rats (N=5) were administered C3 subcutaneously (s.c.) for 32 days (2.5 mg/kg s.c., b.i.d.). Controls were administered saline. Formalin fixed tissue blocks were Golgi-stained and coded slides prepared. Neurons were randomly selected from three areas for quantitative analysis of dendritic branching: layer V pyramids of frontoparietal cortex; CA1 pyramids of the hippocampus, and granule cells of the dentate gyrus. Camera lucida drawings were made from 6 neurons from each area and Sholl analysis (Method of Concentric Circles) was carried out to assess the amount of dendritic material and its distribution.
     In all three populations there was evidence of dendritic neuroplasticity. This was predominately seen by increased arborization which was significant for CA1s (p=0.001), for layer V pyramids (p=0.004), and for the outer half of the arbor of granule cells of the dentate gyrus (p =0.001). Since the vast majority of synapses are on dendritic branches, it is feasible that GAG-mediated structural plasticity would also result in additional synaptic connections. Since Alzheimer’s dementia is associated with synaptic loss, C3 treatment may be efficacious in maintaining brain circuitry and minimizing cognitive loss.

(Supported by NIH grant #1 R4 AG15740-01 to RFM, IH, and JML)

Presented at the meeting of the Society for Neuroscience, San Diego,
November, 2001

DIETARY DEFICIENCY OF DOCOSAHEXAENOIC ACID (DHA) IMPAIRS NORMAL DENDRITIC DEVELOPMENT IN RAT NEOCORTEX.

R.F. Mervis¹, A. Bachstetter^1,2, H. Waters^1, A. Ahmad³,
T. Moriguchi³, and N. Salem³.

¹NeuroStructural Research Laboratories, Columbus, OH 43212,
²Capital University, Columbus, OH 43209,
³Laboratory of Membrane Biochemistry & Biophysics, NIAAA, NIH, Rockville, MD 20852.

     Docosahexaenoic acid (DHA), an n-3 fatty acid, is a critical structural component of neuronal and synaptic membranes. Since DHA is rapidly deposited during the period of normal brain development, reduced dietary intake of n-3 fats leads to lowered brain DHA that, in turn, may cause losses in brain function. Currently, infant formula in North America is devoid of DHA. Rats were made deficient in brain DHA (DEF group) by limiting n-3 dietary fats through 3 generations. Controls received an n-3 adequate diet (ADQ group) that contained flaxseed oil and DHASCO. Rats were sacrificed at 21 and 68 days of age (N=8/group) and fronto-parietal cortex was Golgi-stained. From coded slides, 40 layer V pyramids were randomly selected from each group and camera lucida drawings were made of the basilar dendritic trees. Statistical analysis (Sholl Analysis, method of concentric circles) showed that in the 21 d group, there was 15% less dendritic material in the DHA DEF group compared to the ADQ group (p = 0.0001, Wilcoxin test). Between 21 to 68 days, dendritic pruning reduced the arbor of the ADQ control group by 25%, and of the DEF group by 13%. Thus, by 68 days the extent of dendritic branching in pyramid neurons from both groups was equivalent. Since DHA deficiency can result in long-term cognitive deficits, our findings suggest possible remaining neurostructural consequences that may result from n-3 deficiency.

Presented at the meeting of the Society for Neuroscience, San Diego
November, 2001

DENDRITIC AND SOMA SIZE ALTERATIONS IN HIPPOCAMPUS AND CORTEX OF RATS FOLLOWING NEONATAL BORNA VIRUS INFECTION

M. Hornig^1,2, R.F. Mervis⁴, H. Waters⁴; A. Bachstetter^4,5, H. Singleton^4,5, D. Chian²,
M. Sindelar², W.I. Lipkin^1,2,3*

Departments of Neurology¹, Microbiology & Molecular Genetics²,
Anatomy & Neurobiology³, UC Irvine, Irvine, CA, USA
⁴NeuroStructural Research Labs, Columbus, OH, USA
⁵Capital University, Columbus, OH, USA

     Altered neuronal morphology and cell losses are reported in postmortem brain of subjects with autism, including small neurons, decreased dendritic complexity in hippocampus, cerebellar Purkinje cell loss, and at least one report of cortical involvement. Neonatal infection of Lewis rats with Bornavirus, a neurotropic RNA virus, results in neuropathologic, behavioral, and neurochemical abnormalities resembling features of autism. Understanding the processes influencing synaptic dysfunction and selective neuronal losses by apoptosis in this animal model may reveal mechanisms important in neurodevelopmental disorders, including autism. Using Golgi-stained tissue, we evaluated changes in neuronal morphology in dentate gyrus and layer V of fronto-parietal cortex in neonatally infected rats. At 3 weeks postinfection, dendritic branching was decreased in dentate gyrus granule cells (Sholl analysis, p=0.0004) and in layer V pyramidal cells (p=0.0006). Distal dendrites were especially affected. The soma of the granule cells (but not of cortical neurons) were significantly smaller (p=0.0043). Other changes included branch atrophy, spine loss, and varicosities. These findings suggest a morphologic substrate underlying processes that disrupt synaptic plasticity or increase vulnerability to apoptosis during brain development, and support the use of this animal model for uncovering fundamental aspects of neurodevelopmental disorder pathogenesis.
Supported by: NIH grants K08 MH01608 and RO1 HD37546

Presented at the Yale University School of Medicine Symposium on
Cognitive Enhancers, Anti-Alzheimer and Neuroprotective Drugs
New Haven, October, 2000

Neurotrophic Effects of the Glycosaminoglycan C3 on Dendritic Arborization and Spines in the Adult Rat Hippocampus: A Quantitative Golgi Study.

