Extrinsic tension results in FGF-2 release, membrane permeability change, and intracellular Ca++ increase in immature cranial sutures

There are numerous studies cataloging the temporal profiles of the various growth factors during the morphogenesis of cranial sutures. There are also many clearly documented mutations of the receptors of some of these growth factors such as fibroblast growth factor (FGF)R-2 and FGFR-3 in clinical craniosynostosis. It is obvious, and often concluded, that growth factors play a role or are involved in craniofacial development. However, precisely what that role is, what causes the changes in the growth factor levels, and why these changes occur in the particular temporal and spatial patterns observed remains elusive. Using simple physics, we applied a plasma membrane disruption model and the principles of complex adaptive systems to arrive at a conjecture of calvarial morphogenesis. The purpose of this article is to introduce the concept of complex adaptive systems, to propose our conjecture, and to provide experimental proof of some key steps in this conjecture: tension induces rapid and demonstrable physiological responses in some cells within the immature cranial sutures. These responses include increases of intracellular Ca++, plasma membrane permeability, and the release of growth factors, e.g., FGF-2. Paired coronal sutures from 1-week-old Sprague-Dawley rat pups were subjected to either 0.59 N of tensile force or no force for 5 minutes in a protein-free medium. FGF-2 levels in the media were measured by slot blot analysis. Western blot analysis was used to determine FGF-2 levels in the sutures. To determine cell membrane permeability changes, fluorescein-conjugated dextran, with a molecular weight of 10 kd, was added to the media during the 5 minutes with or without tensile force. Laser confocal microscopy was used to compare the amount of entry of this impermeant tracer and the pattern of permeability change at the tissue level. To determine the intracellular pCa++, the sutures were first loaded with a calcium indictor, FURA-2 AM, and then subjected isotonically to 0.059 N of tension. The intracellular pCa++ was expressed as ratio of Ca++-bound FURA-2 to Ca++-free FURA-2. The experimental findings were as follows: 1) Sutures, in response to tension, release FGF-2. 2) Sutures contain higher levels of FGF-2 when strained. 3) There is an increase in the sutural cell membrane permeability as a result of tensile strain. 4) The cells along the leading edges of the ossification fronts (at the insertion sites of Sharpey's fibers) demonstrated the maximum permeability increase. 5) There was an immediate (within seconds) increase in intracellular Ca++. and 6) This increase in intracellular Ca++ caused by tension was reversible and independent of the extracellular Ca++ ion availability. In summary, these data support, in part, the conjecture that growth of the brain places strain on the cells within the immature sutures, which causes the iteration of a set of cellular subroutines. These subroutines integrate to generate the emergent property of directed cranial expansion with dissipation of the initiating strains.

Effects of sodium and chloride on neuronal survival after neurite transection

An in vitro investigation was undertaken to study the roles of Na+ and Cl- in mammalian spinal cord (SC) neuron deterioration and death after injury involving physical disruption of the plasma membrane. Individual SC neurons in monolayer cultures were subjected to UV laser microbeam transection of a primary dendrite. Neurons lesioned in modified ionic environments (MIEs) where 50%-75% of the NaCl was replaced with sucrose had higher survival (65%-75%) than neurons lesioned in medium with normal (125 mM) NaCl (28%; p < 0.001). Subsequent experiments found a comparable increase in lesioned neuron survival in MIEs in which only Na+ was replaced with specific ionic substitutes; however, replacement of Cl- was not protective. Electron microscope examinations of neurons fixed <16 min after lesioning showed a dramatic decrease in vesiculation of the smooth endoplasmic reticulum and Golgi apparatus in the low NaCl or low Na+ MIEs. It is hypothesized that Na+ entry after membrane disruption may stimulate elevation of [Ca+2]i leading to ultrastructural disruption and death of injured neurons. The results of these studies suggest that a low NaCl MIE may be useful as an irrigant to limit damage spread and cell death within CNS tissues during surgery or after trauma.

Inosine and guanosine preserve neuronal and glial cell viability in mouse spinal cord cultures during chemical hypoxia

Murine spinal cord primary mixed cultures were treated with the respiratory inhibitor, rotenone, to mimic hypoxic conditions. Under these conditions neurons rapidly underwent oncosis (necrosis) with a complete loss in viability occurring within 260 min; however, astrocytes, which accounted for most of the cell population, died more slowly with 50% viability occurring at 565 min. Inosine preserved both total cell and neuronal viability in a concentration-dependent manner. The time of inosine addition relative to hypoxic insult was critical with the most effective protection occurring when inosine was added just prior to or within 5 min after insult. Inosine was ineffective when added 30 min after hypoxic insult. The effect of guanosine was similar to that of inosine. Treatment of cultures with BCX-34, a purine nucleoside phosphorylase inhibitor, prevented protection by inosine or guanosine, suggesting involvement of a purine nucleoside phosphorylase in the nucleoside protective effect.

