Search

The birth, growth, proliferation and death of cells involve a rigorous and continuous process in place to ensure the survival of living organisms. The cell cycle prevails at the core of this process to facilitate the division of a parent cell as well as the duplication of its genetic matter. Although checkpoints exist to steer the course of a cell from one phase to the other, malfunctions at any point of the four active phases of the cell cycle can have detrimental effects. Cancer is thought to be a consequence of such a malfunction in the cell cycle which endows a cell with enhanced replicative potential, immunity to anti-growth signals and the ability to evade apoptosis. This characteristic has been exploited in cancer chemotherapy since a significant number of anticancer drugs manifest their action via cell cycle modulatory effects. The plant family Amaryllidaceae is distinguished for its alkaloid principles which exhibit potent (at the sub-nanomolar level in some cases) and cell line specific antiproliferative activities, with apoptosis induction a key feature of these properties. As a consequence there has been sustained interest in these chemical entities as a source of new anticancer drugs. This has been matched by the large body of work that has emerged over the past two decades addressing their cytotoxic potential, establishing a structure-activity relationship basis as well as probing their mode of action. This review focuses on studies which highlight how Amaryllidaceae alkaloids modulate the cell cycle of cancer cells.

Cells cultured from Alzheimer disease leptomeninges or skin were grafted into the cortex of adult thymectomized rats. At 3 days post-implant, plaque-like aggregates were found in the cortex, corpus callosum, septum and caudate nucleus. These structures were immunopositive for human amyloid precursor protein (APP), human amyloid beta peptide (Aβ), cathepsin D, apolipoprotein E and ubiquitin. Aberrant tau + neurites, reactive astrocytes and microglia were associated with many aggregates. Although birefringent amyloid occupied the central area of most aggregates, these structures had disappeared by l month post-implant. Aβ and APP produced by grafted non-neural human cells can penetrate rat brain and form plaque-like structures, which can be effectively cleared by the rat.

Cell membranes contain numerous species of glycosphingolipids (GSL) and glycerophospholipids. Changes in the surface expression of these moieties that accompany malignant transformation are potential targets for immune therapies. There is increasing evidence that tumour specificity resides in the carbohydrate composition and topography of gangliosides and in the asymmetric distribution of phospholipids in the bilayer membrane. However, since these epitopes are not antigenic, circumvention of immunological tolerance to these ‘self antigens’ is required to generate an effective immune response. New methodologies accomplish this by coupling these hapten-like lipid epitopes to immunogenic protein carriers. Indeed, the generation of immune responses to endogenous lipid antigens occurs in several autoimmune syndromes and may serve as natural models for therapeutic immune intervention that influence pathways critical to tumour angiogenesis, proliferation and apoptosis.

We determined whether the angiogenesis and growth of murine colon carcinomas growing in the wall of the cecum is dependent on infiltrating leukocytes. Syngeneic BALB/c or SCID mice were treated with a myelosuppressive, maximally tolerated dose of doxorubicin. Parental or multidrug resistant CT-26 colon carcinoma cells were implanted into the cecal wall 3 days after the second intravenous injection of doxorubicin. Control mice developed large, well-vascularized tumors, whereas doxorubicin-pretreated mice did not. Intravenous injection of spleen cells from normal BALB/c or SCID mice one day prior to tumor cell implantation reversed the decreased vascularity and tumorigenicity. The production of proangiogenic molecules and microvessel density in tumors directly correlated with the number of infiltrating leukocytes, suggesting that tumor-infiltrating leukocytes are essential to angiogenesis of murine colon carcinomas.

Although humoral responses to Borrelia burgdorferi (Bb) have been shown to be protective in some animal models of Lyme disease, the role of T cells in this disease is less well understood. This work describes three Bb-specific T cell lines that prevent disease progression in syngeneic mice. The T cell lines were generated in C3H mice immunized with Bb in complete Freund's adjuvant. All lines were Bb-specific, CD4+, TCRαβ+, and they proliferated and produced interferon-γ and interleukin-2 on stimulation with Bb. Injection of the cell lines into naive C3H recipients significantly reduced the number of organisms recoverable from the blood and tissues of infected mice and protected them from developing Bb-induced periarthritis. These studies demonstrated that Th1 cells can confer resistance to Bb infection in susceptible mice and suggested that the timing of this T cell response may be critical for determining disease outcome.

The following series of experiments explores the post-xenografting differentiation of a naturally occurring, non-neuronal cell cultured from the leptomeninges of an 84-year-old woman. In culture, flat process-bearing human cells from the leptomeninges were positive for GFAP and 200 kDa neurofilament protein (negative for 68, 160 kDa neurofilament protein). The C3 spinal cord was exposed in 30 adult athymic rats. The hindlimb dorsal columns were transected at C3 and the nerve fibers aspirated to form a pocket, into which 10(6) fast blue-labeled, human leptomeningeal-derived cells were placed. The C3 spinal cord was studied immunohistochemically over 60 days. Three days later the dorsal horn contained fast blue-GFAP-positive astrocyte-like cells that were negative for neurofilament protein. By 7 days, large, process-bearing, fast blue-GFAP-positive (neurofilament protein-negative), astrocyte-like cells joined the native astrocytes of the pia-glia membrane and were in the gray matter of the spinal cord. Some of these astrocyte-like cells were also positive for the human specific histocompatibility complex, HLADR. These data extend the age, species and tissue of origin for pluripotential cells for CNS transplantation.

Astrocytes appear to be excluded from blood vessels. In order to test this observation, pial blood vessels were seeded upon cultures of purified E14 fetal astrocytes. One to seven days later co-cultures were immunohistochemically stained for laminin and fibronectin. Confocal microscopy revealed that laminin-positive cultured astrocytes migrated from the coverslip to and on to the surface of the blood vessels. The fetal astrocytes migrated along the basement membrane but did not enter the blood vessel. Viable (live) basement membrane is a barrier to astrocyte intravasation and is a pathway for astrocyte migration. Published by Elsevier Science Ltd.

Purpose: These studies were designed to develop an orthotopic model for human bladder cancer and to isolate variant metastatic cell lines.Materials and Methods: The human bladder cancer cell line 253J was implanted into the muscular wall of the bladder of athymic nude mice. By in vivo recycling, we selected for 2 variant cell lines: 253J B-V, a bladder line isolated after 5 serial passages in the bladder, and 253J lung-IV, established from a lung tumor nodule that was recycled through the bladder.Results: These 2 cell lines showed enhanced tumorigenicity, as measured by a decreased latent period, and rapid growth as compared with the parental cell line. Moreover, orthotopic implantation of these cell lines resulted in metastasis to the lungs. These in vivo-selected, metastatic cell lines exhibited unique karyotypic alterations, increased anchorage-independent growth, overexpression of basic fibroblast growth factor, altered expression of adhesion molecules and the ability to migrate through Matrig...

