Skip Navigation
Johns Hopkins Bloomberg School of Public HealthCAAT

Technical Report No. 6

Final Scientific Reports of CAAT Grantees


Thomas K. Baumann, PhD
Division of Neurosurgery
Department of Pharmacology
School of Medicine, L472
Oregon Health Sciences University
Portland, OR 97201


After one to 29 days in culture, the chemosensitivity of 43 adult rabbit trigeminal (primary afferent) neurons was studied using the whole-cell variant of the patch-clamp electrophysiological recording technique. Most neurons (30 out of 43 tested) responded to a brief (15 to 30 seconds) superfusion with low concentrations of the irritant resiniferatoxin (RTX, 10 to 100 nm). In different neurons, responses consisted of either a net inward, a net outward, or an early inward current followed by an outward current. Across the population, the amplitude of the whole-cell current grew with increased RTX concentration. The proportion of neurons affected by RTX increased also (8/15, 9/14, and 11/12 neurons were affected by 20, 50, and 100 nM RTX, respectively). In several neurons (n=4) rapid inward current deflections (signaling the discharge of action potentials) were superimposed on the slow currents generated by RTX. The effects of RTX were similar to those reported previously for capsaicin, but the concentrations needed to evoke comparable effects with RTX were three orders of magnitude lower. We conclude that this in vitro system (based on cultured adult rabbit trigeminal neurons) could be used to screen for the sensory effects of chemical irritants.


The long-term goal of this project is to develop a cell-culture which would allow the measurement of the excitation of nociceptive neurons by irritant chemicals. Using this model, it should be possible to predict in vitro whether or not a new ingredient in a commercial product may produce an unpleasant (painful) sensation when applied to the skin or the eye by the consumer.

The subjects of this work are primary afferent trigeminal neurons which respond to irritant chemicals. In the whole animal, these neurons contribute to the "common chemical sense". The common chemical sense (13) evolved to detect substances which are potentially harmful to the body. Using the patch-clamp electrophysiological recording technique, we show that cultured adult rabbit trigeminal neurons respond to known pain-producing irritants (capsaicin and resiniferatoxin). Thus, the in vitro model appears to present a viable alternative to in vivo testing of irritant chemicals.


Cell Culture

The method to grow adult rabbit trigeminal neurons in culture was described in detail recently (1). Briefly, adult rabbits were sacrificed by an intravenous injection of a euthanasia solution (Beuthanasia-D, Schering, 0.2 ml/kg body weight). Both trigeminal ganglia were dissected and subjected to enzymatic dissociation with trypsin and collagenase. Dissociated neurons were washed twice by centrifugation and suspended in a growth medium (L-15 with supplements, which included glucose, glutamine, a variety of amino acids and vitamins, nerve growth factor, and horse serum). Neurons were seeded on poly-ornithine and laminin coated glass coverslips held in 35 mm diameter tissue culture dishes. The cultures were maintained in a n incubator at 37° C in a humidified air atmosphere. The growth medium was refreshed by replacing one third of the volume three times a week.


Ringer solution (5 mM K) had the following composition in mM: NaCl 137.7, KCl 5.0, CaCl2 1.0, MgSO4 1.2, H3PO4 2.0, D,L-alanine 5.0, glucose 5.5, pH adjusted to 7.35 with NaOH. High potassium (25 mM) Ringer solution was prepared by substituting an appropriate amount of KCl for NaCl. Stock solution of resiniferatoxin (RTX, 1 µM) was prepared by further dilution in ethanol. The stock solutions RTX were stored at -20°C. Serial dilutions (100, 50, and 20 nM RTX) were prepared on the day of the experiment by dissolving the appropriate volume of an RTX stock solution in 5 mM K Ringer solution.

Neurophysiological Recording

Glass coverslips with neurons were placed in a recording chamber (volume ca. 0.6 ml) on the stage of an inverted microscope and perfused continuously (at a rate of approximately 1.5 ml/minute) with 5 mM K Ringer solution. Standard whole-cell recording techniques were employed (10) to measure neuronal responses to electrical and chemical stimulation. The patch electrodes were filled with a 160 mM KCL, 8.13 mM EGTA, 10 mM HEPES solution (adjusted to pH 7.25 with KOH).

