The relationship of expression of statin, the nuclear protein of nonproliferating cells, to the differentiation and cell cycle of astroglia in cultures and in situ

Cell division and aging of the organism

The capacity to regenerate cell compartments through cell proliferation is an important characteristic of many developed metazoan tissues. Pre- and post-natal development proceeds through the modifications occurring during cell division. Experiments with cultivated cells showed that cell proliferation originates changes in cell functions and coordinations that contribute to aging and senescence. The implications of the finite cell proliferation to aging of the organism is not the accumulation of cells at the end of their life cycle, but rather the drift in cell function created by cell division. Comparative gerontology shows that the regulation of the length of telomeres has no implications for aging. On the other hand there are interspecies differences in regard to the somatic cell division potential that seem to be related with the "plasticity" of the genome and with longevity, which should be viewed independently of the aging phenomenon. Telomeres may play a role in this plasticity through the regulation of chromosome recombination, and via the latter also in development.

Prosaposin in the secretome of marrow stroma-derived neural progenitor cells protects neural cells from apoptotic death

Functionally, adult stem cells not only participate in replication and differentiation to various cell lineages, but also may be involved in rescuing cells from apoptosis. Identifying functional factors secreted by stem cells, as well as their target cells, may advance our understanding of stem cells' multifaceted physiologic functions. Here, we report that mouse bone marrow stromal cell-derived neuroprogenitor cells (mMSC-NPC) provide a protective function by secreting a key factor, prosaposin (PSAP), capable of rescuing mature neurons from apoptotic death. This factor is identified as the lead protein in the secretome of mMSC-NPC cultures by tandem mass spectroscopic profiling, and further validated by western blotting and immunocytochemistry. The secretome of MSC-NPC reduces toxin-induced cell death in cultures of rat pheochromocytoma neuronal cells, human ReNcell CX neurons, and rat cortical primary neurons; removal of PSAP by immunodepletion annuls this protective effect. This neuronal protection against toxin treatment was validated further by the recombinant PSAP peptide. Interestingly, the secretome of neuronal culture does not possess such a self-protective action. We suggest that upon injury, a subgroup of MSCs differentiates into neural/neuronal progenitor cells, and remains in this intermediate stem cell-like stage, defending injured neighboring mature neurons from apoptosis by secreting PSAP.

From Weismann's theory to present day gerontology 1889-2003

From Weismann's theory to present day gerontology--Weismann's theory was based on the concept that through natural selection the division potential of somatic cells become finite thus limiting the regeneration of the soma and the life span of the organism. Indeed, the somatic cells of some animals have a finite division potential but what became apparent is that the implications for aging are more complex. Experiments showed that at each cell division the genetic information received by each daughter cell differs; cells are this way progressively modified through division creating a functional drift that is responsible in part for the continuous modifications going on in the organism from its very beginning to its extinction. Comparative biology showed that the finite or the infinite division potential of somatic cells has a complex connotation with developmental characteristics of the respective organism with implications for longevity that are far from being understood.

Nonlamin components of the lamina: a paucity of proteins

Current models of nuclear organization propose that nuclear functions are modulated in part by reversible tethering of chromatin loops to structural elements of the nucleoplasm and the nuclear envelope. Lamins are the best-characterized proteins of the lamina portion of the nuclear envelope and are involved in binding chromatin to the inner nuclear membrane. However, they are not a universal feature of eukaryotic nuclei and do not account fully for the putative functions of the lamina in all organisms. It is possible that nonlamin components of the lamina may substitute for lamins in organisms from which they are absent and modify the properties of lamins during development and the cell cycle. We review the properties of the relatively small number of such components that have been reported, including the young arrest (fs(1)Ya) protein of Drosophila, statin, circumferin, and the MAN antigens. The experimental evidence indicates they are a diverse group of proteins, and that at least some have the potential to modulate the interactions of chromatin, lamins, and the nuclear membranes.Key words: nuclear envelope, lamina, YA protein, statin, circumferin.