Glioblastoma is very rarely found outside the central nervous system. The ability of rat C6 glioblastoma cells to intravasate into central nervous system and pial blood vessels is tested using a rat homografting model and two in vitro models. In vivo, scanning electron microscopy demonstrates that upon grafting C6 cells into implantation pockets in rat cortex, blood vessels can be spared in large digestion cysts formed in host brain parenchyma. Immunocytochemistry of the grafted rat cortex reveals that the glioblastoma cells are upon the blood vessel basement membrane, surrounded by the extracellular matrix material, fibronectin. The endothelial cells of the blood vessel are inside the laminin and fibronectin, and there were areas of endothelial cell hyperplasia. C6 cells are not observed inside blood vessels. In vitro, C6 cell cultures seeded with blood vessels from fresh rat pia exhibit the same relationship of the C6 glioblastoma cells to the blood vessel as those in the other models. The C6 cells migrate upon the pial blood vessel basement membrane but do not intravasate into the blood vessel. To ascertain whether structure and components of the blood vessel basement membrane are important factors in glioblastoma cell exclusion from blood vessels, C6 cells are seeded upon artificial basement membrane hydrated gel wafers. C6 cells migrate into the artificial basement membrane gel wafer by 1 day after seeding. These data indicate that glioblastoma cells are confined to the central nervous system by an inability to pass through vital basement membrane.

We determined whether the intravenous administration of multilamellar vesicle liposomes (MLV) containing a lipopeptide analogue of a fragment from the cell wall of gram-negative bacteria (CGP 31 362) can render BALB/c mouse alveolar macrophages tumoricidal in situ and reduce the incidence of spontaneous lung metastasis of syngeneic renal carcinoma (RENCA) cells. Alveolar macrophages (a) incubated in vitro with MLV containing CGP 31 362 (MLV-31 362) and (b) harvested from mice injected i.v. with MLV-31 362 were rendered cytotoxic against the RENCA cells. Maximum cytotoxic activity of the macrophages was induced by injecting 5 µmol MLV consisting of 250 mg phospholipids and 0.5 mg CGP 31 362. The single i.v. injection of 5 µmol MLV-31 362 produced activation of macrophages that lasted for up to 4 days. Repeated i. v. injections of MLV-31 362 produced a continuous antitumor activity in alveolar macrophages. To study the lipopeptide's effects on metastasis, we injected the left kidneys of BALB/c mice with RENCA cells. The kidney with growing tumor was resected 10 days later and, after a further 2 days, groups of mice were injected i.v. with MLV-31 362 or with MLV-HBSS (twice weekly for 3 weeks). Treatment with MLV-31 362 significantly decreased the median number of spontaneous lung metastases. These data demonstrate that the systemic administration of MLV-31 362 can activate murine lung macrophages in situ and reduce the incidence of spontaneous RENCA lung metastases.

Cultured C6 glioma cells were prelabeled with the plant lectin Phaseolus vulgaris leuco-agglutinin (PHAL) and grafted as a cell suspension (10(6) cells in 5.0 microliters) into freshly made cortical implantation pockets in adult host rats. Animals were killed 1-21 days post-implantation (DPI). The brains were removed, dehydrated, embedded in paraffin and sectioned at 8 microns. Paraffin sections were processed for light level immunofluorescent double labeling for PHAL, a marker for graft derived cells, and glial fibrillary acidic protein (GFAP), a specific marker for C6 glioma cells and astrocytes. Cells positive for both PHAL and GFAP were graft-derived C6 cells. By 7 DPI a large mass developed which extended above the surface of the brain and invaded (displacement of host tissue by a cell mass) the host parenchyma. This mass increased in size over the next 14 days. The invading tumor mass contained double labeled cells at all time periods examined. In addition to the invasion process, grafted C6 cells spread through the host parenchyma by migration (movement of single cells). Individual graft-derived C6 (GFAP/PHAL positive) cells migrated into host cortex surrounding the implantation pocket, corpus callosum ventral to the implantation pocket, ipsilateral internal capsule and bilaterally in the habenula.

In a 'double blind' study, 2 series of adult rats were trained to traverse a narrow bar or a horizontal ladder for a water reward. Hindlimb placement (measured as hindlimb foot slips) of the subjects trained on the narrow platform was ranked. The subjects traversing the ladder were videotaped and the number of hindlimb slips counted. After reaching criterion (10 complete traverses on each of 2 consecutive days), all animals had the fasciculus gracilis (FG) of the third cervical spinal cord segment (C3) aspirated to sever hindlimb dorsal column afferents. After aspiration of C3FG the subjects were randomly placed in an aspiration-only (controls) or aspiration + graft group. All grafts were prelabeled with the plant lectin Phaseolus vulgaris leucoagglutinin (PHAL) as a graft-derived cell marker. The grafts were either unoriented whole pieces of E14 fetal spinal cord or 105 purified, cultured, E14 fetal spinal cord astrocytes. Subjects were tested at intervals over 90 days. Aspiration of C3FG and a graft of whole pieces of E14 fetal spinal cord resulted in a statistically significant improvement in hindlimb placement at 21 and 90 days (P < 0.05) when compared to controls. In contrast, animals with cultured, purified, E14 fetal spinal cord astrocyte grafts had significantly worse (P < 0.05) hindlimb placement than controls. Immunohistochemical double staining for PHAL and glial fibrillary acidic protein in the same cell showed that astrocytes from both types of grafts migrated to the nucleus gracilis (NG) of the medulla of the host. The grafted fetal astrocytes from the whole piece grafts prevented denervation atrophy of the host NG neurons, whereas cultured grafted fetal astrocytes did not demonstrate this effect. After grafting donor tissue that contained astrocytes into the injured spinal cord there were two accompanying classes of astrocytes. One class derived from whole piece, E14, fetal spinal cord grafts migrated to the NG of the host, prevented atrophy of host NG cluster and interneurons, and improved hindlimb placement. The other class, derived from cultures of E14 fetal spinal cord astrocytes, migrated to the NG of the host, failed to maintain the size of cluster neurons and interneurons of the host NG, and resulted in a greater hindlimb placement deficit when compared to controls. These data suggest that improvement of hindlimb placement due to graft-derived fetal astrocytes in injured host spinal cord was due to the ability of the astrocytes to maintain host neurons and neuronal networks.