Electrical Stimulation

Prior to the application of chemical stimuli to each neuron, certain electrophysicological characteristics were measured. The recording sequence was as follows. Once a whole-cell recording was established, the recording was switched to current clamp mode and the resting potential of the neuron determined. Action potentials were evoked using brief (40 ms) depolarizing current pulses. The amplitude of depolarizing current pulses progressed from 99 up to 1999 pA in increments of 100 pA. To avoid membrane damage in highly excitable neurons, stimulation was typically terminated 300 pA above threshold for action potential discharge. These measurements provided information about the threshold current and the time course of the action potential. Current pulses of longer duration (160 ms) were used to study adaptation of the action potential discharge. Responses to both brief and long duration current pulses were measured under two conditions: a) applied from the resting potential, and b) after first polarizing the neuron to -60 mV by passing a DC current. Following the measurements of action potential characteristics, the recording was switched to voltage clamp mode and the presence/absence of inward rectifier current was determined by measuring the size and time-course of inward currents evoked by hyperpolarizing voltage pulses. The pulses were 1500 ms in duration and ranged (in 5 nV steps) from +10 to -60 mV relative to a holding potential of -60 mV.

Chemical Stimulation

Following electrical stimulation, the voltage-clamp recording was maintained with the transmembrane voltage held at -60 mV and the whole-cell current monitored for the rest of the experiment. Between chemical stimuli, the neurons were continuously superfused with 5 mM K Ringer solution. Chemical stimuli consisted of a 15 second application of high potassium (25 mM) Ringer solution, followed several minutes later by one or more applications of RTX (either 10, 20, 50, or 100 nm) lasting 15 or 30 seconds each. All stimuli were applied several minutes apart to allow for recovery. Solutions were selected by a manually operated valve and supplied continuously by gravity through supply lines connected to a glass pipette whose orifice (375 µm inner diameter) was placed within 1 mm of the neuron being studied. To confirm that all chemical stimuli indeed reached the neuron, levels of the solutions in the reservoirs were monitored and the applications of RTX were bracketed by application of high potassium Ringer solution, which always evoked a sizable net inward current. Only one neuron per coverslip was studied to exclude the possibility of inadvertent desensitization by previous application of RTX.


The responses of 43 adult rabbit trigeminal (primary afferent) neurons to electrical stimulation and to a brief superfusion with low concentrations of resiniferatoxin (RTX, 10 to 100nm) were studied after one to 29 days in culture. In most neurons (25 out of 43 tested) RTX elicited a net inward current which lasted for the duration of the application of the chemical (15 or 30 seconds). In several neurons (n=4) rapid inward current deflections (signaling the discharge of action potentials) were superimposed on the slow currents generated by RTX. A net outward current, not associated with a discharge of action currents (which in one neuron followed an inward current with action current discharge) was observed in a minority of neurons (5/43) (not illustrated). Thirteen neurons did not respond to RTX.

Figure 1 illustrates the response of a trigeminal neuron to electrical and chemical stimulation. This particular neuron was slowly adapting and depolarizinga current pulse (200 pA, 160 ms duration) evoked two action potentials (Figure 1A). The threshold for electrical stimulation was relatively low (100 pA). Hyperpolarization of the cell body revealed a pronounced inward rectifier current (Figure 1B). High potassium Ringer solution excited this neuron, which responded with a net inward current and discharge of action currents (Figure 1C). RTX (50 nM) evoked an even stronger discharge of rapid action currents, superimposed on a slow net inward current (Figure 1D).

Figure 1. Response of an adult rabbit trigeminal neuron in culture to electrical and chemical stimulation

A: discharge of action potentials (top trace) in response to a current step stimulus of 160 ms duration and 200 pA amplitude (bottom trace), starting at the resting potential of -50 mV. B: Currents (top traces) evoked by voltage steps (+20 to -60 mV, superimposed on a holding potential VH=-60 mV, lower traces). C: Slow net inward current and rapid current deflections (action currents) evoked by a brief (15 s) superfusion of the nueron with 25 mM K Ringer solution (VH=-60 mV). D: Slow net inward current and rapid action currents stimulated by a brief superfusion with a regular Ringer solution containing 50 nM resiniferatoxin (RTX) (VH=-60 mV). Current-clamp recording in A, voltage-clamp recordings in B through D. The variation in the amplitudes of action currents in C and D is due to undersampling.