Proliferative status of cells in adult human dentate gyrus

Experiments in rodents and marmoset monkeys indicate that granule neurons of the dentate gyrus may be renewable throughout the entire life of the animal. Whether this occurs in larger primates remains a matter of contention. However, a recent study of brain samples from five adult humans who had been injected with the thymidine analog bromodeoxyuridine indicates that new neurons might indeed be produced in the dentate gyrus. In this study, hippocampus specimens removed from 18 adult humans for treatment of epilepsy were examined. The cell cycle marker Ki67, which is expressed from late G1 to M phase, was demonstrated by immunohistochemistry, and H2b/H3/H4 histone mRNAs, which are expressed during S phase, were demonstrated by in situ hybridization. Only 0.17% of cells in the subgranular layer, the site of neuronal progenitor cells, were Ki67 immunoreactive but the identity of these could not be proven. Although the histone in situ hybridization technique was shown to work in human fetal brain, no M phase cells could be demonstrated in the hippocampus. The generation of new granule neurons in the human hippocampus must occur at a very slow rate. The approaches used in this study are likely unsuitable for studying cell populations with low turnover rate. Further work is needed to determine the fate of newly generated cells in the dentate gyrus. This information is of importance to our understanding of the mechanisms of learning and memory.

Lesioning of the inferior olive using a ventral surgical approach. Characterization of temporal and spatial astrocytic responses at the lesion site and in cerebellum

Activated astrocytes, intrinsic components of both local and remote (axonal target regions) central nervous system injury responses, are now recognized as active metabolic and regulatory mediators in many neurological disorders. To further define these responses, we devised a new ventral surgical approach to unilaterally lesion the inferior olivary nuclear complex, which has a single predominant remote target, the cerebellum. Activated astrocyte number, volume, and density, as well as the total volume of brainstem involved in the astrocytic response, all peaked at postlesion day (pld) 4, returning toward, but not to, unoperated control values at pld 24 (p < 0.05). In contrast, the peak astrocyte response in the cerebellum was delayed, being greatest at pld 6 (p < 0.05 compared to control or pld 2). These responses were associated with increases in overexpression of S100 beta, an astrocyte-derived neurite growth factor, and with an increase in cerebellar steady-state levels of a neuronal injury response protein, the beta-amyloid precursor protein (beta-APP). This is similar to correlated increases in these two proteins that are found in epilepsy and Alzheimer disease. Our studies defining remote astrocytic and neuronal responses may be important for understanding glial-neuronal mechanisms underlying the spread of neuropathological changes in conditions such as Alzheimer disease.

Remodelling of the nuclear periphery during muscle cell differentiation in vitro

We have examined the composition and ultrastructure of the nuclear periphery during in vitro myogenesis of the rat myoblast cell line, L6E9. Immunofluorescence labelling and immunoblotting showed that lamins A/C and B were all present in undifferentiated cells, but that they increased significantly before extensive cell fusion had occurred, with lamins A/C increasing proportionately more. Electron microscopic observations were consistent with these results, showing an increase in the prominence of the lamina during differentiation. On the other hand, immunofluorescence labelling suggested that the P1 antigen began to disappear from the nuclear periphery as the cells were fusing, after the increase in lamin quantity, and was no longer detectable in multinucleated cells. Unexpectedly, however, P1 was readily detected in isolated nuclei, whether prepared from myoblast or differentiated cultures, as well as in both myoblast and myotube nuclear matrices. It appears probable, therefore, that the fading of P1 labelling is due to masking of the epitope by a soluble factor recruited to the nuclear periphery as cell differentiate. These data, together with evidence that the genome is substantially rearranged during L6E9 myogenesis [Chaly and Munro, 1996], suggest that L6E9 cells are a useful model system in which to study the interrelationship of nuclear envelope organization, chromatin spatial order, and nuclear function.