The endogenous antioxidant glutathione as a factor in the survival of physically injured mammalian spinal cord neurons

Glutathione is part of the system of cellular defenses against lipid peroxidation and other free radical-mediated damage. An established in vitro trauma model was utilized to evaluate whether glutathione is a factor in the survival of mammalian spinal cord neurons following physical injury. Cultured murine spinal neurons were subjected to a standard lesion: transection of a primary dendrite 100 microm from the perikaryon. Prior reduction of glutathione with ethacrynic acid or buthionine sulfoximine caused a dose-dependent decrease in neuronal survival 24 hours after dendrotomy. Prior glutathione augmentation with gamma-glutamylcysteine or L-2-oxo-4-thiazolidine carboxylic acid significantly increased survival, but N-acetyl-cysteine was not protective. Gamma glutamylcysteine effected the most rapid increase in glutathione (peak at 10 min), and survival was 72% +/- 10 when 0.2 mM gamma-glutamylcysteine was added immediately after dendrotomy compared with 38% +/- 4 in the control group (p < 0.0001). These results indicate that the level of glutathione is a factor in spinal cord neuron survival after physical trauma, and that glutathione augmentation may be an effective acute phase spinal cord injury (SCI) intervention strategy.

Increased jun immunoreactivity in an in vitro model of mammalian spinal neuron physical injury

Dendrites were transected from murine spinal neurons. Unlesioned neurons showed dark nucleolar and patchy cytoplasmic jun immunostaining. By 0.5 and 2 h, most lesioned neurons stained intensely throughout the soma. However, at 24 h only dead neurons displayed intense somal staining, and 100% of the surviving cells stained like unlesioned controls. Correlation of immunostaining patterns with viability, injury, and death suggests jun gene expression may influence the survival of neurons after physical injury.

The effects of ciliary neurotrophic factor on murine spinal cord neurons subjected to dendrite transection injury

Ciliary neurotrophic factor (CNTF) has been found to increase neuronal survival during development and after axotomy. The present study tested the effects of CNTF on lesioned and uninjured mouse spinal cord (SC) neurons grown in tissue culture. An initial toxicity study found that a 24-72 h exposure of SC cultures to concentrations of CNTF above 1000 ng/ml caused stress and death of unlesioned neurons and glia. Pre-selected SC neurons were then subjected to transection of a primary dendrite 100 microns from the edge of the perikaryon (approximately 50% average survival at 24 h). Application of CNTF at concentrations ranging from 0.5 to 1000 ng/ml immediately after lesioning had no statistically significant effects on SC neuron survival 24 h after dendrotomy. Separation of control (no CNTF) and CNTF-treated cells into groups of putative alpha-motor (multipolar with somal diameters > or = 25 microns) and non-alpha-motor neurons (< 25 microns somal diameters) also failed to reveal any significant differences in survival. The lack of protection by CNTF of lesioned SC neurons in mature (21-28 DIV) cultures may reflect a loss of sensitivity to CNTF that occurs with development. Alternatively, protection by CNTF may require co-factors or factors that are released from target or other cells after injury but that are not present in SC cultures.

Recurrent synovial chondromatosis treated with meniscectomy and synovectomy

Synovial chondromatosis is a rare benign intraarticular metaplasia of synovium. This process may result in the production of detached particles of highly cellular cartilage in the involved joint spaces. It is most often reported in the larger joints of the body including the knee, hip, elbow, and ankle. Since Axhausen in 1993 reported the first case affecting the temporomandibular joint, several articles have been listed in the literature regarding the presentation, diagnosis, and management of this form of an arthropathy. This is a case of a recurrent synovial chondromatosis that was approached with a meniscectomy and a complete synovectomy.