Across the population, the amplitude of the net inward current grew with increased RTX concentration, while outward currents remained relatively small and similar in size at all RTX concentrations used (Figure 2). The proportion of neurons responding to RTX with a net inward current increased with RTX concentrations (8/15, 9/14, and 11/12 neurons tested with 20, 50, and 100 nM RTX, respectively). Thus, there appears to be a subpopulation of trigeminal ganglion neurons which responds to RTX. The intensity of the irritant stimulus is reflected in the amplitude of the net inward current and possibly also in the proportion of neurons excited by RTX.

Figure 2. Peak amplitudes of currents produced by superfusion with different concentrations of RTX.

Dots are measurements from individual neurons. Horizontal lines indicate median amplitude of currents evoked by a given concentration of RTX.


The present study shows that cultured adult rabbit trigeminal neurons are excited by the pain-producing chemical resiniferatoxin (RTX). The responses to RTX appear similar to those observed in these neurons in a previous study using capsaicin (4). Compared with capsaicin, the concentrations of RTX needed to achieve analogous effects are about three orders of magnitude lower. This agrees with results obtained in cultured rat dorsal root ganglion (DRG) neurons (17). In functional tests in vivo, RTX was found to be two times to several thousand times more potent than capsaicin, depending on the assay used (14). We conclude that cultured adult rabbit trigeminal neurons could be used to screen for the sensory effects of irritants acting through the capsaicin receptor (15).

In the intact organism, different somatosensory modalities (the sensations of touch, vibration, flutter, cold, warmth, and mechano-, thermo- and chemonociception) are served by different subsets of primary afferent neurons (5, 9, 11, 16). Intracellular recordings, obtained from some of these subsets, revealed that the membrane properties are different in different subsets (8, 12). Whole-cell recordings from adult rabbit trigeminal ganglion neurons in culture revealed that their electrophysiological properties are also not uniform (2, 3). The neurons differ in the shape of their action potentials, the speed of adaptation to depolarizing stimuli, and in the strength of inward rectification evoked by hyperpolarizing stimuli. Thus, cultured trigeminal neurons maintain in vitro at least some of the differentiated characteristics observed in DRG neurons in vivo. It remains to be shown whether a given set of electrophysicological characteristics is tied to chemosensitivity of nociceptive primary afferent neurons both in vivo and in vitro.

None of the currently available alternative tests (see 6, 7 for a review) is able to assess the sensory aspects of an exposure to a chemical. Based on the present experiments, it appears that cultures of adult rabbit trigeminal neurons may bridge this gap. Once fully validated, the sensory nerve culture model will offer a powerful alternative to in vivo irritant screening exemplified by the Draize test and thus substantially reduce the discomfort and the number of experimental animals required to screen potential irritants. It has orders of magnitude, higher temporal resolution, and requires considerably less tissue than the biochemical techniques currently available to measure cell damage or irritation.