Increased levels of statin, a marker of cell cycle arrest, in response to hippocampal neuronal injury

Injured neurons in the CNS are known to synthesize high levels of proliferation related oncogene products and heat shock proteins without dividing. Statin is a cell cycle regulated nuclear phosphoprotein, selectively associated with the non-proliferative state in a wide variety of cell types. In the present study, neuronal statin was examined following lethal or sublethal neuronal injuries in the hippocampus of Alzheimer's disease patients, in rats receiving kainate lesions to the dorsal hippocampus and in entorhinal cortex lesioned rats. Immunolabelling of nuclear statin showed that statin immunoreactivity increased preferentially in CA1 pyramidal neurons of the hippocampus in Alzheimer's disease. In kainate lesioned rats, statin immunoreactivity was markedly induced in the CA3 hippocampal region in association with neuronal loss. Entorhinal cortex lesioned rats showed a transient induction of statin between 2 and 6 days post lesion in CA1 neurons. However, cell counts in entorhinal cortex lesioned rats remained unaltered in the CA1 and granule cell layers during the entire 30 day time course, indicating that increased statin levels are not secondary to neuronal degeneration and are not necessarily accompanied by irreversible neuronal death. It is concluded that, in addition to proliferation related gene products, neuronal injury induces an increase in levels of statin, a nuclear marker of cell cycle arrest. Furthermore, statin may be a potentially useful marker of injurious neuronal stress, even under conditions that do not necessarily lead to irreversible cell death.

Cytokine stimulation increases intracellular calcium and alters the response to quisqualate in cultured cortical astrocytes

Cytokine levels are elevated in the central nervous system (CNS) in a variety of disorders and may contribute to abnormalities in CNS function associated with the disorders. To begin to understand the mechanisms through which elevated cytokines affect CNS cells, we have examined the effects of cytokines on astrocyte physiology within minutes of application as well as 24 h later. Both standard cultured cortical astrocytes and those induced to further differentiate by pre-treatment with forskolin were examined. Such treated astrocytes may more closely resemble those in brains exhibiting elevated cytokine levels. The cytokine focused upon was interleukin-1-beta (II-1 beta). Gamma-interferon (gamma-IFN) and tumor necrosis factor-alpha (TNF-alpha) were also examined in some studies. Changes in calcium levels produced by acute application of these cytokines were measured. The most pronounced effect was an immediate calcium elevation in response to II-1 beta in the forskolin pre-treated astrocytes. Longer term treatment with IL-1 beta in forskolin pre-treated astrocytes enhanced the calcium response to quisqualate stimulation, a glutamate neurotransmitter receptor agonist. These results suggest that situations that cause chronic changes in cytokine levels and involve astrocytic differentiation, such as chronic CNS infection or Alzheimer's disease, could change astrocytic responses to normal stimuli. Such changes may result in altered astrocytic support of neurons and therefore cause changes in CNS function.

Expression of the nuclear membrane protein statin in cycling cells

Statin is a 57 kD protein previously reported to be expressed by cells in G0. We have studied the detailed distribution of statin immunoreactivity in normal human and rat tissues, and correlated this with investigation of in vitro model systems. By laser confocal microscopy, statin immunoreactivity is localized to the nuclear membrane. In contrast to previous reports, using in vitro model systems we found that statin was also expressed by replicating cells as judged by both co-localization with [3H]thymidine-labelled and Ki67-labelled cells. Furthermore, in a nude mouse xenograft model the number of statin-labelled cells exceeded the number of quiescent cells as assessed by both fraction of labelled mitosis methods and labelling with [3H]thymidine and Ki67. We conclude that although there is an association between expression of the 57 kD nuclear membrane protein statin and growth arrest, this is not absolute and it is expressed in a sub-population of cycling cells. The properties of statin closely resemble those of nuclear lamins, members of the intermediate filament family.