The role of excitatory amino acids in hypothermic injury to mammalian spinal cord neurons

Hypothermia has been reported to be beneficial in CNS physical injury and ischemia. We previously reported that posttraumatic cooling to 17 degrees C for 2 h increased survival of mouse spinal cord (SC) neurons subjected to physical injury (dendrite transection) but that cooling below 17 degrees C caused a lethal NMDA receptor-linked stress to both lesioned and uninjured neurons. The present study tested whether cooling below 17 degrees C increases extracellular levels of excitatory amino acids (EAA). SC cultures were placed at 10 degrees C or 37 degrees C. Glutamate (Glu) and aspartate (Asp) levels were higher in the medium of the cooled cultures after 0.5 h (23 +/- 4 nM/microgram vs. 4 +/- 1 nM/microgram and 4 +/- 1 nM/microgram vs. 1 +/- 0 nM/microgram, respectively). The concentration of each EAA then declined and reached a plateau at 2-4 h that was still significantly higher than control levels (p < 0.0001, two-factor ANOVA, three cultures per group). Other amino acids (glycine, asparagine, glutamine, serine) showed an opposite pattern, with higher levels in the 37 degrees C group. Both NMDA and non-NMDA antagonists prevented the lethal cold injury. Survival of SC neurons cooled at 10 degrees C for 2 h and rewarmed for 22 h was 58% +/- 25% in the control group, 94% +/- 5% in the CNQX-treated group, 97% +/- 5% in the DAPV-treated group, and 99% +/- 2% in the group treated with both antagonists [p < 0.0006, one factor ANOVA, five cultures (> 120 neurons) per group]. These results show that death of neurons cooled to 10 degrees C is caused by elevated extracellular Glu and Asp and requires activation of both the NMDA and non-NMDA receptor subtypes.

Reduction of NaCl increases survival of mammalian spinal neurons subjected to dendrite transection injury

Neurites were transected from spinal neurons in media with normal (125 microM) or reduced NaCl (sucrose substitution). After 12 h the normal ionic environment (conditioned medium with serum) was restored. A one-factor ANOVA comparison found a significant difference in 48 h survival (P = 0.0001). Survival was highest when NaCl was reduced 50% (74% +/- 19 vs. 22% +/- 19 in normal NaCl). Na(+)- and Cl-mediated deterioration may contribute to both gray and white matter injury in CNS trauma.

Effects of methylprednisolone on lesioned and uninjured mammalian spinal neurons: viability, ultrastructure, and network electrophysiology

An in vitro investigation was undertaken to provide information regarding the effectiveness of methylprednisolone sodium succinate (MPSS) as a treatment for the primary mechanical injury of spinal cord (SC) trauma. Exposure of uninjured mouse SC cells to MPSS for 24 h caused neuronal stress when the concentration exceeded 150 micrograms/mL; neuronal death occurred at concentrations above 600 micrograms/mL. The concentration range for MPSS protection of SC neurons subjected to a defined physical injury (laser microbeam transection of a primary dendrite 100 microns from the perikaryon) was very narrow: survival in the 30 micrograms/mL group differed significantly from the untreated control group (68.5% +/- 14.1 vs. 47.1% +/- 14.1), treatment with 20 or 60 micrograms/mL MPSS did not increase survival, and treatment with 100 micrograms/mL MPSS accelerated ultrastructural deterioration and increased the likelihood of death. Enhanced survival of lesioned neurons was observed when 30 micrograms/mL MPSS was applied within 15 min of dendrotomy but not when MPSS was administered 2 h after lesioning. Multimicroelectrode plate (MMEP) studies of SC network electrical activity indicated that MPSS associated readily with neuronal membranes. This finding was consistent with the hypothesis that MPSS may protect lesioned neurons by stabilizing damaged membranes, enhancing lesion resealing, and limiting the spread of ion-mediated damage. However, comparisons of neurite die-back 24 h after dendrotomy found no significant difference between MPSS-treated and control neurons. Application of 30 or 100 micrograms/mL MPSS increased the spontaneous burst activity of SC networks grown on MMEPs, however, there was no evidence that the increased excitability at these concentrations was the result of specific actions of MPSS on GABA or NMDA synapses.

Responsiveness to ATP with an increase in intracellular free Ca2+ is not a distinctive feature of calbindin-D28 immunoreactive neurons in myenteric ganglia

The aim of this study was to test the hypothesis that ATP elevates cytosolic free Ca2+ levels ([Ca2+]i) in myenteric neurons expressing the Ca2+ binding protein, calbindin-D28. A laser microbeam marked the location of cultured neurons on coverslips and provided unequivocal relocation of ATP-responsive neurons after immunocytochemistry. All myenteric multipolar neurons displayed ATP Ca2+ transients, and 42% also expressed calbindin-D28 reactivity. Statistical analysis of the kinetics and shape of ATP Ca2+ transients revealed no differences between calbindin and non-calbindin neurons. The identity of other responsive neurons is unknown. Less than 8% of ganglion cells with ATP Ca2+ transients were immunopositive for the glial protein S-100. We conclude that one of the actions of ATP in myenteric ganglia is to increase [Ca2+]i which may activate gKCa leading to membrane hyperpolarization in AH, Dogiel Type II neurons expressing calbindin-D28. An efficient buffering mechanism for handling large purinergic Ca2+ loads is a common feature of all types of myenteric ganglion cells.