  1. Baumann, T.K. (1993). Cultures of adult trigeminal ganglion neurons. In Vitro Biological Models Vol. 1, Methods in Toxicology (eds. Tyson, C.A. and Frazier, J.M.) New York: Academic Press (in press).
  2. Baumann, T.K., and R.H LaMotte. (1988). Electrophysiological properties of adult rabbit trigeminal neurons in dissociated cell culture. Society of Neuroscience Abstracts 14: 560.
  3. Baumann, T.K., and R.H. LaMotte (1989). Electrophysiological properties and chemosensitivity of cultured rabbit trigeminal ganglion neurons. Society of Neuroscience Abstracts 15: 440.
  4. Baumann, T.K. and R.H. LaMotte (1989). Whole-cell patch-clamp recordings of the responses of cultured rabbit trigeminal ganglion neurons to capsaicin - an irritant and neurotoxin. In In Vitro Toxicology, New Directions. Alternative Methods in Toxicology Vol. 7 (ed. Goldberg A.M.), Mary Ann Liebert, Inc.: New York, pp. 57-66.
  5. Baumann, T.K., Simone, D.A., Shain, C., and R.H. LaMotte (1991). Neurogenic hyperalgesia: the search for the primary cutaneous afferent fibers that contribute to capsaicin-induced pain and hyperalgesia. Journal of Neurophysiology 66: 212-227.
  6. Frazier, J.M., Gad, S.C., Goldberg, A.M., McCulley, J.P., and D.R. Meyer (1987). A critical evaluation of alternatives to acute ocular. In Critical Evaluation of Alternatives to Acute Ocular Irritation Testing. Alternative Methods in Toxicology Vol. 4 (ed. Goldberg, A.M.) Mary Ann Liebert, Inc., NY, pp. 1-4.
  7. Gad, S.C. (1989). Acute ocular irritation evaluation: in vivo and in vitro alternatives and making them "standard" for testing. In Benchmarks: Alternative Methods in Toxicology (ed. Mehlman, M.A.) Princeton Scientific Publ. Co., Princeton.
  8. Koerber, H.R., Druzinsky, R.E., and L.M. Mendell(1988). Properties of somata of spinal dorsal root ganglion cells differ according to peripheral receptor innervated. Journal of Neurophysiology 60: 1584-1596.
  9. LaMotte, R.H., and J.N. Campbell, (1978). Comparison of responses of warm and nociceptive C-fiber afferents in monkey with human judgements of thermal pain. Journal of Neurophysiology 41: 509-528.
  10. Marty A., Neher, E. (1983). Tight-seal whole-cell recording. In Single Channel Recording (eds. Sakmann, B. and E. Neher) Plenum Press, New York: pp. 107-122.
  11. Perl, E.R. (1984). Pain and nociception. Handbook of Physiology 3:915-975.
  12. Rose, R.D., Koerber, H.R., Sedivec, M.J., and Mendell, L.M. (1986). Somal action potential duration differs in identified primary afferents. Neuroscience Letters 63: 259-264.
  13. Silver, W.L. (1987). The common chemical sense. In Neurobiology of Taste and Smell (eds. Finger, T.E. and W.L. Silver) New York: Wiley, pp. 65-87.
  14. Szallasi, A. and P.M. Blumberg (1989). Resiniferatoxin, a phorbol-related diterpene, acts as an ultrapotent analog to capsaiscin, the irritant constituent in red pepper. Neuroscience 30: 515-520.
  15. Szallasi, A. and P.M. Blumberg (1990). Specific binding of resiniferatoxin, an ultrapotent capsaicin analog, by dorsal root ganglion membranes. Brain Research 524: 106-111.
  16. Willis, W.D. and R.E. Coggeshall (1978). Sensory mechanisms in the spinal cord. New York: Plenum.
  17. Winter, J., Dray, A., Wood, J.N., Yeats, J.C., and S. Bevan, (1990). Cellular mechanism of action of resiniferatoxin: a potent sensory neuron neurotoxin. Brain Research 520: 131-140.


Jan Krysztof Blusztajn1,2, Brygida Berse5, Maximilian Follettie6, Amy Venturini1, Darrell A. Jackson1, Ulrike Schüler1, Henry J. Lee3 and Bruce Wainer3,4.

Departments of 1Pathology and 2Psychiatry, Boston University School of Medicine, 85 East Newton Street, Room M1009, Boston, MA 02118-2394;
Departments of 3Pharmacological and Physiological Sciences, and 4Pathology, The University of Chicago, Chicago, Il;
5Department of Pathology, Beth Israel Hospital, Boston, MA;
6Department of Biology, Massachusetts Institute of Technology, Cambridge, MA.


In order to study differentiated properties of brain cholinergic neurons using long term culture, we have developed a cell line (SN56.B5.G4) derived from fusion of mouse postnatal day 21 septal neurons with murine neuroblastoma cells, N18TG2. We found that these cells exhibit several features of cholinergic neurons including choline acetyltransferase (CAT) activity, sodium-dependent-high-affinity uptake of choline, and depolarization-evoked acetylcholine (ACh) release. In cells treated with 1 mM of a cAMP analog, N6,O2'-dibutyryl-adenosine-3'-5'-cyclic monophosphate (dbcAMP) or with 10 µM forskolin (an activator of adenylyl cyclase) for two days, the activity of CAT and ACh content were two-to-three-fold higher relative to controls. CAT activity and ACh content were also elevated up to four-fold in cells treated with 1 µM of all trans retinoic acid (RA) for two days. These effects were time- and dose-dependent. The EC50 values for dbcAMP, forskolin, and RA were 1.3 mM, 0.7 µM, and 10 nM, respectively. The effects of tRA and forskolin were additive resulting in a five-fold increase in ACh content in cells treated with 1 µM RA and 10 µM forskolin for two days, relative to controls. Northern blot analysis revealed that CAT mRNA was more abundant in cells treated with 10 µM forskolin or 1 µM RA relative to controls, suggesting that these treatments may stimulate the CAT gene transcription. The enhancement of ACh synthesis by agents which increase intracellular cAMP levels and by RA in SN56.B5.G4 cells suggests that ACh synthesis in vivo may be regulated by: 1) activation of receptors for neurotransmitters, hormones, or growth factors which activate adenyl cyclase; and 2) activation of the retinoid receptors. The data also indicate that RA and elevated intracellular cAMP levels stimulate ACh production by two different mechanisms. Taken together the data demonstrate that SN56.B5.G4 cells are a useful model to study the differentiated features of the cholinergic phenotype and to test the effects of drugs and toxic agents on cholinergic neurons.