Current therapy of human superficial bladder cancer includes the intravesical administration of antitumor drugs and immunomodulators. The purpose of these studies was to determine whether phospholipid liposomes that bind to human bladder cancer cells can improve the delivery of interferon alpha (IFN-alpha) to neoplastic urothelium. The antiproliferative activity of free IFN-alpha and IFN-alpha encapsulated in liposomes was assessed in vitro against the human transitional cell carcinoma line 253J. The cells were exposed to free and liposome-encapsulated IFN-alpha for short periods ranging from 30 minutes to four hours, and inhibition of cell growth was determined three days later. The production of greater than 25 percent cytostasis of 253J cells by free IFN-alpha required four hours of continuous exposure. In contrast, IFN-alpha encapsulated in liposomes produced 35 percent and 60 percent cytostasis after a 30-minute and four-hour exposure, respectively. Liposome-encapsulated IFN-alpha was also effective (50 percent cytostasis) against a subline of 253J cells selected for resistance against free IFN-alpha. Liposomes containing IFN-alpha were stable in the presence of human urine. In vivo studies in mice showed that intravesical administration of radiolabeled IFN-alpha or radiolabeled liposomes did not yield significant systemic absorption and deposition in distant organs. Collectively, these results suggest that the encapsulation of IFN-alpha within multilamellar liposomes may augment its antiproliferative activity, overcome some forms of tumor cell resistance to IFN-alpha, and prove useful for intravesical therapy of superficial bladder cancer.

Human leptomeningeal (arachnoid and pia mater) cells in culture have been demonstrated in replicated studies to express typical neuronal proteins such as neurofilament protein and neuron-specific enolase. In addition, they can express glial fibrillary acidic protein. The present study examines the possibility that neuronal and glial proteins might be present in rat and human leptomeningeal cells in situ. The neuronal proteins 160 kDa and 200 kDa neurofilaments, neuron-specific enolase and microtubule-associated protein 2 were, however, not immunolocalized in either the pia mater or arachnoid. Glial fibrillary acidic protein and galactocerebroside were also not detected, while fibronectin and vimentin immunoreactivities were robust in all layers of the leptomeninges. Together with the previously reported expression of some neuronal and astroglial markers in cultured human leptomeninges, these observations suggest that culture alters the properties of leptomeningeal cells.

Purpose: We sought to determine whether blockade of platelet-derived growth factor receptor (PDGF-R) activation by oral administration of a PDGF-R tyrosine kinase inhibitor (STI571) alone or in combination with i.p. paclitaxel can inhibit the progression of tumors caused by human ovarian carcinoma cells growing in the peritoneal cavity of female nude mice. Experimental Design: In several different experiments, paclitaxel-sensitive and paclitaxel-resistant metastatic human ovarian carcinoma cells were injected into the peritoneal cavity of nude mice. Seven days later, groups (n = 10) of mice began receiving a control treatment, STI571 alone, paclitaxel alone, or a combination of STI571 and paclitaxel. The mice were necropsied after 45 days of treatment. Results: Treatment with combination therapy significantly reduced tumor weight (relative to control or single-agent therapy) in all three human ovarian cancer cell lines. Immunohistochemical analyses revealed that PDGF-R activation was blocked by STI571 administered alone or in combination with paclitaxel. Tumor-associated endothelial cells expressed both PDGF-R and phosphorylated PDGF-R. In mice receiving combination therapy, tumor-associated endothelial cells underwent apoptosis, leading to decreases in microvessel density and tumor cell proliferation relative to control and single-agent therapy. Conclusions: These results show that administration of a PDGF-R tyrosine kinase inhibitor in combination with paclitaxel impairs the progression of ovarian cancer in the peritoneal cavity of nude mice, in part, by blockade of PDGF, an endothelial cell survival factor, which results in the increased apoptosis of tumor-associated endothelial cells.

Objectives . We determined the expression of platelet-derived growth factor (PDGF), PDGF-receptor (PDGF-R), and phosphorylated PDGF-R (p-PDGF-R) on tumor cells and tumor-associated endothelial cells in clinical specimens of human ovarian carcinoma and human ovarian cancer cells growing in culture and in the peritoneal cavity of nude mice. Methods . Ten specimens of high-grade serous ovarian carcinoma were analyzed using immunohistochemistry (IHC). IHC was used to detect ligand and receptor expression in the human ovarian cancer cells from Hey A8 and SKOV3ip1 growing in culture. Cells from these lines were also implanted orthotopically into the peritoneal cavity of nude mice. IHC was used to determine ligand and receptor expression in tumors that formed in the peritoneal cavity. Results . All 10 evaluable samples expressed both PDGF AA and BB on tumor cells. Tumor cells were positive for PDGF-Rα in 10/10 samples, PDGF-Rβ in 8/10 samples, p-PDGF-Rα in 6/10 samples, and p-PDGF-Rβ in 4/10 samples. p-PDGF-Rα was positive in 4/10 tumor-associated endothelial cell samples and p-PDGF-Rβ was positive in 3/10 samples. Human ovarian cancer cells expressed PDGF, PDGF-R, and p-PDGF-R when growing in culture or in the peritoneal cavity of nude mice. PDGF-R and p-PDGF-R were also present on tumor-associated endothelial cells as demonstrated by simultaneous staining with CD31 antibody. Conclusions . PDGF and the corresponding receptors were expressed in autochthonous human ovarian cancer lesions on both tumor cells and tumor-associated endothelial cells. The ligand and receptor were also present on Hey A8 and SKOV3ip1 human ovarian cancer cells growing in vitro and in the peritoneal cavity of nude mice.

Since vascular endothelial growth factor (VEGF) plays a major role in tumor angiogenesis, we determined whether blockage of VEGF receptor signaling using a novel tyrosine kinase inhibitor (PTK 787) decreases the growth and metastasis of human pancreatic carcinoma growing orthotopically in nude mice. Human pancreatic L3.6pl cells were injected into the pancreas of nude mice. Seven days later, groups of mice were given daily oral administrations of PTK 787 alone, twice weekly i.p. injections of gemcitabine, or combination therapy. The mice were necropsied when control mice became moribund (day 35). Therapy with PTK 787 alone, gemcitabine alone, or the combination of both agents produced respectively 60%, 70%, and 81% inhibition in the volume of pancreatic cancers. The combination therapy significantly decreased the incidence of lymph node and liver metastasis, leading to a significant increase in survival. Microvessel density (MVD) was significantly decreased in tumors treated with either PTK 787 alone or PTK 787 plus gemcitabine. MVD directly correlated with tumor cell proliferation and inversely correlated with apoptosis of tumor cells and associated endothelial cells. Collectively, our results demonstrate that blockade of VEGF-R signaling may provide an additional approach to the therapy of pancreatic cancer.