R. F. Mervis¹, J. McKean¹, S. Zats^1,2, A. Gum^1,2, R. Reinhart¹, B. Dudas³, U. Cornelli³,
J.M. Lee³, S.A. Lorens³, J. Fareed³, and I. Hanin³.

¹NeuroStructural Research Laboratories, Columbus, Ohio 43212.
²Ohio Dominican College, Columbus, Ohio
³Dept. of Pharmacology and Experimental Therapeutics,
Loyola University Chicago Stritch School of Medicine, Maywood, Illinois 60153.

Glycosaminoglycans (GAGs) may be effective as a therapeutic strategy in the treatment of Alzheimer’s Disease (AD). Previous work from this group using a rat model showed that a single intra-amygdaloid injection of A beta (25-35) could induce abnormal tau-immunoreactive perikarya in the hippocampus. Furthermore, administration of the GAG C3, an ultra low molecular weight heparin mixture of 4-10 oligosaccharides (MW~2.1) could decrease Abeta25-35-induced tau-2 immunoreactivity in this model of AD.

In this study, we evaluated the effects of stereotaxic intra-amygdaloid injection of A Beta (25-35) 5nml/3ul)and of C3-treatment (administered subcutaneously, s.c.) on dendritic morphology of Golgi-impregnated CA1 pyramids of the hippocampus. Subjects were young-adult, male F344 rats.

There were four groups:

• the controls were rats with intra-amygdaloid (i-a) injection of vehicle (trifluoroacetic acid, TFA, 3ul) followed by saline (32 days, s.c., 1ml/kg, b.i.d.)(N=10);
• treatment groups consisted of:
• i-a injection of Abeta unmollfollowed by 32 days of saline (s.c., 1ml/kg, b.i.d.)(N=5); i-a injection of Abetanmollfollowed by GAG-treatment (32 days of C3, 2.5 mg/kg s.c., b.i.d.)(N=5); andi-a injection of vehicle (TFA, 3ul) followed by the C3 GAG (32 days, 2.5 mg/kg s.c. , b.i.d.)(N=5).

Formalin-fixed coronal blocks of rat frontoparietal cortex, which included the underlying hippocampus, were stained using the Rapid Golgi method. Slides were sectioned at 120 m. All brains were coded, and 6 well-impregnated CA1 neurons were randomly chosen from each subject from the hemisphere ipsilateral to the intra-amygdaloid injection of the A25-35or the vehicle. For analysis of dendritic branching, camera lucida drawings were made of the basilar tree. The drawings were then quantified by the Sholl analysis (method of concentric circles) which provides a profile of the extent and distribution of the dendritic material. Dendritic spines were quantified on 4-5 randomly selected terminal tip segments from each neuron.

Results. Compared to the controls (TFA vehicle), dendritic branching of the CA1s was not diminished by A however, dendritic spine density was significantly attenuated (p<0.001). GAG (C3) treatment in the Atreated animals had only a minor effect on increasing dendritic branching (p=0.039, Wilcoxon signed-rank test) and had no impact on spines. However, C3 GAG-treatment alone (e.g., in the TFA vehicle-injected controls) had a remarkable neurotrophic impact on both dendritic branching and spines. C3 treatment increased total dendritic length by 36% (p<0.001) while increasing branching complexity (p<0.05). Simultaneously, dendritic spine density on the terminal tips was increased by 17%. Combining the C3-induced neurotrophic changes in both branching and spines, it was determined that 32 days of C3-treatment resulted in an overall 61% increase in total synaptic contacts on the basilar tree of the CA1 pyramids in these young-adult rats (p<0.001).

These results indicate that the GAG C3 appears to have a remarkable neurotrophic impact on the hippocampal circuitry of the brain of the normal young-adult rat. The apparent inability of the GAG treatment to produce the same effect in the A-injected AD model may suggest that a more extended time frame (and/or a different dosage) is required for the damaged neurons to respond to C3 treatment. These and other related studies (including the assessment of the C3 treatment on neurons in aged rats) are in progress. In addition, while it possible that C3 treatment could enhance learning or memory, it is also feasible that the extensive neurotrophic response in the young rat may be deleterious to learning (due to the "noise" induced by the presence of inappropriate synapses). This, too, is the subject of future study.

The present findings demonstrate a clear and dramatic effect of the GAG C3 on inducing an extensive neurotrophic effect on both dendritic branching and spines in the undamaged hippocampus of the adult rat. While awaiting the results of future and more targeted studies, they also suggest that such treatment may be useful in the treatment of AD or in attenuating age-related dendritic atrophy. The study also demonstrates the value of analysis of dendritic branching and spines from Golgi-impregnated tissue to demonstrate and quantify this structural neuroplasticity.

(Supported by HHS STTR grant # AG15740-01awarded to RFM and IH)

Presented at the Society for Neuroscience, October, 2000, New Orleans

ATROPHIC DENDRITIC BRANCHING OF CORTICAL PYRAMIDAL NEURONS RESULTS FROM INSULIN-LIKE GROWTH FACTOR 1 (IGF1) GENE DELETION:

A QUANTITATIVE GOLGI ASSESSMENT.