Degeneration and/or malfunction of cholinergic neurons underlies the pathophysiology of Motor Neuron Disorders, Familial Dysautonomias, Alzheimer's Disease (AD), Tardive Dyskinesia, and Huntington's Chorea. Crucial to the development of experimental approaches to study these diseases, and to design treatment strategies, is the establishment of a homogeneous cell preparation which expresses all aspects of the cholinergic phenotype. Such cell preparations will not only provide alternatives to research on experimental animals but will also be models for investigating how cholinergic neurons might behave in a defined environment. Cholinergic cell lines offer such a model system because they are homogeneous, and permit easy analysis of a variety of treatments in a well-controlled environment. We have developed cell lines derived from fusion of the murine neuroblastoma cells, N18TG2 (which lack cholinergic markers), with postnatal day 21 mouse brain septal neurons (1). Here we summarize some features of one such cell line, SN56.B5.G4, and show that these properties are similar to those characteristic of septal neurons (2). The cholinergic properties of SN56.B5.G4 cells are strikingly sensitive to pharmacological agents, and thus we expect that they will become a useful model for testing effects of drugs, both beneficial and toxic, on cholinergic neurons.

Materials and Methods

Cell Culture

The SN56.B5.G4 cells were created by fusing N18TG2 mouse neuroblastoma cells with murine (strain C57BL/6) neurons from postnatal day 21 septa (1, 3). We grow the cells at 37°C in an atmosphere of 95% air, 5% CO2 in Dulbecco's Modified Eagle's Medium (DMEM) containing 10% fetal bovine serum (FBS), and 50 µg/ml gentamicin.

ACh Accumulation

To measure [14C]ACh accumulation, the cells were incubated at 37°C in a physiological salt solution (containing in mM: NaCL, 135; KCL, 5; CaCl2, 1; MgCl2, 0.75; glucose, 5; eserine, 0.015; HEPES, 10; pH 7.4) in the presence of [14C]ACh synthesized by the cells extracted, purified by HPLC (4, 5) and its radioactivity was determined.

ACh and CAT Activity Measurements

ACh was determined by HPLC with an enzymatic reactor containing aceytlcholinesterase and choline oxidase and an electrochemical detector using a commercial kit (Bioanalytical Systems Inc., West Lafayette, IN) based on the method of Petter et al (6). Choline acetyltransferase activity was determined in cell homogenates by the method of Fonnum (7).

ACh release

To measure [14C]ACh release, the cells were incubated for 180 minutes at 37°C in L-15 medium containing 10 µM [14C]choline and 15 µM eserine. The cells were washed with L-15 medium (as above), and then incubated for an additional 30 minutes in a physiological salt solution (composition as above) and either 5 mM (control) or 40 mM potassium chloride (the concentration of sodium chloride was reduced to 100 mM). The media were collected and [14C]ACh released from the cells was purified by HPLC (4, 5), and its radioactivity determined.

Analysis of Choline Acetyltransferase mRNA

RNA was isolated from cells by the guanidinium isothiocyanate/phenol method, and 30 µg of total cellular RNA was fractionated on 1% formaldehyde/agarose gels and blotted on Zeta-Probe GT membrane. Hybridization was carried out for 36 hours at 42°C in 50% formamide, 5x SSPE, 5x Denhardt's solution, 1% SDS, 10% dextran sulfate, and 0.5 mg/ml denatured herring sperm DNA. 0.87 kb fragment of muring cDNA corresponding to the 3' end of CAT coding region (8) was used as a probe. Final washes were in 0.2x SSC, 0.1% SDS at 42°C.