Autoradiographic studies of rat astroglial cell proliferation in vitro with and without treatment with basic fibroblast growth factor

Using specific autoradiographic methods, cell cycle parameters of untreated and basic fibroblast growth factor (bFGF)-treated astroglial cells from newborn rats grown in primary culture were directly measured. The mode of proliferation was also analysed. In untreated cultures, S phase duration (Ts = 6.9-13.1 h) and cell cycle time (Tc = 10-18 h) can be modified by about a factor of 2 depending on the culture conditions (serum-supplemented or defined medium, thyroid hormone concentration). However, growth fraction (GF = 0.15) and the ratio Ts/Tc remain stable. With increasing days in vitro (DIV) (DIV 7-DIV 20), Ts (7.8-10.6 h) and Tc (10-21 h) are prolonged and GF (0.14-0.06) decreases, probably due to cell maturation. In general, astroglial cells proliferate exponentially with a GF < 1, but stop proliferating about 30-36 h after the last feeding, probably caused by exhaustion of the medium. However, after refeeding they continue to proliferate. As opposed to in vivo, no transition of non-proliferating cells into the GF occurs. After addition of bFGF, GF increases (e.g. GF at DIV 7 = 0.43), but Ts and Tc are not influenced at DIV 7 and 12. At DIV 20, bFGF additionally shortens Ts and Tc, thereby producing values of Ts, Tc and GF like 'younger' cultures. However, the revitalizing effect on 'mature' cells is only transitory. In general, bFGF leads to a single re-entry of G0 cells into the GF. Thereafter, bFGF does not affect the mode of proliferation.

Expression of the non-proliferation-specific protein, statin, in grey matter neuroglia of the aging rat brain

The monoclonal antibody, S-44, identifies statin, a 57 kDa nuclear protein which appears to be expressed exclusively in non-proliferating cells. We previously demonstrated that in the aging rat corpus callosum approximately one third of neuroglia are statin-negative, suggesting the existence of an unexpectedly large cycling glial compartment. In the present study, double-labeling of individual cultured astroglia with [3H]thymidine and the S-44 antibody provided direct evidence for the non-proliferative status of statin-positive cells. The S-44 antibody was used to immuno-localize statin and thereby determine growth fractions for neuroglia in various grey matter regions of 3-, 18-, and 33-month-old rats. The proportion of statin-negative (cycling) cells for the three ages combined ranged from about 24% in the molecular layer of the dentate gyrus to 38% in the molecular layer of the parietal cortex. In most regions surveyed total glial counts and proportions of statin-positive and -negative cells did not vary significantly as a function of advancing age. These results suggest that (i) as in corpus callosum, pools of cycling neuroglia in various grey matter regions are far in excess of those previously predicted by S-phase labeling with [3H]thymidine or BUdR, and (ii) ratios of proliferating-to-quiescent neuroglia are tightly regulated over much of the animal's adult life span. These conserved ratios may be used as markers of normal CNS senescence, and deviations thereof may indicate the presence and extent of intervening neuropathologic processes.

Expression of a novel nuclear protein is correlated with neuronal differentiation in vivo

We report the production of a monoclonal antibody (MAb 526) that recognizes a novel, developmentally regulated nuclear protein expressed in neurons throughout the rat nervous system. Analysis of whole brain and cell nuclear extracts by SDS-PAGE and immunoblotting determined that MAb 526 recognizes a single nuclear protein (np) of apparent molecular weight 42 kD, designated np526, as well as a slightly larger (ca. 44 kD) cytoplasmic protein. Light microscopic immunocytochemistry showed np526 to be present in neurons of all types throughout the central and peripheral nervous systems. Nuclei of both fibrous and protoplasmic astrocytes were also immunoreactive, but oligodendrocyte nuclei were negative. Positive, but highly variable immunocytochemical staining of nonneural cell nuclei in a variety of other tissues was also observed. Electron microscopic (EM) immunocytochemistry using pre-embedding peroxidase methods revealed that np526 is associated with euchromatin or with the edges of condensed chromatin bundles in neurons, indicating that it is likely to be a chromosomal protein. Most interestingly, the expression of np526 was found to be developmentally regulated in brain. Immunocytochemical analysis of the developing cerebral cortex from embryonic day (E) 16 to postnatal day (P) 4 and cerebellum from P4 to P18 revealed that np526 first appears in central neurons following the cessation of mitosis and that the intensity of nuclear staining increases during subsequent neuronal maturation. To our knowledge, np526 is the first presumptive chromosomal protein whose expression has been precisely correlated with the early postmitotic differentiation of mammalian neurons.