Once a diagnosis of cancer is established, the urgent question is whether it is localized or has already spread to regional lymph nodes and visceral organs. Despite improvements in early diagnosis, surgical techniques, general patient care and local and systemic adjuvant therapies, most deaths of cancer patients result from the relentless growth of metastatic disease that is resistant to conventional therapies. Surgical excision of primary neoplasms is not curative in many patients because by the time of diagnosis, metastasis may well have occurred. The major challenge for treatment of metastasis is the biological heterogeneity of cancer cells within primary lesions and especially within metastases. A wide range of genetic, biochemical, immunological and biological characteristics that differ from subpopulation to subpopulation include the display of cell-surface receptors, enzymes, karyotypes, cell morphologies, growth properties, sensitivities to various therapeutic agents and the ability to invade and produce metastasis [1,2]. This heterogeneity is due to differences in etiology, origin and selection pressures. The outcome of the metastatic process depends on the interaction of metastatic cells with host homeostatic factors, which include various elements of the host immune system. The establishment and progressive growth of metastatic cells requires that they exploit, subvert and suppress, when necessary, natural host defense systems. This chapter will focus on the host-tumor interactions that result in immune modulation of tumor cell populations.

Astrocytoma vasculature patterns differ according to histological grade of malignancy with glioblastoma multiforme (WHO grade IV) showing most extensive endothelial proliferation. Here, we determined whether the vascular patterns of medulloblastoma and supratentorial primitive neuroectodermal tumors (PNETs), which can be hardly distinguished histopathologically, differ. We evaluated the spatial organization of vessels in medulloblastomas and PNETs using antibodies to von Willebrand factor (vWF) and CD34. Medulloblastoma capillaries showed slight endothelial cell hyperplasia. Microvessels sprouted from the capillaries and formed glomeruloid clusters. There were areas with chains of unopposed endothelial cells (3-10 cells). Supratentorial PNETs had highly branched capillaries with extensive endothelial cell hyperplasia. Glomeruloid arrays of microvessels extended from the capillaries. Small fragments of endothelial tubes were scattered throughout the tumor. Therefore, medulloblastomas and supratentorial PNETs showed different spatial organization of tumor vessels which can be used for differentiation of each tumor entity. These vascular patterns may reflect different tumor derived angiogenic stimuli.

We investigated the expression of cytokine genes and tumoricidal properties in human blood monocytes in response to a new synthetic immunomodulating lipopeptide, JBT3002. Incubation of peripheral blood monocytes with free-form JBT3002 or JBT3002 encapsulated in multilamellar phospholipid vesicles (liposomes, MLV-JBT3002) induced tumoricidal properties in a dose-dependent manner. Both MLV-JBT3002 and free-form JBT3002 induced production of tumor necrosis factor α, interleukin-1β, and interleukin-6 in a dose-dependent manner with similar kinetics. Treatment of monocytes with interferon-γ did not significantly alter the expression of cytokine genes but increased the expression of cytokines induced by MLV-JBT3002 and free-form JBT3002. In contrast to monocyte activation by lipopolysaccharide (LPS), activation by JBT3002 was independent of serum and was not inhibited by CD14-neutralizing antibody. Incubation of monocytes with JBT3002 induced a rapid increase in tyrosine phosphorylation of proteins with apparent molecular masses of 42 and 38 kDa, a migration band shift of c-Jun NH2-terminal kinase 1 (JNK1), and activation of extracellular signaling regulated kinases. Consistent with its effect on cytokine expression, stimulation of these intracellular signaling pathways by JBT3002 was not inhibited in serum-free conditions. Collectively, the data indicate that the synthetic lipopeptide JBT3002 is a potent monocyte activator that modulates monocyte function by mechanisms similar to LPS but by a distinct receptor.

Cultured human leptomeningeal fibroblasts grafted into rat frontal cortex were localized to the implant pocket and to adjacent host leptomeninges. Immunohistochemical studies using a panel of human-specific and domain-specific APP antibodies revealed that all grafted cells expressed both APP and A β in situ. Remarkably, these antibodies also labeled rat pial and ependymal cells as well as reactive astrocytes adjacent to vessels. In addition, apical projections and cell bodies of many cortical pyramidal neurons contained human-specific APP immunoreactive material. Groups of subcortical neurons, particularly those of the amygdala, hippocampal formation and suprachiasmatic nuclei, were similarly labeled. The presence of human APP in host brains was confirmed by immunoblotting. Birefringent Congo Red staining was observed in the cortical neuropil and in leptomeningeal vessels. These data indicate that grafted leptomeningeal fibroblasts hyperexpress APP and A β which can diffuse into parenchyma and be taken up by specific rat cells.

Transection of the hindlimb dorsal column fibers leads to a partial deafferentation of the neurons of the nucleus gracilis, the first relay of the ascending dorsal column-medial lemniscal (DC-ML) neural network. In response to the deafferentation, a synaptic renewal cycle is initiated and the neurons of the nucleus gracilis atrophy. The present study examines the molecular changes that occur in synaptic relays of the ascending DC-ML following a hindlimb dorsal column transection. Rats were sacrificed 0.5, 1, 24, or 72 hours postlesion. Steady-state levels of mRNA coding for c-fos were significantly elevated only at 24 hours postlesion in caudal dorsal medulla, which contains the nucleus gracilis. The increased c-fos is neuronal in origin since there is an increased level of c-fos immunoreactivity in both the cluster neurons and the interneurons of the nucleus gracilis. In the third relay of the DC-ML system, the somatomotor cortex, levels of c-fos mRNA were significantly decreased 72 hours postlesion. These data indicate that lesions of the hindlimb dorsal column fibers have transneuronal effects on gene expression that extend to at least the third synaptic relay in the DC-ML system.

Individual astrocytoma cells expressing a cytoplasmic form of pl85 c-neu migrated along basement membrane lined surfaces after xenografing fresh low or high grade human malignant astrocytomas into host rat brain. We now study the migratory capacity of fresh human malignant astrocytoma cells seeded on hydrated gel wafers composed of artificial basement membrane or collagen I, a normal and lesion-related CNS extracellular matrix component. Approximately 107 mechanically disrupted cells (with small clumps) of 3 fresh low grade and 6 fresh high grade astrocytomas were seeded on the surface of artificial basement membrane and collagen I wafers (11 × 16 mm). The wafers were then prepared for scanning electron microscopy and immunohistochemstry at 1, 3, 5, and 7 days after seeding. Regardless of tumor grade, a morphologically similar class of cells was observed to migrate through collagen I gels in 24 hours and 0.5–1.5 mm into artificial basement membrane gels in 7 days. Immunohistochemistry revealed that the migrated cells from low and high grade astrocytomas were positive for glial fibrillary acidic protein (GFAP)) and expressed cytoplasmic human-specific pl85 c-neu . These data indicate that fresh human malignant astrocytoma cells that contain GFAP and express cytoplasmic pl85 c-neu have a high degree of migratory capacity and could be the cell in the tumor involved in intraparenchymal metastasis and poor patient survival in high grade astrocytomas of the human brain.