Ultrastructural damage and neuritic beading in cold-stressed spinal neurons with comparisons to NMDA and A23187 toxicity

While exposure of cultured spinal neurons to mild hypothermia provides some protection from physical trauma (dendrotomy), profound cooling (< 17 degrees C) causes unrelated neuronal injury and death, which can be prevented by treatment with NMDA receptor antagonists. To investigate the mechanism of hypothermic neuronal injury we examined the ultrastructure of cultured spinal neurons after 2 h of cooling to 17 degrees C or 10 degrees C, with or without the presence of the NMDA receptor antagonist D-2-amino-5-phosphonovalerate, and with or without rewarming to 37 degrees C. These groups were compared to cultures exposed to NMDA or to the calcium ionophore A23187. Patterns of ultrastructural change, involving cytoskeletal disruption, mitochondrial abnormalities and vacuolization of the cytoplasm, suggest a common mechanism of injury in all treatment groups, involving an elevation of intracellular calcium. Some neurons exposed to hypothermia, NMDA or ionophore developed beaded dendrites. Microtubules were fragmented in varicosities but not in the intervening constrictions; other organelles were largely excluded from the constrictions. Varicosities may form when organelles and cytoplasm accumulate as the result of disruption of transport and membrane stabilizing proteins by proteases activated by calcium influx via NMDA mediated channels. The periodic nature of the swellings may reflect inherently discontinuous distribution of molecular subunits of the cytoskeleton.

Chronically indwelling venous cannula and automatic blood sampling system for use with nonhuman primates exposed to 60 Hz electric and magnetic fields

An automated blood sampling system was developed for use with tethered baboons (Papio cynocephalus) during concurrent exposure to 60 Hz 30 kV/m electric fields and 0.1 mT (1.0 G) magnetic fields. The system was controlled by a FORTH-based microcomputer, which operated a pump, a fraction collector, and two pinch valves. A swivel mechanism at the end of the tether allowed the baboons to move freely in their cages. The hardware and software were designed for fail-safe operation. Heparinized saline was infused at a rate of 0.5 mL/min until a sample cycle was initiated. Then, blood was drawn from the animal into a storage tube at a rate of 12.5 mL/min, a sample of undiluted blood was taken from the end of the storage tube near the baboon, and the blood remaining in the storage tube was then flushed back into the animal. Use of the storage tube prevented the peristaltic pump rollers from pressing on tubing containing blood, and return of the blood diluted with saline limited the blood wasted per sample to less than 0.5 mL. The system functioned reliably in three experiments, collecting samples as scheduled 97% of the time. Although it was initially designed for and used successfully with primates in an electric and magnetic field environment, this type of system could be employed in many areas of biomedical research or medical treatment.

A 60 Hz electric and magnetic field exposure facility for nonhuman primates: design and operational data during experiments

A unique exposure facility was designed and constructed to generate large-scale vertical electric fields (EF) of up to 65 kV/m and horizontal magnetic fields (MF) of up to 100 microT (1G), so that the behavioral and neuroendocrine effects of 60 Hz EF or combined electric and magnetic field (E/MF) exposure could be examined using nonhuman primates as subjects. Facility design and operational problems and their solutions are presented, and representative operational data from four sets of experiments are provided. A specially designed, optically isolated, 4 cm spherical-dipole EF probe and a commercially available MF probe were used to map the EF and MF within the fiberglass animal cages. In addition, amplifiers, signal conditioners, and A/D converters provided EF, MF, and transformer signals to a microcomputer at 15 min intervals. The apparatus produced homogeneous, stable E/MF at the desired intensities, and the fiberglass cages did not produce appreciable distortion or attenuation. Levels of recognized EF artifacts such as corona and ozone were negligible. The facility worked as intended, providing a well-characterized and artifact-controlled environment for experiments with baboons (Papio cynocephalus).