R.F. Mervis¹, J. McKean¹, S. Zats^1,2, C. Cheng³, A. Gum^1,2 and Bondy,C.A³

¹NeuroStructural Research Laboratories, Columbus, Ohio 43212, USA;
²Ohio Dominican College, Columbus, Ohio, USA;
³Developmental Endocrinology Branch, NICHD, NIH, Bethesda, MD 20892, USA;

Based on its high level expression in the developing brain and on the finding of mental retardation in the case of human IGF1 gene deletions, IGF1 is thought to play an important role in brain development. The murine Igf1 (-/-) brain demonstrates preserved cell numbers in most structures, but a generalized reduction in neuropil— suggesting a decrease in growth of neuronal processes and compromised circuitry. In this study, we quantified the extent of dendritic branching in cortical layer II-III pyramidal neurons in adult Igf1 targeted gene deletion mice. Formalin-fixed blocks of frontoparietal cortex of 50 day-old Igf1 (-/-) mice (N=7) and wild-type littermate controls (N=6) were stained using the Rapid Golgi method. From coded slides, five layer II/III pyramids were randomly selected from each subject. Camera lucida drawings were made of the basilar dendritic tree of each neuron and the dendritic arbor was assessed using the Sholl analysis and other quantitative means. Sholl analysis showed that the dendritic arbor of II/III pyramids from the Igf1 (-/-) mice had significantly less dendritic material (p=0.0002). The estimated total dendritic length of the basilar tree was reduced by 24% in the knockout mice. Branch point analysis confirmed that these neurons also had a less complex dendritic arbor. Conclusions: This study conclusively shows significant dendritic atrophy of layer II/III pyramids in 50 day-old Igf1 targeted gene deletion mice. Similar, albeit more moderate, dendritic atrophy was also seen in layer V pyramids of these knockout mice. These results support a critical role for IGF1 in the normal maturation of cortical neurons. (Supported by NICHD)

Presented at the Meeting of the Society for Neuroscience, 2000

STRIKING NEUROTROPHIC EFFECT OF CHRONIC GLYCOSAMINOGLYCAN (GAG) TREATMENT ON THE DENDRITIC ARBOR OF HIPPOCAMPAL CA1 PYRAMIDAL NEURONS IN ADULT F344 RATS: A QUANTITATIVE GOLGI ANALYSIS

R.F. Mervis¹, J. McKean¹, T. Pindell¹, R. Rinehart¹, S. Zats², A. Gum², J.M Lee^,3,4,
M.Hejna³, B. Dudas⁴, and I. Hanin⁴

¹NeuroStructural Research Labs, Columbus, OH 43212;
²Ohio Dominican College, Columbus, OH,
³Dept. of Pathology and ⁴Dept of Pharmacology & Exp. Therapeutics,
Loyola Univ. Chicago Medical Center, Maywood, IL 60153.

In order to to test the efficacy of a novel therapeutic strategy in a animal model of Alzheimer’s disease (AD), neurostructural assessment of the extent of the dendritic arbor was made of CA1 pyramids in adult rats in which amyloid- (25-35) (or vehicle) had been stereotaxically injected directly into the amygdala. This was followed by chronic sub-cutaneous administration of the GAG compound. Previously, amyloid- was foundto result in trans-synaptic cytoskeletal pathology including tau2 immunoreactivity distally in cingulate cortex and hippocampus (Sigurdsson et al, 1996). In this study there were four groups of subjects: (1) Young-adult (3 mon-old) F344 rats (N=8) were given a single intra-amygdala injection of A (25-35) (5nm/3ul) and killed 32 days post-operatively; (2) Vehicle-injected Controls (N=12); (3) A+ GAG-treated Ss (N=6), and (4) the s.c. GAG-treatment alone (N=5). Formalin-fixed tissue blocks were Golgi-stained and CA1 pyramids (6/brain) were randomly selected from coded slides. The basilar dendritic trees were analyzed using Sholl analysis. While intra-amydgaloid A+ chronic GAG treatment (32 days) resulted in a small increase (+6%) in dendritic arbor above control levels, chronic GAG treatment alone resulted in a and highly significant increase (+32%) in the amount of dendritic branching in the hippocampal neurons in the young adult rats. These findings lend support the neurotrophic potential of of GAGs and suggests that it may be useful as a therapeutic strategy in AD. (Supported by NIH/SBIR Grant #1-R41-AG15740-01)

Published in the Journal of Neuroscience (2000)

HIPPOCAMPAL ABNORMALITIES AND ENHANCED EXCITABILITY IN A MURINE MODEL OF HUMAN LISSENCEPHALY

Mark Fleck^1,4, Shinji Hirtsune^2,5, Michael J. Gambello^2,6, Emily Phillips-Tansey¹, Gregory Suares¹,
Ronald F. Mervis³, Anthony Wynshaw-Boris^2,6, and Chris J. McBain¹