Fresh suspensions of human glioblastoma multiforme were preincubated in the plant lectin Phaseolus vulgaris leucoagglutinin (PHAL) and implanted into cortical pockets in adult rat brain. Brains were investigated periodically over 30 postoperative days and the migration of the human glioblastoma cells was traced with anti-PHAL immunofluorescence or the overexpression of human specific p185c-neu a specific marker of a class of human malignant astrocytoma cells. The principal pathway of migration of the implanted human cells in the rat brain was ventrally through cortical gray matter and into the corpus callosum, with rapid lateral distribution in this and other parallel and intersecting white matter fascicles. Human glioblastoma cells also migrated on basement membrane lined blood vessels, pia-glia membrane and spaces of Virchow-Robin, as well as the subependymal space of the ventricles. These paths of migration of human glioblastoma cells in the rat brain are consistent with the pathways of spread of glioblastoma in the human brain as described by Scherer over 50 years ago, indicating that multifocal malignant astrocytomas have common migratory pathways in mature mammalian brain.

Alterations in the immunogenic properties of tumor cells frequently accompany selection for multiple-drug-resistant (MDR) variants. Therefore, studies were performed to examine the hypothesis that overexpression of membrane P-glycoprotein, commonly observed in MDR tumor cells, is associated with enhanced immunogenic properties. Immunogenicity was determined by (a) the ability of drug-sensitive parental UV2237M fibrosarcoma cells and drug-resistant UV2237M variant cells to immunize normal mice against rechallenge with parental tumor cells and (b) the ability of normal syngeneic mice to reject cell inocula that caused progressive tumor growth in immunocompromised mice. Variant UV2237M cell lines included subpopulations selected for a six- to ten-fold increase in mRNA for P-glycoprotein and expression of the MDR phenotype (resistance to doxorubicin) and cells sensitive to doxorubicin (and no expression of MDR properties) but resistant to ouabain. All UV2237M drug-resistant cells were highly immunogenic in immunocompetent mice, regardless of their MDR phenotype. Additional studies showed that CT-26 murine adenocarcinoma cells, sensitive or resistant to doxorubicin (expressing high levels of P-glycoprotein), injected into normal syngeneic Balb/c mice produced rapidly growing tumors. The data do not demonstrate a correlation between the immunogenic properties of drug-resistant tumor cells and the expression of P-glycoprotein.

The immunoglobulin (IgG, IgM, IgA) content of normal, reactive and migrated rat astrocytes was studied in lesioned adult rat cortex and after fetal cortex grafts. Implantation pockets were aspirated in the somatomotor cortex (under bregma) of adult Sprague-Dawley rats. A fetal (E14) rat hemicortex incubated in the plant lectin Phaseolus vulgaris leucoagglutinin (graft premarker) was placed in the aspiration pocket or the pocket left empty (control). Sections of brain were immunohistochemically double labeled for immunoglobulins-GFAP or Phaseolus vulgaris leucoagglutinin-GFAP 3–60 days postoperative. Mature or fetal astrocytes in situ did not contain immunoglobulins. In pocket-only animals, individual reactive astrocytes lining and subjacent to the pocket were positive for rat IgG, IgM and IgA (3–60 days). In grafted animals, graft derived astrocytes (Phaseolus vulgaris leucoagglutinin-GFAP positive) were intermingled with host reactive astrocytes (Phaseolus vulgaris leucoagglutinin negative) lining and subjacent to the pocket. Both classes of astrocytes contained immunoglobulins IgG, IgM and IgA. Immunoglobulin positive graft derived astrocytes migrated into the corpus callosum, cingulum and habenula. These results demonstrate that astrocytes sequester immunoglobulins and probably other serum proteins as a critical function in restoring homeostasis to the injured or diseased nervous system.

Background Understanding the mechanism of prostate cancer metastasis is essential to the design of a more effective therapy. An effective therapy for this disease will depend on the development of a clinically relevant in vivo model. Purpose We describe the development of such a model by using orthotopic implantation of human prostate cells in BALB/c nude mice. Method We compared the tumorigenicity of and the incidence of metastasis of human prostate cancer PC-3M and LNCaP-FGC (LNCaP) cell lines subsequent to prostatic (orthotopic) or subcutaneous (ectopic) implantations in male nude mice. Results LNCaP cells produced tumors only in the prostate. Enhanced tumorigenicity at the orthotopic site was found for PC-3M cells. Lymph node metastases were observed in practically all mice given an injection of PC-3M cells in the prostate, but they were uncommon with subcutaneous injection of these cells. Bilateral orchiectomy did not alter the tumorigenicity of either PC-3M or LNCaP cells or the incidence of lymph node metastasis by PC-3M cells. LNCaP tumors in the mouse prostate (but not PC-3M tumors) elaborated detectable levels of human prostate-specific antigen (PSA) in the serum, even when tumors were small (1.5 mm in diameter). Immunohistochemistry analysis revealed the presence of the PSA marker in tissue sections of LNCaP but not of PC-3M tumors. Conclusions The implantation of human prostate cancer cells in an ectopic environment does not permit expression of metastatic potential. In contrast, intraprostatic implantation does. Implications These data suggest that the orthotopic injection of human prostate cancer cells into the nude mouse may provide a valuable model to study the biology and therapy of human prostate cancer.

Fetal basal ganglia astrocytes and C6 glioma cells were plated on the surface of 1.5 cm thick hydrated collagen I wafers. Both cell types migrated through the entire thickness of the wafer within 1 day after plating. The collagen in the wafer was digested and the fine collagen I fibrils were clumped into large strands. By 2-3 days, the collagen strands were digested from the wafers and replaced by a mass of fetal astrocytes or C6 cells joined by their processes. The collagen I digestion and cell migration suggested protease production. In a second series of experiments, cultured C6 cells and E14 fetal astrocytes were immunohistochemically stained for the presence of plasminogen activators as an index of protease production. Both tissue (tPA) and urokinase (uPA) types were observed. Fetal astrocytes and C6 cells were also positive for guanidinobenzoatase, a serine protease associated with migrating cells. These data demonstrate that rapid migration of the cells on and through collagen I fibrils is concomitant with expression of plasminogen activators and proteases which can either activate or function as collagenases and release the cells from the substrate.