In vitro investigations of the effects of nonfreezing low temperatures on lesioned and uninjured mammalian spinal neurons

This two-part investigation explored the parameters and mechanisms of: (1) injury to spinal cord (SC) neurons by nonfreezing low temperatures, and (2) hypothermic protection of SC neurons subjected to a defined, physical injury (dendrite transection). Conclusions from the studies of hypothermic injury were: (1) morphologic and ultrastructural signs of stress developed in SC neurons as the temperature was decreased below 17 degrees C; (2) most neurons showing stress during cooling died upon rewarming to 37 degrees C; (3) spontaneous SC network activity was not significantly changed by cooling to 17 degrees C for 2 hours and rewarming, but cooling to 10 degrees C for 1 hour caused a reduction of burst frequency after rewarming, and cooling to 10 degrees C for 2 hours resulted in electrical silence after rewarming; and (4) application of N-methyl-D-aspartate (NMDA) antagonists before cooling prevented neuronal death, ultrastructural damage, and loss of activity upon rewarming, but application after cooling (before rewarming) was not protective. Conclusions from the studies of hypothermic protection were: (1) cooling at 17 degrees C for 2 hours followed by rewarming to 37 degrees C significantly increased lesioned neuron survival, but protection was lost when the period at 17 degrees C was increased to 6 hours; (2) NMDA blockade under normothermic (37 degrees C) or hypothermic (17 degrees C or 10 degrees C for 2 hours) conditions was not more protective of lesioned neurons than cooling to 17 degrees C (no NMDA antagonist); and (3) 200 microM thiopental or 100 microM pentobarbital increased lesioned neuron survival to a degree comparable to cooling for 2 hours at 17 degrees C.

An in vitro study of the effects of methylprednisolone on lesioned and uninjured mammalian spinal neurons

Cultured spinal neurons were subjected to dendrite amputation 100 microns from the perikaryon and treated with methylprednisolone (MP). Survival was significantly increased by 30 micrograms/ml MP but not by 10, 20 or 60 micrograms/ml. Survival was reduced by 100 micrograms/ml MP. These results suggest: (1) MP protects neurons subjected to physical trauma, and (2) the effective dose range is very narrow. These findings may have implications for MP's observed bimodal effects in spinal cord injury.

Contributions of sodium and chloride to ultrastructural damage after dendrotomy

To determine the contributions of sodium and chloride to ultrastructural changes after mechanical injury, we amputated primary dendrites of cultured mouse spinal neurons in low calcium medium in which sodium chloride had been replaced with either choline chloride or sodium isethionate or sodium propionate. Uninjured cultured neurons were also exposed to the sodium ionophore, monensin. A third set of neurons was injured in medium in which all sodium and calcium chloride had been replaced with sucrose. Neurons injured in low-calcium, low-sodium medium exhibited few ultrastructural changes, except very near the lesion, where there was some dilation of mitochondria and cisternae of the smooth endoplasmic reticulum (SER). Mitochondria in other regions of the neurons developed an electron opaque matrix, and those nearer to the lesion converted to the condensed configuration, characterized by expanded intracristal spaces as well as a dense matrix. If sodium but not chloride was present in the medium, there was some dilation of the Golgi cisternae after injury, as well as some increased electron opacity of the mitochondria. Monensin treated neurons also exhibited dilation of the Golgi cisternae. Neurons injured in sucrose-substituted medium showed none of the changes associated with injury in normal culture medium. These results indicate that sodium influx through the lesion is involved in the dilation of the SER, which is seen even in low-calcium medium, and that a permeant anion, such as chloride, is also involved. This dilation of the SER may result from uptake of calcium released from mitochondria in response to elevated cytosolic sodium. Dilation of the Golgi cisternae appears to be a response only to elevated intracellular sodium. Condensation of the mitochondria after injury is thought to be due to increased demands for ATP synthesis and may involve a "futile cycling" of calcium across the mitochondrial membrane, involving sodium-mediated calcium release in response to elevated intracellular calcium.

In vitro studies of multiple impact injury to mammalian CNS neurons: prevention of perikaryal damage and death by ketamine

We have developed an in vitro model of rapid acceleration injury (RAI) to study the effects of multiple impact (220 g/impact, 3-5 s intervals) trauma on cultures of mammalian CNS cells. Our initial investigations have shown that: (1) multiple impacts delivered tangential to the plane of growth caused neuronal death while normal impacts did not; (2) glia were not affected by tangential or normal RAI; (3) most neuronal death occurred within 15 min; (4) the threshold for neuronal death was above 440 g (cumulative); (5) neuronal death reached a maximum of 50% at cumulative accelerations greater than or equal to 1100 g; (6) somal swelling and increased nuclear prominence were often observed after tangential RAI, and the frequency of these changes increased with the cumulative acceleration; and (7) ketamine prevented neuronal death and morphological changes during tangential RAI. We hypothesize that neuronal sensitivity to multiple impact RAI depends on the density of N-methyl-D-aspartate (NMDA) complexes in the dendrosomatic membranes. We also hypothesize that the events leading to neuronal death during multiple impact injury are: (1) calcium leakage through NMDA channels causes weakening of the cytoskeleton; (2) loss of cytoskeletal integrity allows nuclear shifting during impact; and (3) nuclear pressure disrupts the plasmalemma causing a lethal influx of calcium.