¹Laboratory of Cellular & Molecular Neurophysiology, NICHD, NIH, Bethesda, MD 20892; ²Laboratory of Genetic Disease Research, National Human Genome Research Institute, NIH, Bethesda, MD 20892;
³NeuroStructural Research Laboratories, Columbus, OH 43212;
⁴Dept of Pharmacology and Neuroscience, Albany Medical College, Albany, NY 12208;
⁵The Institute for Animal Genetics, Odakura, Fukushima 961, Japan;
⁶Depts of Pediatrics and Medicine, UCSD School of Medicine, La Jolla, CA 92093

Abstract
     Human cortical heterotopia and neuronal migration disorders result in epilepsy; however, the precise mechanisms remain elusive. Here we demonstrate severe neuronal dysplasia and heterotopia throughout the granule cell and pyramidal cell layers of mice containing a heterozygous deletion of Lis1, a mouse model of human 17p13.3-linked lissencephaly. Birth-dating analysis using bromodeoxyuridine revealed that neurons are born at the appropriate time but fail in migration to form a defined cell layer. Heterotopic pyramidal neurons in Lis1+/- mice were stunted and possessed fewer dendritic branches, whereas dentate granule cells were hypertrophic and formed spiney basilar dendrites from which the principal axon emerged. Both somatostatin- and parvalumin-containing-containing inhibitory neurons were heterotopic and displaced into both stratum radiatum and stratum lacunosum-moleculare. Mechanisms of synaptic transmission were severely disrupted, revealing hyperexcitability at Schaffer collateral-CA1 synapses and depression of mossy fiber-CA3 transmission. In addition, the dynamic range of frequency-dependent facilitation of Lis1+/- mossy fiber transmission was less than that of wild type. Consequently, Lis1+/- hippocampi are prone to interictal electrographic seizure activity in an elevated [K+](o) model of epilepsy. In Lis1+/- hippocampus, intense interictal bursting was observed on elevation of extracellular potassium to 6.5mM, a condition that resulted in only minimal bursting in wild type. These anatomical and physiological hippocampal defects may provide a neuronal basis for seizures associated with lissencephaly.

Published in CNS Drug Reviews, 5, suppl 1 (1999)

QUANTITATIVE ASSESSMENT OF NEURONAL DAMAGE IN A TRANSGENIC MURINE MODEL OF ALZHEIMER’S DISEASE

R.F. Mervis¹, D. Campbell¹, T. Pindell¹, J. McKean¹,
I. Lieberburg², D. Schenk², K. Kahn^2, and D. Games².

¹NeuroStructural Research Labs, Columbus, OH 43212
²Elan Pharmaceuticals, South San Francisco, CA 94080

Abstract
     Golgi-stained preparations reveal the dendritic arbor and spines of randomly impregnated neurons in exquisite detail. This permits highly quantitative assessments of dendritic parameters (e.g., branching and spines) which reflect the health of the neurons and the extent of circuitry in that brain region.
     The PD-APP transgenic (tg) mouse overexpresses human amyloid precursor protein (APP717Và F) and is an animal model of Alzheimer’s Disease (AD). These mice develop A-beta plaques in an age-related manner, and show aberrant axonal sprouting. Although increased amyloid deposition has not been demonstrated to result in neuronal loss in these mice, injured neurons should show dendritic atrophy amd related alterations. Using Golgi preparations we characterized dendritic changes in the hippocampus of aged tg PDAPP mice and non-transgenic (non-tg) wild-type controls.
     Coronal blocks of formalin-fixed tissue encompassing the hippocampal formation from 26 month-old PD-APP mice and age-matched controls were stained using the Rapid Golgi method. From coded slides, CA1 pyramids (n=66) and granule cells (n=67) in the dentate gyrus were randomly selected for dendritic branching analysis. Camera lucida drawings were analyzed using the Sholl Method of Concentric Circles. This analysis produces a profile delineating the extent and distribution of the dendritic field. Dendritic spines along selected branch segments were quantified directly from the microscope at 1,008x.

Results:

• CA1 Pyramids – (basilar tree) Branching: The basilar tree of the CA1 pyramids from the PD-APP mice showed significant widespread atrophic changes throughout the entire arbor characterized by both loss of dendritic material and simplification of branching. Spines: There was a 30% loss of total estimated spines on the basilar tree.
• Granule Cells -- Branching: The dendritic arbor of the granule cells showed significant atrophic changes specifically restricted only to the outer 2/3rds of the molecular layer. This portion of the dendritic tree receives afferent input from the entorhinal cortex, an area prominently affected in AD. Spines: There was a 45% loss of total estimated spines on the distal outer 1/3rd of the granule cells in the molecular layer. Middle and inner molecular layer segments showed no spine loss.

Summary:
     Our findings are in general agreement with previous studies of hippocampal dendritic atrophy in AD brains. Moreover, since these hippocampal neurons receive afferent innervation from the entorhinal cortex and this latter region is widely damaged in AD, our results suggest that amyloid deposition and partial deafferentiation of entorhinal input may both contribute to hippocampal damage with concomitant disruption of brain circuitry. These findings support the use of these PD-APP mice which overexpress amyloid as an animal model of AD. It also serves as validation to the efficacy and sensitivity of quantitative Golgi studies for assessing the extent of neuronal damage and, subsequently, for testing the efficacy of various treatment strategies.