C6 tumor cells (10(6] were grafted as suspensions into freshly made implantation pockets in rat host cerebral cortex. Specimens were prepared for transmission and scanning electron microscopy 1 to 7 days postimplantation (DPI). By 3 DPI vacuolated C6 cells had migrated on or invaded the host brain. C6 cells were observed on the glia limitans on the surface of the brain, in the corpus callosum, subependymal space, and perivascular space and had invaded the cortex under the implantation pocket. In addition to the tumor mass that was observed under the implantation pocket, by 7 DPI individual C6 cells had migrated into the corpus callosum and internal capsule. Migrated C6 cells were observed in a perineuronal position in the hippocampus and other gray matter structures inferior to the corpus callosum. Micropockets were found around each C6 cell and the processes of these cells had replaced host parenchyma. The preferred routes of migration were on basal lamina and parallel and intersecting nerve fiber bundles. Invasion occurred through gray and white matter. The movement of homografted C6 cells in the brain suggests that these cells actively migrate as individual cells in addition to invading as a mass.

Scanning electron microscopic studies revealed that the cell bodies of cultured C6 cells were densely covered with villi and that membrane blebs were rare. Exposure to a hydroxyl radical generating system resulted in rapid ultrastructural changes. Within 2.5 minutes, 98% of the cells lost their villi. The number of blebbed cells increased during the first 7.5 minutes of exposure, until virtually all cells were blebbed. After 20 and 30 minutes of exposure, the plasma membrane of the blebs ruptured, resulting in the escape of cell contents and cell necrosis. Trypan blue exclusion, a measure of cell viability, decreased between 1 and 7.5 minutes of exposure. No further significant decline in viability was observed for the remainder of the experiment. The hydroxyl radical generating system also initiated a rapid onset of membrane lipid peroxidation, as measured by accumulation of thiobarbiturate reactive material. After a short lag period, thiobarbiturate reactive material increased rapidly, with maximum lipid peroxidation occurring by 20 minutes. Increasing the concentration of the radical generating system components shortened the time course but not the level of maximal thiobarbiturate reactive material production. These data suggest that plasmalemmal lipid peroxidation plays a role in rapid morphological changes and necrosis that occur after free radical insult.

Cortically homografted C6 glioma-astrocytoma cells both invade the rat host brain as a mass and migrate as individual cells. In contrast, fetal astrocytes derived from homografted whole pieces of fetal cortex migrate only as individual cells throughout the brain of the rat but are not capable of invasion. Our experiment explored the migratory capacity (over 7 days) of cultured purified fetal astrocytes and C6 cells after seeding 10(6) cells on a hydrated artificial basement membrane wafer (Matrigel). The artificial basement membrane wafer was not a suitable substrate for the growth of cultured fetal astrocytes. In contrast, C6 cells migrated as individual cells from the surface of the wafer into the substrate. Individual C6 cells migrated 1.8 mm in the first 4 days and then ceased migration. The C6 cells were observed at the base of a digestion tube that extended from and was open to the surface of the wafer. At 3 days, micropockets were observed to form around each C6 cell at the base of each tube. By 7 days, the majority of pockets observed were large and contained several C6 cells. These multiple cell groups appeared to be progenitors of tumor masses. These data indicate that C6 glioma-astrocytoma cells, which in vivo appear to be a model for glioblastoma multiforme, primarily migrate as individual cells through artificial basement membrane and secondarily form tumor masses. Progenitor tumor masses form by coalescence of individual C6 cell micropockets or the division of a single cell in an individual micropocket.

Neuropeptide receptors were visualized in homografts of fetal cortex (E14) into adult rat cortex (immediate or 7 day delay) and spinal cord usingin vitroautoradiographic techniques to explore the expression of peptide receptors in the same graft tissue in different central nervous system implantation sites. Receptors for bombesin (BN)-like peptides developed in the grafts by 3 weeks postimplantation regardless of location or age of implantation pocket in host. After 4 weeks, the density of BN receptors was confined to the graft. In grafts to spinal cord, however, high densities of BN-like receptors were not confined to the graft but were distributed throughout the spinal cord. In contrast, the density of vasoactive intestinal polypeptide (VIP) and substance P (SP) receptors was moderate and low to undetectable in the fetal grafts. The development of the peptide receptors studied was graft donor tissue specific since they were not altered by central nervous system implantation site.

The effects of spinal cord hemisection on protein labeling in the medulla, pons, midbrain, and cortex in the Cebus monkey were studied by measuring the uptake of [3H]lysine into the trichloroacetic acid (TCA) precipitable and non-precipitable fractions of brain. Animals were sacrificed 3, 6 or 13 days after hemisection. Two normal animals and a 3-day sham operate were utilized as controls. Differences included individual animals, operated vs. unoperated animals, brain areas, and variation of brain areas over all postoperative days. The overall trend in brain radioactivity over postoperative days indicates stress and non-stress related components of brain response to spinal injury. Left-right side differences (L greater than R) were noted in the uptake of [3H]lysine into specific areas of brain over postoperative time, including leg areas of motor cortex, occipital cortex, and superior colliculus. A biphasic action of leg sensory cortical areas over time was noted, where L greater than R for the first 6 postoperative days and R greater than L at day 13. The factors possibly mediating radiolabeled amino acid uptake into protein are discussed in relation to demonstrated anatomical and neurochemical regenerative processes.

Mammalian spinal motoneuron dendrites respond with cyclic degeneration and regeneration after ventral root crush. In the following experiments, the cross-sectional dendritic profile of rat lamina IX, medial, motoneurons under the T2 vertebra were analyzed after mincing the dorsal horn (normals, 14, 30, 60, 90 days postoperative (DPO); N = 6 animals/DPO). The spinal cords were impregnated by the tungstate modification of the Golgi technique. Individual lengths along dendritic segments between branching points were measured from coded slides, the data were computerized, and the dendrites were reconstructed by computer. Interanimal statistical comparisons were made by ANOVA a priori and Newman-Keuls test a posteriori. At 14 DPO, there was a statistically significant (P less than 0.05) increase in the number of dendritic segments, dendritic and serpentine length, and number of segments emanating from the soma compared with normal intact rats and with all other postoperative days. At 30 and 60 DPO, these parameters returned to normal values; however, there were many long, unbranched dendrites. At 90 DPO, there was a statistically significant (P less than 0.05) decrease in motoneuron dendritic serpentine length and segments. These data show that partially deafferented rat spinal motoneurons undergo a biphasic response; an initial growth phase followed by a degenerative phase.