Paradoxical effect of hypothermia on survival of lesioned and uninjured mammalian spinal neurons

Survival of cultured spinal neurons 2 h after dendrotomy increased significantly when the temperature was reduced to 17 degrees C or 7 degrees C. Survival of lesioned neurons maintained at 17 degrees C for 2 h and then at 37 degrees C for 22 h was still significantly higher than survival of neurons maintained at 37 degrees C for 24 h. Below 17 degrees C both lesioned and uninjured neurons swelled. Upon rewarming to 37 degrees C, many swollen neurons died. Ketamine prevented hypothermic swelling and neuronal death.

NMDA antagonists prevent hypothermic injury and death of mammalian spinal neurons

Prolonged (2-6 h) cooling of monolayer cultures of dissociated murine spinal cord at temperatures below 17 degrees C caused pronounced swelling of neuronal perikarya and dendrites. The numbers of swollen neurons in a culture increased as the temperature was reduced, and at 7 degrees C-10 degrees C all of the neurons were swollen. On rewarming the cultures to 37 degrees C, the majority of the swollen neurons died (up to 74% at 10 degrees C). Glial cells were not affected. Addition of the NMDA antagonists D-2-amino-5-phosphonovalerate (DAPV, 100 microM), ketamine (100 microM), and dibenzocyclohepteneimine (MK801, 10 microM) to spinal cord cultures before lowering the temperature to 10 degrees C minimized the dendrosomatic swelling and reduced neuronal mortality from 74% to 10%. These data show a surprising sensitivity of some neurons to nonfreezing low temperatures and suggest direct involvement of the NMDA receptor in hypothermia-related neuronal death.

Neuronal survival and dynamics of ultrastructural damage after dendrotomy in low calcium

To determine the contributions of calcium to development of ultrastructural damage and neuronal death after mechanical injury, we amputated primary dendrites from over 300 cultured mammalian spinal neurons under normal (1.8 mM) or low (less than or equal to 30 microM) calcium conditions. Two general categories of early ultrastructural change were seen in both normal and low calcium: (1) a lesion-dependent gradient of damage that moved centripetally through the proximal segment and penetrated the soma within 15 min and (2) dilation of the somal Golgi/smooth endoplasmic reticulum (SER), which preceded the wave of deterioration from the lesion. Although the somal Golgi/SER changes were similar in both normal and low calcium, the damage gradient in low calcium differed from the damage gradient in normal calcium. (1) Microtubules and neurofilaments were preserved, (2) mitochondria became more electron dense but did not develop electronlucent foci or high amplitude swelling, and (3) an extensive vesicular gradient formed consisting of rows of swollen SER vesicles. Sodium ionophores have been reported to cause similar changes. Survival studies showed that calcium reduction significantly delayed neuronal death. Survival was 63 +/- 16% vs 35 +/- 8% (p less than 0.003) at 2 h and 30 +/- 7% vs 23 +/- 8% at 6 h in low and normal calcium, respectively. Dead neurons that had been lesioned in low calcium also showed greater ultrastructural preservation than neurons that died after dendrotomy in normal calcium. We hypothesize that under low calcium conditions, the large sodium injury current plays an important role in neuronal deterioration and death after mechanical trauma.

Calcium antagonists fail to protect mammalian spinal neurons after physical injury

Most investigations of calcium antagonists as treatments for experimental spinal cord injury (SCI) have not demonstrated significant reduction of tissue damage or improvement in neurologic outcome. Many of these studies were prompted by reports that these agents increase blood flow to ischemic tissues. However, in vitro studies of renal and neuronal tissues subjected to an anoxic stress have shown that the calcium antagonists can confer direct protection on stressed parenchymal cells. We have used a tissue culture model of nerve cell injury to investigate whether calcium antagonists increase the probability of survival of spinal cord neurons after a defined physical trauma. Preliminary toxicity studies determined the maximum nontoxic dosages of verapamil (80 microM), nifedipine (10 microM), and chlorpromazine (10 microM) for neurons in our cultures. Preselected neurons (100-200 per study) were subjected to amputation of one primary dendrite at a distance of 100 microns from the perikaryon. Erythrosine B tests of viability conducted 24 h after lesioning failed to demonstrate that neurons injured in the presence of any one of these agents had an increased probability of survival compared to operated control neurons. Viability evaluations conducted 2 h after injury with phase contrast microscopy showed no evidence of slowed deterioration. Correction for other lesion physical parameters (lesion diameter and the extent of proximal segment retraction) also failed to reveal any increased protection by these agents. We conclude that calcium antagonists alone will not be useful for treatment of the primary injury of SCI.