SN56.B5.G4 Cells Extend Neurites

SN56.B5.G4 cells grown in basal medium extended few neurites. Since the analogs of the second messenger, cyclic AMP (cAMP), have been shown to cause neurite outgrowth in several muring neuroblastoma cell lines (9), rat pheochromocytoma cells (PC12) (10), and the neuroblastoma x glioma hybrid cells (NG108-15) (11), we added 1 mM N6,O2'-dibutyryl-adenosine-3'-5'-cyclic monophosphate (dbcAMP), a cell permeant analog of cAMP, or 10 µM forskolin, an activator of adenyl cyclase, to the medium. Cells were treated with these new drugs, divided slowly, and developed a network of neurites. Since the dbcAMP molecule can be hydrolyzed liberating free butyric acid, we test the effect of 2 mM butyrate in our cultures. Butyrate-treated cells were rounder than controls, and few neurites were observed.

SN56.B5.G4 Cells Synthesize ACh from Choline Taken Up by a Sodium-Dependent High-Affinity Transport

In the initial step of Ach synthesis in nerve endings, choline is taken up from the extracellular space by a sodium-dependent high-affinity uptake system (SDHACU) (12). We determine the apparent affinity for choline of the ACh synthetic process by incubating the cells for 10 minutes in a medium of varying [14C]choline concentration and measuring [14C]ACh accumulation. [14C]ACh accumulation was saturable with choline and exhibited an apparent Km of 4.6 µM, i.e. in the range characteristic of SDHACU. The uptake was sodium dependent; when cells were incubated in medium in which sodium was replaced by lithium, accumulation of [14C]ACh from 1 µM [14choline was diminished to 29% of control. Others obtained similar results using primary cultures of rat septum (13, 14). These data indicate that SN56.B5.G4 cells express SDHACU, and that their ACh is synthesized from choline taken up by this system.

SN56.B5.G4 Cells Release ACh upon Depolarization

Non-differentiated SN56.B5.G4 cells prelabeled with 10 µM [14C]choline and then incubated for an additional 30 minutes in a physiological salt solution containing either 5 or 40 mM K+ released little [14C]ACh. However, when the SN56.B5.G4 cells were grown in the presence of 1 mM dbcAMP, 10 µM forskolin, or 2 mM butyrate for 48 hours ACh release was reliably observed and this release was enhanced by depolarization. The spontaneous and the depolarization-evoked ACh release occurred both in neurite-free (butyrate-treated) and neurite-bearing (dbcAMP-, or forskolin-treated) cells. These data demonstrate that differentiated SN56.B5.G4 cells are capable of depolarization-evoked ACh. It would be important to determine what components of the ACh releasing mechanism are missing in undifferentiated cells.

ACh Synthesis in SN56.B5.G4 Cells Is Enhanced by Pharmacologic Agents

In cells treated with 1 mM dbcAMP or with 10 µM forskolin for two days, the activity of CAT and ACh content were two-to-three-fold higher relative to controls. CAT activity and ACh content were also elevated up to four-fold in cells treated with 1 µM of all trans retinoic acid (RA) for two days. These effects were time- and dose-dependent. The EC50 values for dbcAMP, forskolin and RA were 1.3 mM, 0.7 µM, and 10 nM, respectively. The effects of RA and forskolin were additive resulting in a five-fold increase in ACh content in cells treated with 1 µM RA and 10 µM forskolin for two days, relative to controls. One reason for the elevated CAT activity in cells treated with forskolin or with RA would be increased rate of CAT gene transcription. Evidence obtained with Northern blot analysis of the RNA purified from cells treated for two days with 10 µM forskolin or with 1 µM RA are consistent with this prediction. CAT mRNA content was higher in the treated cells relative to controls.


The SN56.B5.G4 cells have been selected from other septal lines based on CAT activity. However in order to serve as a useful model of brain cholinergic neurons it was important to establish whether these cells exhibit other features of the cholinergic phenotype. The ACh content of these cells is similar to NS20 neuroblastoma cells (2 nmol/mg protein) (15), but lower than that of the human neuroblastoma LA-N-2 cells grown in a similar medium (approximately 10 nmol/mg protein) (5). By comparison ACh content of rat striatum is 0.3 nmol/mg protein (16) and that of purely cholinergic synaptosomes from Torpedo electric organ is 130 nmol/mg protein (17). In addition to CAT activity, SN56.B5.G4 cells express SDHACU, a property which sets them apart from a variety of CAT-expressing cell lines including NS20 neuroblastoma cells (18), NG108-15 neuroblastoma cells x glioma cells (19), PC12 pheochromocytoma cells (20), and LA-N-2 neuroblastoma cells (5), all of which synthesize ACh from choline taken up by the low-affinity carrier. Thus the SN56.B5.G4 cells resemble septal neurons, which are capable of expressing SDHACU in primary cultures (13).