Presented at the Society for Neuroscience, Miami Beach, October, 1999

HIPPOCAMPAL DENDRITIC ATROPHY IN AGING PD-APP TRANSGENIC MICE OVEREXPRESSING HUMAN AMYLOID PRECURSOR PROTEIN:
A quantitative Golgi analysis of ca1 pyramids and dentate granule cells.

R.F. Mervis¹, D. Campbell¹, T. Pindell¹, J. McKean¹,
I. Lieberburg², D. Schenk², K. Kahn^2, and D. Games².

¹NeuroStructural Research Labs, Columbus, OH 43212
²Elan Pharmaceuticals, South San Francisco, CA 94080

Abstract
     PD-APP transgenic (tg) mice overexpress human amyloid precursor protein (hAPP) V717F, develop A-beta plaques, and mirror other Alzheimer-like pathology. Here we further characterize neurostructural changes in the hippocampus of these mice. Coronal blocks of formalin-fixed tissue encompassing the hippocampal formation from aged (26 mons-old) PD-APP mice and age-matched controls were stained using the Rapid Golgi method. CA1 pyramids (n=66) and granule cells (n=66) in the dentate gyrus were randomly selected from coded slides. Camera lucida drawings were made of their dendritic arbors and analyzed using the Sholl Method. The basilar tree of the CA1 pyramids from the PD-APP mice showed significant atrophic changes including loss of dendritic material and simplification of branching. By contrast, the dendritic arbor of the granule cells showed atrophic changes only in the outer 2/3rds of the molecular layer, e.g., that dendritic region receiving input from the entorhinal cortex, an area severely affected in AD. These findings support the validity of this animal model of AD and provide an anatomical substrate for the cognitive impairment described in these mice. (Supported by Elan Pharmaceuticals)

Presented at the Society for Neuroscience, Miami Beach, October, 1999

GLYCOSAMINOGLYCANS REVERSE THE DENDRITIC ATROPHY OF CINGULATE PYRAMIDS FOLLOWING INTRA-AMYDGALOID INJECTION OF AMYLOID-b IN ADULT F344 RATS: A Quantitative Golgi Analysis

J. McKean¹, T. Pindell¹, J.M Lee^2,3, M.Hejna², I. Hanin³, and R.F. Mervis¹

¹NeuroStructural Research Labs, Columbus, OH 43212;
²Department of Pathology and ³Department of Pharmacology & Experimental Therapeutics,
Loyola University,Chicago Medical Center, Maywood, IL 60153

Abstract
     A potential animal model of Alzheimer’s disease (AD) was assessed, based on the neurotoxic effects of amyloid-b (25-35) stereotaxically injected directly into the amygdala. Previously, this was found (Sigurdsson et al, 1996) to result in progressive trans-synaptic cytoskeletal and astroglial reactions, including tau2 immunoreactivity distally in cingulate cortex and hippocampus. The present study reveals additional information regarding neurostructural damage of cingulate neurons in this model. Young-adult (3 mon-old) F344 rats (N=8) were given a single intra-amygdala injection of Ab (25-35) (5nm/3l) and killed 32 days post-operatively. Controls (N=7) were vehicle-injected. Formalin-fixed fronto-parietal tissue blocks were Golgi-stained. Layer V cingulate neurons (6/brain) were randomly selected from coded slides. Camera lucida drawings of the basilar tree were analyzed using Sholl analysis. Results revealed small, but consistent and statistically significant dendritic atrophy in Ab injected rats (Wilcoxin test, p<.05). The findings lend additional validity to this animal model of AD and suggest that it may be useful in the evaluation of therapeutic strategies.
(Supported by NIH/SBIR Grant #1-R41-AG15740-01)

Presented at the Meeting of the Society for Neuroscience, Miami Beach, October,1999

HIPPOCAMPAL DISORGANIZATION AND DENDRITIC ATROPHY FOLLOWING ABNORMAL NEURONAL MIGRATION IN LIS1-DEFICIENT MICE:
a quantitative analysis of golgi-stained tissue.

T. Pindell¹, R.F. Mervis¹, J. McKean¹, M.W. Fleck², S. Hirotsune³, M. Gambello³,
A. Wynshaw-Boris³, and C. J. McBain⁴.

¹NeuroStructural Research Labs, Columbus, OH 43212;
²Dept. Pharmacology & Neuroscience, Albany Medical College, Albany, NY 12208;
³Lab of Genetic Disease Research, National Human Genome Research Institute, NIH, Bethesda, MD
⁴Lab. of Cellular & Molecular Physiology, NICHD, NIH, Bethesda, MD 20892

Abstract
     Human Type I lissencephaly is a severe neuronal migration disorder that results from mutation or deletion of the gene encoding the b subunit of platelet-activating factor acetylhydrolase (Lis1). Mice with a heterozygous deletion of Lis1 show abnormal neuronal migration. In the hippocampus of Lis1+/- mice both normally located, and a "stranded" subgroup of CA1 pyramidal neurons are found. We characterized dendritic parameters of both populations of CA1 neurons. Formalin-fixed coronal blocks from mutants (N=4) and wild type (WT) controls (N=7) were Golgi stained. From coded slides, normally located CA1 pyramids (in the str. pyramidale) and stranded CA1s were randomly selected for analysis. Camera lucida drawings were quantified. Dendritic analysis showed that normally located CA1s in both the KO mice and WT controls were not structurally different. However, stranded CA1s were significantly atrophic and had far more simplistic dendritic arbors. Nevertheless, spine density on these stranded CA1 neurons was normal. Such structural abnormalities may explain in part the reduced electrographic seizure threshold observed in Lis1+/- mice. (Supported by NIH.)