Proximal segment retraction increases the probability of nerve cell survival after dendrite transection

Dendrites were transected 100 microns from nerve cell somata. The probability of cell survival increased as a function of the extent of proximal segment retraction. The retraction effect is probably the result of increased resistance to calcium and sodium currents at the lesion. It is speculated that limitation of nerve fiber die-back and neural tissue damage after injury may depend upon the ability of severed neurites to retract.

The sequence of ultrastructural changes in cultured neurons after dendrite transection

Cultured mouse spinal neurons were fixed at three different intervals after dendrite amputation: within the first 15 min, at 2 h and at 24 h. Dendrites were amputated at lesion distance of either 50 microns (31% probability of cell survival) or 100 microns (53% probability of cell survival) from the edge of their perikarya. When fixed within 15 min, operated neurons showed a two-phase gradient of ultrastructural damage which spread from the transection site towards the perikaryon. At 2 h after dendrite amputation all neurons operated close to their perikarya were categorized as either viable, moribund or dead, based on their appearance with phase contrast microscopy. These categories of response to physical trauma corresponded to distinctly different ultrastructural changes. Moribund neurons were filled with membrane-bound vesicles which were derived from swollen mitochondria and grossly dilated cisternae of the smooth endoplasmic reticulum. The cytoplasm of dead neurons contained large clear areas and many condensed, dark mitochondria. Both moribund and dead neurons lacked cytoskeletal elements. All of these ultrastructural changes are hypothesized to be the result of an increase in the intracellular concentrations of free calcium. Although evidence of residual mitochondrial swelling was present in some surviving neurons at 24 h, the ultrastructure of others was comparable to that of control cells. Some surviving neurons had terminal swellings at the ends of the severed neurites which were very similar to retraction balls of transected axons after CNS trauma.

Adhesion of cultured mammalian central nervous system neurons to flame-modified hydrophobic surfaces

Surface wettability is an excellent indicator of the ability of cells to adhere to a culture substrate. We have determined that brief exposure of a hydrophobic culture surface to a propane flame may increase wettability more than 1200% via the deposition of ionic combustion products. Previously nonadherent mouse spinal cord cells will adhere to and differentiate morphologically on a hydrophobic surface after flaming. Central nervous system cells remain adhered to flamed surfaces for periods of 2 mo. or longer and demonstrate spontaneous electrical activity during that time. Secondary modification of a flamed surface with polylysine further enhances the strength of single cell adhesion, thereby retarding mobility and promoting neurite extension. Flaming also enhances the wettability of common culture materials such as glass and polystyrene, as well as metal. Flaming of hydrophobic substrates through masks permits creation of discrete adhesion islands and patterns which may be used for a variety of investigations requiring maintenance of different cell types in separate regions of a culture surface.

A photoetched cell relocation matrix for long-term, quantitative observations of selected individual neurons in culture

A photoetched matrix of indium tin oxide (ITO) on glass has been developed and tested as a tool to assist in the relocation and identification of individual neuronal cells in culture. The matrix is formed by 10-15 micron wide and 300 A thick ITO lines which subdivide a 1-cm2 area into 625 smaller squares. Each of the smaller squares measures 400 micron on a side and contains a photoetched two-letter "address". The address code allows precise relocation of specific regions of a culture as well as verification of the identities of individual neurons selected for repeated observation. Marks at 50 micron intervals along the sides of the address squares permit quantitative analysis of morphological changes, cell migration, reaggregation, etc. The ITO is transparent and does not interfere with visualization of even fine details of cells with high power microscopy.

In vitro simulation of neural trauma by laser

A serious lack of knowledge about central nervous system trauma is encountered on the cellular level where the inability to create precise experimental lesions of known magnitude has limited our understanding of the reactions of single cells to injury. We used a laser cell surgery technique developed in this laboratory to manipulate neurons in a controlled environment, in order to observe pathologic reactions during and immediately after the injury. This technique is especially suited for axonal and dendritic amputations close to the perikaryon. The laser provided three different physical modes of injury to neurites: direct vaporization of cytoplasm, pressure wave damage from external vaporization of substrate material, and photobiologically-induced localized cytoskeletal destruction leading to the slow pinching of processes followed by transection. Our data indicated a great similarity between laser impact damage and the cellular damage produced by physical trauma to the central nervous system.