Qualitative behavioral deficits after bilateral lesion of the rat dorsal funiculus at T12–13 were studied during prelesion, postoperative (14 to 21 days postlesion), and terminal (the last 4 days before killing at 30, 45, 60, 90, or 120 days postlesion) periods. Utilizing each animal (36 with lesion, 5 sham stress) as its own preoperative control, hind limb performance was rated for five behavioral components (slips, recovery, manner of traverse, movement, tail position) on four apparatuses (wide platform, narrow platform, ladder, parallel bars). Significant postoperative deficits ( N = 36) occurred on slips, recovery, and manner; Spearman coefficients between apparatus pairs suggested that either of the platforms and/or parallel bars could have demonstrated most component deficits. Sham-stress controls ( N = 5) showed no significant deficits in postoperative performance. Postoperatively, lesion size correlated with some performance deficits. Also, significant postoperative deficits were obtained with 25 to 87% damage of the dorsal column, though solely in the manner of traverse as an individual component; additional destruction (90 to 100%) of the dorsal column added several other deficits. Larger dorsal funicular lesions which included at least 50% of the corticospinal tract significantly affected performance still further, particularly on the parallel bars. Terminal performance ( N = 24), in general, was significantly poorer than that at prelesion periods on 23 (of 29 possible) comparisons. Terminal performance degraded considerably at 90 and 120 days compared with post-operative behavior. Because, in related deafferentation studies, changes in bouton renewal and neuronal excitability can occur at these late postlesion periods, eroded behavioral deficits may reflect a loss of synaptic elements critical in mediating stable behavior.

The short term effects of spinal cord hemisection on protein synthesis in the Cebus monkey were studied by measuring the uptake of [ 3 H]lysine into the trichloroacetic acid (TCA) precipitable and non-precipitable fractions of spinal cord. Animals were sacrified 3, 6, or 13 days after hemisection. Two normal animals and a 3 day sham operate were utilized as controls. All data were pooled and subjected to analysis of variance. Significant group differences were resolved by a standardt-test over pooled operative group samples and a pairedt-test for left-right differences within groups. Compared to normal animals, significant increases in the mean uptake of precursor into TCA precipitable and non-precipitable fractions of spinal cord samples were noted in all postoperative groups. The observed increase in the uptake of [ 3 H]lysine into TCA soluble and insoluble fractions at 3 days after hemisection were attributed to stress, as were increases in the activity of the TCA insoluble fractions in the sham operate. By 6 days, the increases attributed to stress in the average protein activity seemed to subside, resulting in an overall decrease in the TCA precipitable protein activity with an increase in TCA soluble fraction activity. Between 6 and 13 days the activity of the TCA soluble and insoluble fractions returned to the 3 day levels with a sharp increase in the activity of both fractions in the area of the hemisection. This increase was attributed to increases in neuronal, neuroglial, and infiltrating cell metabolism possibly associated with regenerative changes.

IgG-positive astrocytes have been reported in scrapie-induced and Alzheimer's cortical plaques, multiple sclerosis, and CNS tissue around abcesses, metastatic tumors and primary tumors of glial origin. The present experiments ascertain if this immunoglobulin positivity is specific for these cases or a function of astrocytes around any site of injury in the CNS. The spinal cords of 30, 300-g Sprague-Dawley male rats were lesioned by passing a 26 gauge needle through the cord at T 6 . After periods as long as 9 months, the spinal cords were processed for paraffin immunohistochemistry with antisera to IgM, IgG or double labeled for these immunoglobulins and GFAP (glial fibrillary acidic protein) a specific cytoplasmic marker for astrocytes. From 1 through 7 days after lesioning double labeled astrocytes in and around the site of injury are both IgM- and IgG-positive. From 14 days through 9 months postlesion, double labeled astrocytes surrounding the lesion are only positive for IgG. These data indicate a relationship between immunoglobulin availability, continued blood-brain barrier perturbation to immunoglobulins and the ability of reactive astrocytes to sequestor immunoglobulins. IgM is an early determinant for reactive astrocytes and IgG positivity is determinant for reactive astrocytes at any time period.

Fetal spinal cord grafts survive for long periods in host spinal cord. We explored the ability of such grafts into a lesion in C3 fasciculus gracilis to influence the expected deterioration of hind limb performance following this lesion. Ten adult Sprague-Dawley rats were placed on a schedule of 23 h off and 1 h on water. These subjects were then trained to traverse a narrow platform for a water reward. Criterion was established as 10 trials/day completed in 15 s a trial for 2 consecutive days. Animals were ranked for hind limb and forelimb performance utilizing slips, recovery, and manner of traversing the platform. In a triple-blind experimental design, the animals' identification was coded and laminectomy performed at C3. The fasciculus gracilis was bilaterally transected at the rostral and caudal borders of the segment and aspirated. Five subjects were chosen at random immediately after the lesion was made for the implantation of two 1-mm segments of E14 spinal cord from fetuses of timed pregnant Sprague-Dawley dams at the time of the lesion. Coded lesion and lesion-transplant animals were then tested after 21, 30, 45, 60, and 90 days until criterion or for 4 days. Hind limb performance of the lesion-only (ANOVA chi 2 = 13.04, df 5; P less than 0.05) and lesion-transplant groups (ANOVA chi 2 = 11.82, df 5; P less than 0.05) showed that the C3 fetal transplants significantly decreased (P less than 0.05) the severity of hind limb deficit at 21 days postlesion and significantly (P less than 0.05) decreased the expected foot slips and limb recovery to the platform that was expected and did develop in lesion-only animals at 90 days. There were no significant differences in forelimb performance (ANOVAs P greater than 0.05). There were few small-caliber nerve fibers in the fasciculus gracilis at the level of C1 indicating that reinnervation was not a major factor in the reduction of the deficit.

Successful transplantation of the fetal brain into adult host brain has been accomplished. These studies explore the growth, differentiation, and viability of E11, E12, and E15 rat fetal cortex and fetal spinal cord implantation into the spinal cord of adult rats (donor and host, Sprague-Dawley). Under deep Chloropent anesthesia, 70 rats had 1-mm cubes of fetal cortex inserted with pressure or by stylus injection subpially between the dorsal horn and dorsal column (left side), or implantation of whole segments of fetal spinal cord. Animals were prepared for light microscopy 14 and 21 days and 1, 2, and 3 months later. Implants by both fetal tissues had a 69% survival rate. The younger the fetal implant the higher the success of the implant (E11 greater than E15). The diameter of fetal spinal cord implants reached the diameter of control postnatal animals after 30 days. The implants not only increased in mass (up to 7-fold in some cases) but differentiated and matured (apolar, unipolar, bipolar, and multipolar) neurons were observed one to three months postimplantation. By 30 days postimplantation, fetal neurons had large, often crenated nuclei, with a large single prominent nucleolus. The most successful implants were the young E11 fetal spinal cord into the adult host spinal cord. These implants represent an initial successful transplantation of fetal spinal cord into adult spinal cord.