Immuno-electron microscopical localization of vimentin and glial fibrillary acidic protein in mouse astrocytes and their precursor cells in culture

Immuno-electron microscopy was used to localize the distribution of vimentin and glial fibrillary acidic protein (GFAP) in mouse astrocytes and their precursor cells in primary cultures. In astroblasts and astrocytes, vimentin and GFAP form intermediate filaments (IF), which are heteropolymers, as previously observed in gliomas. Astrocytes and their precursor cells may have IF composed of GFAP-vimentin heteropolymer or vimentin alone, but IF composed of GFAP only were not seen. It seems that the formation of IF that are GFAP-vimentin heteropolymers is a feature of normal astroglia development and that the ratio of GFAP to vimentin in these IF reflects the degree of differentiation and functional state of the cell.

Morphology of astroglia in colony cultures following transient exposure to potassium ion, hypoösmolarity and vasopressin

Brain swelling is the major cause of delayed neuronal damage following injury to the central nervous system. Swelling of mouse astroglial cells was studied in colony cultures by light and electron microscopy. Swelling of suspended astroglial cells was studied by flow cytometry. Swelling caused by hypoösmolarity solution was more pronounced than that caused by 15 or 60 mM K+. Under both conditions swelling in both immature and mature astroglia was followed by a regulatory volume decrease. Arginine vasopressin caused mild astroglial swelling and atrial natriuretic peptide did not significantly affect cell volume. All changes in extracellular environment were associated with changes in the morphology of microvilli and varying amounts of membrane ruffling. Immature cells exhibited a delayed response to the application of atrial natriuretic peptide and less membrane ruffling following exposure to 60 mM K+ than mature astroglia. These nonspecific morphological changes are likely associated with changes in membrane ion pump activity.

The neuronal cytoskeleton

In this paper we have described the organization of F-actin and actin-binding proteins (ABP): alpha-actinin, myosin, tropomyosin, caldesmon, vinculin, talin, and spectrin, in differentiating astroglia in colony cultures. We observed that the microfilament (MF) network arrangements differ at various stages of astroglia development, but the composition of MF bundles and stress fibers is the same at all developmental stages. F-actin is closely colocalized with myosin, tropomyosin, caldesmon, and alpha-actinin. The striated pattern of myosin, tropomyosin, and caldesmon are superimposable. Tropomyosin and caldesmon extend along F-actin but are interrupted for short periods, whereas myosin is interrupted for longer periods. alpha-actinin colocalizes with tropomyosin and caldesmon but not with myosin. In astroglia at different stages of development spectrin is arranged in the form of fine networks spreading through the cell and does not follow the arrangement of MF bundles. Only F-actin, alpha-actinin, and vinculin can be detected at cell-cell junctions. In the areas of the focal contacts, F-actin, alpha-actinin, vinculin, and talin are present. They overlap each other, although talin and vinculin extend toward the cell membrane beyond F-actin and alpha-actinin. Astroglia undergo well-defined states of nonmotility, motility, and nonmotility again during differentiation. The changes in motility are paralleled by changes in the organization of F-actin and ABP: as GFAP-containing intermediate filaments increase in differentiating astroglia, the F-actin and ABP are down-regulated, leading to non motility.