Neuronal survival or death after dendrite transection close to the perikaryon: correlation with electrophysiologic, morphologic, and ultrastructural changes

We investigated the probability of survival of mouse spinal neurons in monolayer cultures after transection lesions of dendrites made within 400 microns of the perikarya. Based on a total of 650 lesioned neurons, the following observations were made. First, neuronal survival is a function of lesion distance from the perikaryon and of process diameter at the lesion site. For an average lesion diameter of 3 microns, dendrite transections at 50 microns, 100 microns, and 150 microns were associated with survival probabilities of 30%, 53%, and 70%, respectively. Second, the fate of the injured cells was definitely established 24 hours after injury and very likely was determined as early as 2 hours. Third, early stages of deterioration leading to cell death were associated with cytoplasmic phase brightness on light microscopy, correlating with an appearance of numerous, small, electron-lucent vacuoles and swollen mitochondria on electron microscopy. The cytoplasm of these moribund cells stained darkly and contained no visible microtubules or neurofilaments. Fourth, the magnitude and time course of injury potentials recorded at the somata were a function of the lesion distance and did not return to prelesion levels within 30 minutes after transection. Fifth, at 24 hours after injury, the average membrane potential of lesioned neurons was 8% below that of control neurons. Sixth, at a lesion distance of approximately 300 microns both the injury potential and the probability of cell death approach zero. We conclude that, in the model system used, neuronal survival after dendrite amputation depends on physical parameters of the lesion that determine the magnitude of the injury current reaching the soma. Survival is not assured if the injury is inflicted within 250 microns of the cell body, and cell death is likely for lesions within 50 microns of the soma. The below-normal membrane potentials at 24 hours after injury suggest a possible greater vulnerability of recovering neurons to secondary insults. The characteristic mitochondrial disruption and loss of microtubules implies that the calcium component of the injury current contributes to cell death.

Laser microbeam surgery: ultrastructural changes associated with neurite transection in culture

The exposure of neuronal and glial cell processes to a large number (up to 300) of 12-nsec laser pulses at a wavelength of 337 nm and energy densities below the threshold for nonlinear absorption results in a gradual, gentle process transection in the laser focus. Within 10 to 20 sec after cessation of firing, the process pinches in the target area. During this time, mitochondria become swollen and bleached, the plasma membrane develops an obvious tautness, microtubules disappear, and organelles accumulate to either side of the process constriction. Depending on the irradiation parameters, a local pinching may proceed to a transection in about 30 sec or it may reverse to yield a normal-appearing process in approximately 5 min. Severe process pinching is accompanied by a sudden depolarization that may last for 2 to 5 min and is usually followed by a repolarization to the original resting potential even if the process has transected. Spiral retraction of cut processes and cytoplasmic spillage observed after mechanical transections are not seen with this laser method. Process stretching is minimized or eliminated. Extensive vacuolization often associated with mechanical transections does not develop unless substrate involvement in the form of shock waves is apparent. For the performance of cell surgery in culture, this method appears to offer a reliable approach to morphological alteration of single cells and to the tailoring of two-dimensional neuronal networks. It should also allow more quantitative and better-controlled studies of axonotmesis, degeneration, and regeneration on the single cell level, and it may be used as a probe for the investigation of cytoskeletal dynamics. A mechanism describing the cytoskeletal changes associated with laser-induced cell process transection is proposed.

Recording of spontaneous activity with photoetched microelectrode surfaces from mouse spinal neurons in culture

A matrix of photoetched gold conductors integrated into the floor of a tissue culture chamber has been used to record from mammalian spinal cord neurons grown on the insulation layer of the multielectrode plate. Spontaneous activity has been monitored from tissue microfragments less than 150 micrometers in diameter and from thin sheets of spinal cell aggregates. Maximum spike amplitudes of 360 microV with signal-to-noise ratios of 8:1 have so far been achieved and the spontaneous activity maintained for several days. Recording electrode impedances measured between 4 and 7 M omega at 1 kHz. Conductor tips were deinsulated with laser pulses that formed shallow craters 2 micrometers deep and 12 micrometers in diameter. Addition of colloidal gold or platimum black was not necessary to achieve satisfactory recordings.