SN56.B5.G4 cells grown in basal medium fail to release ACh reliably. However, differentiated cells, released ACh and thus release was almost doubled by depolarization. The permissive effect of dbcAMP, butyrate, and forskolin on ACh release in these cells may be due to either differentiation of the excitable properties of cell membranes, including expression of specific ion channels, or differentiation of ACh release mechanisms such as vesicular storage of ACh, or to synthesis of proteins involved in vesicular release.

The ability to release ACh in dbcAMP-treated cells acoompanyied neurite outgrowth and stimulation of CAT activity and ACh synthesis. The latter effect of dbcAMP was maximal after two days of treatment suggesting that it was mediated by changes in CAT gene expression, translation, or CAT protein turnover or that ChAT was activated by a factor (perhaps an enzyme which modifies ChAT), whose expression required two days to develop fully. If these effects of dbcAMP were due to the cAMP moiety of this molecule, then cells treated with forskolin, which activates the adenyl cyclase and increases the intracellular cAMP concentration, should respond similarly. Consistent with this prediction the forskolin-treated (10 µM; two days) cells developed neurites and had CAT activity similar to that of dbcAMP-treated cells, and three-fold higher than the controls. The molecule of dbcAMP permeates into cells due to its butyrate moieties. Hydrolysis of dbcAMP yields free butyrate which has been shown to stimulate CAT activity in neuroblastoma cells (9, 21, 22). In SN56.B5.G4 cells butyrate caused elevations in CAT activity and stimulated ACh synthesis. The latter effect may be due to the conversion of butyrate to acetylCoA necessary for ACh synthesis (23).

CAT activity and Ach synthesis and storage are elevated in SN56.B5.G4 cells treated with RA and/or forskolin. These effects appear to be caused by an increased abundance of CAT mRNA in cells treated with these agents. This could be due to stimulation of CAT gene transcription or to a slowed degradation of the CAT message. The former possibility seems likely because most of the actions of RA are due to stimulation of transcription via the retinoid response elements. It would be important to determine whether these regulatory elements are present in the CAT promoter region. However, the idea that CAT message is stabilized by tRA is consistent with the work of Casper and Davies (24) who shoed that a-amanitin, an inhibitor of transcription, did not abolish the stimulatory effects of RA on CAT activity in a human neuroblastoma cell line. Further, these authors observed that inhibitors of translation failed to attenuate the elevations in CAT activity caused by RA, and concluded that RA caused CAT activation by a posttranslational process. Although posttranslational modifications of CAT protein may partly account for the mechanism of the effect of RA in our study, it is clear that we must also consider the elevation in CAT mRNA levels.

Stimulation of CAT activity by RA was not seen in a study of primary spetal neurons isolated from embryonic rat brain performed by Knusel and Hefti (25). SN56.B5.G4 cells are derived from postnatal day 21 mice. It is possible that the responsiveness to RA (e.g. the expression of retinoid receptors) is developmentally regulated and is not yet present in the embryonic cholinergic neurons studied by Knusel and Hefti but that it occurs in the parental primary neurons of our cell line. It should be possible to answer this question using in situ hybridization techniques with a retinoid receptor probe.

The enhancement of ACh synthesis by agents which increase intracellular cAMP levels and by RNA in SN56.B5.G4 cells suggest that ACh synthesis in vivo may be regulated by: 1) activation of receptors for neurotransmitters, hormones, or growth factors which activate adenylyl cyclase; and 2) activation of the retinoid receptors. The data also indicate that RA and elevated intracellular cAMP levels stimulate ACh production by two different mechanisms.