Statin immunolocalization in human brain tumors. Detection of noncycling cells using a novel marker of cell quiescence

Surgical specimens of 35 human brain tumors were examined with a novel monoclonal antibody, S-44, immunoreactive to statin, a nuclear protein specifically expressed in quiescent (noncycling) G0-phase cells. Benign tumors typically were statin positive with labeling indices (LI) between 22% and 96%: acoustic schwannomas (n = 3, mean = 29.9 +/- 19.4%); meningiomas (n = 4, mean = 59.0 +/- 15.1%); pituitary adenomas (n = 3, mean = 79.9 +/- 28.2%), and an epidermoid cyst (41.0%). By contrast, the statin LI of 18 of 24 (75%) malignant brain tumors was less than or equal to 2%: medulloblastomas (n = 7, mean = 0.3 +/- 0.2%); anaplastic astrocytomas (n = 3, mean = 1.6 +/- 2.7%); glioblastomas (n = 10, mean = 10.3 +/- 14.4%); metastatic carcinomas (n = 3, mean = 3.0 +/- 4.6); and a germinoma (0.2%). The vascular endothelium among diverse tumors typically was statin positive. All 21 tumors with a statin LI less than 10% were malignant, and all nine tumors with a statin LI greater than 40% were benign. The statin LI of benign tumors (n = 11, mean = 55.1 +/- 26.7%) was significantly higher than that of the malignant tumors (n = 24, mean = 5.2 +/- 10.5%, P less than 0.001). The absence of statin expression is a new way to determine the malignancy of human brain tumors. The statin LI may be useful to guide the prognosis and treatment of individual patients. The mechanisms that control statin expression are important in therapy seeking to shift the proliferating, cycling cells to the quiescent, G0 compartment.

Dynamic features of cells expressing macrophage properties in tissue cultures of dissociated cerebral cortex from the rat

Two previously identified forms of macrophage were investigated in primary cultures of cerebral cortical cells. Dynamic features were revealed through time-lapse video recording and aspects of macrophage function were assessed. The two cell forms were shown to be different pre-mitotic stages of a single cell type. The cell cycle for these cells involved an initial large, flat, quiescent cell which retracted to yield a slightly rounded form with numerous processes. This latter form lost processes and developed profuse filopodia as it became very rounded just prior to division; both resulting daughter cells then regained the initial large flat appearance. These cells possessed several properties of macrophages, including phagocytosis, nucleoside diphosphatase enzyme, and CR3 receptors. These properties were transient, expressed just before and after mitosis, but subsequently down-regulated in the flat daughter cells. Because of this feature, it was difficult to determine the exact size of this cell population; however, the observed rate of proliferation suggests it may be substantial. It is suggested that these cells correspond to non-microglial macrophages of brain tissue and, because of their significant down-regulation, they may be difficult to detect. This may be important in studies of brain accessory immune cells in tissue culture.

Statin expression in the untreated and SarCNU-exposed human glioma cell line, SK-MG-1

Cytokinetic analyses of gliomas and other neoplasms rely exclusively on the use of proliferation-dependent markers such as [3H]-thymidine and BuDR incorporation and the detection of growth-dependent proteins such as proliferating cell nuclear antigen (PCNA) and Ki-67. In normal tissues, the monoclonal antibody S-44 recognizes statin, a nuclear protein expressed only in nonproliferating cells. In the present study, indirect immunofluorescence microscopy using S-44 identified nuclear statin in 5.9% of a population of untreated human SK-MG-1 glioma cells in vitro. Incremental doses of the alkylating agent sarcosinamide chloroethylnitrosourea (SarCNU) induced a linear increase in the fraction of statin-positive SK-MG-1 cells. Labeling of nuclear statin with the monoclonal antibody S-44 may be a potentially useful marker of the cytotoxic effects of anticancer drugs in gliomas and other neoplastic tissues.