Presented at the meeting of the Neurotrauma Society, Miami Beach, October, 1999

NEOCORTICAL NEUROPROTECTION BY BONE MORPHOGENIC PROTEIN-7 FOLLOWING MIDDLE CEREBRAL ARTERY OCCLUSION IN THE RAT:
A Novel and Sensitive Histopathological Assessment using Golgi-Stained Tissue.

Ronald F. Mervis*¹, J.M. McKean¹, T.P. Pindell¹, P.L. Kaplan²J-M Ren³, S.P. Finklestein³

¹NeuroStructural Research Laboratories, Columbus, OH 43212,
²Creative BioMolecules, Inc., Boston, MA 02199,
³Dept. of Neurology, Massachusetts General Hospital and Harvard Medical School, Boston, MA 02114

Abstract
     Bone Morphogenic Protein-7 (BMP-7) a.k.a Osteogenic protein-1 (OP-1) has previously been shown to enhance recovery of function when administered 24 hours after unilateral middle cerebral artery occlusion (MCAO) in the rat. Using the microscopic appearance of Golgi-stained neuropil as a highly sensitive histopathological technique for assessing tissue damage, we now show that BMP-7 reduced the extent of tissue damage surrounding the infarct.
     Adult S-D rats were subjected to permanent right MCAO. Approximately twenty-four and seventy-two hours after infarction the animals were administered BMP-7 (~0.4nmol) or vehicle by injection into the cisterna magna. All subjects were sacrificed 21 days post-MCAO. Brains were immersion-fixed in formalin and 3mm-thick coronal blocks within the infarcted region were removed and stained using the rapid Golgi method.
     Neocortex from the infarcted hemisphere was categorized as: (1) missing; (2) heavily damaged (severely necrotic with no Golgi-impregnated neurons); (3) moderate-to-mildly damaged (with well-impregnated neurons and a grainy neuropil background); and, (4) undamaged (impregnated neurons with a clear neuropil background). From coded slides, neocortical areas ascribed to the different categories were quantified.
     BMP-7 treated rats had less tissue missing (p=0.0012) and more lightly-to-moderately damaged (p=0.0098) and undamaged tissue (p=0.001) than control rats. Since Golgi-stained neurons in the moderately-to-lightly damaged penumbral region had less dendritic material (p=0.002) this population of neurons may be characterized by injured cells which were rescued from subsequent neuronal death by BMP-7.
(Funded by Creative BioMolecules, Inc.)

Presented at the IBC Conference on Alzheimer’s Disease, Boston, MA 1999

AGING PDAPP TRANSGENIC MICE WITH A-beta DEPOSITION SHOW DIFFERENTIAL PATTERNS DENDRITIC ATROPHY OF HIPPOCAMPAL CA1 PYRAMIDS
& DENTATE GRANULE CELLS.
A quantitative Golgi study of an animal model of Alzheimer’s disease

Abstract
     PD-APP transgenic (tg) mice overexpress human amyloid precursor protein (hAPP) V717F, develop A-beta plaques, and mirror other Alzheimer-like pathology. Here we further characterize neurostructural changes in the hippocampus of these mice. Coronal blocks of formalin-fixed tissue encompassing the hippocampal formation from aged (26 mons-old) PD-APP mice and age-matched controls were stained using the Rapid Golgi method. CA1 pyramids (n=66) and granule cells (n=66) in the dentate gyrus were randomly selected from coded slides. Camera lucida drawings were made of their dendritic arbors and analyzed using the Sholl Method. The basilar tree of the CA1 pyramids from the PD-APP mice showed significant atrophic changes including loss of dendritic material and simplification of branching. By contrast, the dendritic arbor of the granule cells showed atrophic changes only in the outer 2/3rds of the molecular layer, e.g., that dendritic region receiving input from the entorhinal cortex, an area severely affected in AD. These findings support the validity of this animal model of AD and provide an anatomical substrate for the cognitive impairment described in these mice. (Supported by Elan Pharmaceuticals)

Presented at The American Association of Veterinary Anatomists
July, 1999

ANALYSIS OF SELECTED GOLGI-STAINED BRAIN REGIONS OF TOTTERING, LEANER AND COMPOUND HETEROZYGOUS, TOTTERING/LEANER MICE.

L. C. Abbott*, J. McKean, T. Pindell and R. F. Mervis.

*Dept. of Veterinary Anatomy & Public Health, Texas A&M University, College Station, TX 77843 NeuroStructural Research Laboratories, 2109 West Fifth Ave., Columbus,OH 43212.