The ependymal cells in rat host spinal cord were studied after transplantation of fetal cerebral cortex homografts. Adult Sprague-Dawley male rats had pieces of E14 fetal rat cortex pressure-injected into the spinal cord at T6. The tissue was prepared for light and electron microscopy and studied for over 2 months postimplantation. At 7 and 14 days postimplantation (DPI), there were dividing ventricular-ependymal cells lining cysts in the host spinal cord. After 21 days, cell division was no longer observed in these cells, and only mature ependymal cells lined these cysts. Mature ependymal cells were either: columnar, pseudocolumnar, cuboidal, squamous, or pseudomorphic; had cilia and villi on only one surface of the cell and shared tight junctions when contiguous. These data show that ependymal cells were mature cellular components of adult spinal cord after transplantation of fetal CNS grafts. In addition, ventricular-ependymal cells continued to divide within the parenchyma of the graft, indicating continued growth of transplanted fetal cerebral cortex in host spinal cord.

The fasciculus gracilis (FG) of the third cervical spinal cord segment (C3) of adult rats was aspirated and fresh whole pieces of rat 14-day gestation cervical spinal cord, prelabeled with Phaseolus vulguris leucoagglutinin, immediately placed in the pocket of one-half the operates. One to three months later there were no transplant derived rostro-caudal nerve fibers in the C1 or C2 segment of the aspiration-only or aspiration-graft FG. However, the neurons of the nucleus gracilis of the host medulla of the aspiration-only group were significantly atrophied whereas ther neurons of the aspiration-graft group were not statistically different from normal. There were normal numbers of neurons in all groups. Using immunohistochemical double labeling for prelabeled astrocytes with PHAL and GFAP, graft-derived astrocytes were found in the necleus gracilis of the host of the host medulla. These data indicate that the trophic action of spinal graft-derived astrocytes maintained neurons in the medulla of the host nucleus gracilis.

There appears to be specificity in the regrowth of nerve fibers in the fish central nervous system. However, the distribution of terminal boutons is unknown. This study was undertaken to ascertain the reinnervation of neurons by boutons in the regenerated goldfish spinal cord. The spinal cord of 60 goldfish was transected and a segment 2 cm caudal to the lesion was examined 5, 10, 15, 20, 25, 30, 45, and 60 days postoperatively. In another group of animals the spinal cord was retransected after 60 days regeneration time and examined 5 days later to show contributions of long regenerated tracts. The spinal cord segment was stained for boutons by the Rasmussen technique. Boutons were counted on the soma of neurons in medial intermediate gray, ventral horn motoneurons, and ventral horn interneurons and in addition, on primary dendrite of motoneurons. Animals were observed daily for the return of normal swimming. Neurons in the intermediate gray region showed a gradual linear decrease in boutons for 30 days and they remained at this level from 30 to 60 days. Motoneuron soma and primary dendrite showed an abrupt decrease in boutons at 5 days, with a linear increase in boutons to normal levels by 60 days. Retransection of the spinal cord after 60 days did not result in bouton loss on soma or dendrite. These data indicate that the motoneuron was originally innervated by descending long tracts but was now reinnervated by other tracts. Ventral-horn interneuron soma showed an initial loss of boutons followed by return to normal numbers by 30 days and hyperinnervation between 30 and 60 days. Retransection resulted in the return of boutons to normal levels. Thus, this interneuron was reinnervated by regenerated descending tracts. Despite this synaptic reorganization of spinal neurons, normal swimming returned. We conclude that specificity of regeneration is not selective for soma or dendrite of spinal neurons but is probably selective for the neuronal pools or circuits necessary for swimming.

Summary L1210 leukemia cells are 2–4 fold more sensitive to the cytolytic effects of melittin, the membrane-active toxin of bee venom, than normal DBA/2 mouse spleen and bone-marrow cells. Lysis of the normal cells was abolished when either 75 mM galactosamine, glucosamine or 100μM β-lactoglobulin was added to the melittin-cell reaction, but lysis of the leukemia cells was unaffected. The amino-groups appeared necessary for blocking melittin-mediated lysis since glucose, galactose and the N-acetyl derivatives were not inhibitory. Bone-marrow cells were more readily protected from lysis than spleen cells. Since melittin-inhibitor complexes were not detected by gel chromatography and the inhibitor could be added to the cell suspension after melittin, the evidence suggests that bone-marrow cells are rich in membrane binding sites for carbohydrates that decrease in mature spleen cells and are virtually absent after neoplastic transformation.

Cellular surface carbohydrates have been studied extensively in relation to cell interactions. However, the presence of intracellular carbohydrates not associated with cell surface membranes has not been well-defined histochemically. We describe a technique for the ultrastructural localization of intracytoplasmic carbohydrates using a wheat germ agglutinin peroxidase conjugate. The ultrastructural aspects of wheat germ agglutinin (WGA) binding to intracellular structures of rat somatomotor cortex layer V neurons and spinal cord motoneurons were examined using sonication to enhance penetration of the lectin–peroxidase conjugate. In addition to the membrane-associated structures previously described by others, WGA binding was observed on neuronal nuclear chromatin, nuclear pores, ribosomes, microtubules, vesicular organelles, and to synaptic complexes, primarily in the presynaptic bouton. The results support data indicating that complex carbohydrates may be important in neuronal molecular processing.

Alteration of spinal motoneurons can be induced by axotomy, deafferentation, or deafferentation of spinal interneurons. In our experiment, dendritic profiles of rat lamina IX motoneurons were studied 1.5 to 2.5 mm rostral to a T2, T3 (interface) hemisection [six per group of normal, 7-, 14-, 30-, 60-, and 90-day postoperative (DPO) animals]. The spinal cord was impregnated using the Golgi technique and individual dendritic segments measured from coded slides, the data were computerized, and the length, number of segments, and dendritic profile were reconstructed. These data were heteroscedastic (alpha less than 0.005) making them inappropriate for ANOVA. However, the data did show two size classes of motoneurons at 14, 30, 60, and 90 DPO. At 7 DPO, there was a reduced dendritic field. At 14 to 60 DPO, there were gigantic motoneurons that were many times larger than normal or other neurons in these operated groups. At 90 DPO a large motoneuron was present. These data show that individual motoneurons have different regenerative dendritic responses after spinal injury.