Abstract
     Tottering (tg/tg), leaner (tgla/tgla), and compound heterozygous tottering/leaner (tg/tgla), mutant mice exhibit, to different degrees: ataxia, petit-mal-like epilepsy and an intermittent, myoclonus-like movement disorder. The tottering (tg) and leaner (tgla) mutations occur in the alpha1A calcium channel subunit which is highly expressed by many CNS neurons. This suggests a correlation between calcium channel dysfunction and occurrence of the neurological disorders exhibited by these mice, but the pathogenesis is far from clear. We initiated a morphologic study of the CNS to gain information concerning the influence these mutations have on CNS neurons. We have carried out a quantitative morphometric analysis of tg/tgla cerebellar Purkinje cells, as 90+ % of calcium channels expressed by these cells contain alpha1A subunits. We also examined the cerebral cortex to study neurons in another brain region. Brains from adult mutant and control mice were prepared using the Golgi-Cox method. Decreased branching area and spine density were observed for tg/tgla cerebellar Purkinje cells. However, tg/tgla cerebral layer II/III pyramidal neurons showed significantly larger dendritic arbors compared to controls, but no significant difference in spine density was observed. Cerebral pyramidal neurons may undergo compensatory neuroplastic changes, possibly in response to impaired cerebellar circuitry. Supported by a NIH grant (K08NS01681) to L.C.A.

Published in Brain Research, 1998

DENDRITIC ALTERATIONS IN CORTICAL PYRAMIDAL CELLS
IN THE SPARCE FUR MOUSE

K.J. Hopkins¹, J. McKean ², R.F. Mervis ², M-L. Oster-Granite^3
1Dept of Environmental Toxicology, University of California, Riverside, CA 92521;
²NeuroStructural Research Labs, Columbus, OH 43212;
³Division of Biomedical Sciences, University of California, Riverside, CA 92521

Abstract
     Ornithine carbamoyltransferase deficiency, an X-linked trait, leads to toxic hyperammonemia in sparce fur (spf/Y) mice. Quantitative analysis of the basilar dendritic tree of layer V pyramidal cell in frontoparietal cortex stained by the Golgi Kopsch method revealed a significant decrease in both the complexity of the dendritic arbor and in dendritic terminal spine density (60%) in spy/Y mice compared with controls. Such reductions may contribute to behavioral dysfunction observed in spy/Y mice.

Presented at the Annual Meeting of the Neurotrauma Society, Los Angeles, November, 1998

QUANTITATIVE ASSESSMENT OF CORTICAL DENDRITIC ALTERATIONS FOLLOWING TRAUMATIC BRAIN INJURY IN THE RAT: A GOLGI-IMPREGNATION STUDY.

Ronald F. Mervis*^1, Timothy Pindell^1, Jody McKean^1, Joseph S. Soblosky², Michael E. Carey²

¹NeuroStructural Research Laboratories, 2109 West Fifth Avenue, Columbus, OH 43212 USA, ²Department of Neurosurgery, LSU Medical Center, New Orleans, LA 70112 USA.

Abstract
     Traumatic brain injury (TBI) was produced by a controlled pneumatic impactor striking the entire right sensorimotor cortex of the anesthetized rat. Neurologic testing reveals initial deficits with a biphasic pattern of functional behavioral recovery. Golgi-impregnation staining – which permits detailed microscopic assessment of degenerative and regenerative dendritic changes – was used to quantitatively evaluate the basilar trees in superficial (layers II-IV) and deep (layers V-VI) pyramidal cell populations in the impact zone ("damaged region") and adjacent cortical ("undamaged") region. Animals were sacrificed at 6 days (n=3) , 3 months (n=3), and 4.5 months (n=1) after the TBI. There were 5 sham controls. A total of 219 neurons were evaluated. Slides were coded and assessments made blind with respect to the post-trauma times. Parameters quantified included Sholl analyses, measurement of total dendritic lengths, and changes in branch segments. The most salient findings were: Six days following TBI, neurons in both damaged and undamaged cortex revealed widespread dendritic atrophy. Deep pyramidal neurons appeared to be more affected than superficial pyramids. Indices of damage included loss of terminal branches, reduction in dendritic length, and significant reductions in Sholl dendritic arbor profiles. At 3 months post-injury, most dendritic parameters still were suppressed in both damaged and undamaged regions. However, there was an increase in terminal tip segments, possibly predicting new neuronal growth. By 4.5 months we find preliminary but compelling evidence of neuroplasticity in both damaged and undamaged cortex, the undamaged region showing the more robust effect. Golgi-staining thus appears to be a valuable tool to help ascertain the extent of neuronal damage and dendritic plasticity and could be useful in evaluating the efficacy of neuroprotective agents . (Supported by U.S. Army contract DAMD-17-93-C-3008)

Presented at the Annual Meeting of the Society for Neuroscience, Los Angeles, November, 1998

COMPENSATORY CORTICAL NEUROPLASTICITY IN THE COMPOUND, HETEROZYGOUS, TOTTERER-LEANER MUTANT MOUSE.

Ronald F. Mervis*, Timothy Pindell*, Jody McKean*, and Louise C. Abbott¹

*NeuroStructural Research Laboratories, Columbus, OH 43212,
¹Dept. Veterinary Anatomy & Public Health, Texas A&M University, College Station, TX 77843

Abstract
     Compound heterozygous, tottering/leaner mice (tg/tg^la) exhibit ataxia, intermittent myoclonus and absence seizures. These mice carry one allele each of two autosomal recessive mutations: tottering (tg) and leaner (tg^la). Golgi impregnation studies of the cerebellum showed Purkinje cells (PC) from adult tg/tg^la mice have reduced branching area and fewer dendritic spines (Abbott et al, Soc. Neurosci. Abst., 1998). Broad interaction between cortex and cerebellum with respect to motor and memory functions infers that reduced cerebellar circuitry in these mutant mice may influence cortical circuitry. Blocks of fronto-parietal cortex from tg/tg^la (N=5) and control (+/+) (N=4) mice were stained using the Golgi-Cox method. Sections were cut at 120um. From coded slides, five layer II/III pyramidal neurons were randomly selected from each subject. Camera lucida drawings were made of the basilar tree of each neuron. Sholl analysis showed that the dendritic arbors from tg/tg^la mutant mice were significantly larger than controls (Wilcoxin test, p=0.001). There was no significant difference in the spine density along 30um terminal tip segments between tg/tg^la mutants and controls. Overall, the increased amount of dendritic branching in the mutant suggests that cortical synaptic circuitry is enhanced in this group. Previously, it was found that Lurcher mutants (with postnatal loss of PCs) showed enhanced dendritic arbor of cortical pyramids (Goss et al, Soc. Neurosci. Abst. 22:1132, 1996). The tg/tg^la mutant mouse reveals an apparently comparable compensatory neuroplastic response of cortical neurons -- possibly to subserve functions associated with impaired cerebellar circuitry. [Supported by a NIH grant (K08NS01681) to L.C.A.]

Published in the Journal of Neurochemistry, 69, 1997

ENHANCED HEMICHOLINIUM BINDING AND ATTENUATED DENDRITE BRANCHING IN COGNITIVELY IMPAIRED ACETYLCHOLINESTERASE-TRANSGENIC MICE

R. Beeri ⁽¹⁾, N. LeNovere ⁽²⁾, R.F. Mervis ⁽³⁾, T. Huberman ⁽¹⁾, E. Grauer ⁽⁴⁾, J.P. Changeux ⁽²⁾,
and M. Soreq ⁽¹⁾

⁽¹⁾Department of Biological Chemistry, Hebrew University of Jerusalem, Jerusalem, Israel;
⁽²⁾ Molecular Neurobiology Laboratory, Biotechnologies Department, Pasteur Institute, Paris, France
⁽³⁾NeuroStructural Research Laboratories, Columbus, Ohio 43212 USA
⁽⁴⁾Department of Pharmacology, Israel Institute for Biological Research, Ness-Ziona, Israel

Abstract
     In a search for behavioral, neuroanatomical, and metabolic characteristics of Alzheimer’s disease that may result from cholinergic malfunction, we used transgenic mice overexpressing acetylcholinesterase (AChE) mRNA and active enzyme in brain neurons. Mapping by in situ hybridization revealed that transgenic and host AChE mRNAs were distributed similarly. In a Morris water maze working memory paradigm, adult transgenic mice did not display the characteristic improvement found control mice either between or within test days and spent less time than control mice in the platform zone. In 5-week-old transgenic mice, the basilar dendritic trees of layer V pyramidal neurons from the frontoparietal cortex were essentially as developed as in age-matched controls. However, branching totally ceased after this age, whereas in control adults it continued up to at least 7 months. Therefore, dendritic arbors became smaller in adult transgenic mice than those of controls. Furthermore, the average number of spines was significantly lower on dendritic branches of 7-month-old but not 5 week-old transgenics compared to controls. Binding of tritiated hemicholinium-3, a blocker of the high affinity uptake characteristic of active cholinergic terminals, was over two-fold enhanced in the brain of transgenic mice. In contrast, no differences were observed in the mRNA and ligand binding levels of several different subtypes of nicotinic and muscarinic acetylcholine receptors. These findings suggest that three different hallmarks associated with Alzheimer’s disease – namely, progressive cognitive failure, cessation of dendrite branching and spine formation, and enhanced high-affinity choline uptake – are outcomes of cholinergic malfunction.

Presented at the Annual Meeting of the Society for Neuroscience, New Orleans, October, 1997

ALTERED NEURONAL MORPHOLOGY AND NEURONAL CELL DENSITY IN HIV INFECTED PRIMATES

M. Piralta, M. Ruiz, F.J. Denaro, Texas Tech University Health Sciences Center, Lubbock, TX
R.C. Switzer III, NeuroScience Associates, Knoxville, TN
A. Lackner, J. Lane, New England Regional Primate Center, Harvard Medical School
R. F. Mervis, NeuroStructural Research Laboratories, Columbus, Ohio

Abstract
     Identification of altered neuronal morphology has been reported in altered postmortem CNS tissue from patients with AIDS. In experimental models it has been found that HIV peptides can also cause alteration in the neuronal morphology in rodents. Such alteration in neuronal morphology can have important implications for understanding both the neurologic and cognitive symptoms. In the present study, Golgi impregnated tissue from SIV infected primates was analyzed. Changes in neuronal morphology were identified which include: atrophic neurons, neurons with decreased spines and, in some instances, varicosities. In addition, special Nissl sections (75um thick) were prepared for analysis with a computer-assisted optical dissector. Regional neuronal counts, adjacent to multi-nucleated giant cells did not demonstrate neuronal loss. But areas in which gliosis can be identified (with or without multi-nucleated giant cells) presented with neuronal cell loss. Studies are underway to characterize the spectrum of morphologic changes. In addition, our work with cell counts is continuing. Further data may reveal a pattern associated with neuronal cell loss.
(Supported by NIH grants NS31857, RR00168, NS30769, NS35732)