Taken together, the data presented above show that SN56.B5.G4 cells are characterized by ACh synthesis and storage, SDHACU, and depolarization-evoked ACh release. These are properties characteristic of the cholinergic phenotype (26). Treatment with dbcAMP causes both morphological and neurochemical differentiation. It will be important to determine whether physiologically relevant agents alter the cholinergic phenotype. The list of such molecules includes: nerve growth factor (27), basic fibroblast growth factor (28), ciliary neurotropic factor (29), CAT development factor (30), cholinergic differentiation factor (31) or leukemia inhibitory factor (32), membrane-derived factor (33), and target-derived neuronal cholinergic differentiation factor (34). It is worth noting that granulocyte-macrophage colony-stimulating factor (35) and interleukin-3 (36) have been reported to stimulate CAT activity in spetal neurons as well as in one of our cell lines (SN6.10.2.2) derived from embryonic septum, indicating the molecular mechanisms of action of these and other growth and differentiating factors on the cholinergic phenotype.


Despite the fact that cholinergic function has been studied for well over half a century, to this day there is no experimental system which allows one to investigate the mechanisms of ACh synthesis and release under long term in vitro conditions. Our knowledge of these mechanisms derives from work on intact animals or from in vitro studies using freshly-obtained preparations. Animals are necessary for both of these approaches. Recently several laboratories (22, 24, 37, 38), including ours (1, 3, 5, 39), describing studies on cholinergic cell lines in long-term culture. The initial data, some of which are summarized in this article, are promising, and they warrant further investigation of these models.


Supported by: NSF BNS8808942; NS25787; MH46095; and Center for Alternatives to Animal Testing grants.


  1. Lee, H.J., Hammond, D.N., et al (1990). Immortalized young adult neurons from the septal region: Generation and characterization. Developmental Brain Research 52: 219-228.
  2. Blusztajn, J.K., Venturini, A., et al (1992). Acetylcholine synthesis and release is enhanced by dibutyryl cyclic AMP in a neuronal cell line derived from mouse septum. Journal of Neuroscience 12: 793-799.
  3. Hammond, D.N., Lee, H.J., et al (1990). Development and characterization of clonal cell lines derived from spetal cholinergic neurons. Brain Research 512: 190-200.
  4. Liscovitch, M., Freese, A., et al (1985). High-performance liquid chromatography of water-soluble choline metabolites. Analytical Biochemistry 151: 182-187.
  5. Richardson, U.I., Liscovitch, M., et al (1989). Acetylcholine synthesis and secretion by LA-N-2 human neuroblastoma cells. Brain Research 476: 323-331.
  6. Potter, P.E., Meek, J.L., et al (1983). Acetylcholine and choline in neuronal tissue measured by HPLC with electrochemical detection. Journal of Neurochemistry 41: 188-194.
  7. Fonnum, F. (1975). A rapid radiochemical method for the determination of choline acetyltransferase. Journal of Neurochemistry 24: 407-409.
  8. Ishii, K., Oda, Y., et al (1990). Complementary DNAs for choline acetyltransferase from spinal cords of rat and mouse: Nucleotide sequences, expression in mammalian cells, and in situ hybridization. Molecular Brain Research 7: 151-159.
  9. Prasad, K.N. and S. Kumar (1974). Cyclic AMP and the differentiation of neuroblastoma cells in culture. Control of Proliferation in Animal Cells (eds. Clarkson, B. and R. Baserga) Cold Spring Harbor Laboratory: 581-594.
  10. Green, L.A. and A.S. Tischler (1976). Establishment of noradrenergic clonal line of rat adrenal pheochromocytoma cells which respond to NGF. Proceedings of the National Academy of Science USA 73: 2424-2428.
  11. Daniels, M.P. and B. Hamprecht (1974). The ultrastructure of neuroblastoma glioma somatic cell hybrids. Expression of neuronal characteristics stimulated by dibutyryl adenosine 3',5' cyclic monophosphate. Journal of Cell Biology 63: 691-699.
  12. Suszkiw, J.B. and G. Pilar (1976) Selective localization of a high affinity choline uptake system and its role in ACh formation in cholinergic nerve terminals. Journal of Neurochemistry 26: 1133-1138.
  13. Keller, F., Rimvall, K. et al (1987). Choline and acetylcholine metabolism in slice cultures of the newborn rat septum. Brain Research 405: 305-312.
  14. Bostwick, J.R., Landers, D.W., et al (1989). Purification and characterization of a central cholinergic enhancing factor from rat brain: its identity as phosphoethanolamine. Journal of Neurochemistry 53: 448-458.
  15. Kato, A.C., Lefresne, P., et al (1977). Choline stimulates the synthesis of acetylcholine from acatate and the accumulation of acetate in a cholinergic neuroblastoma clone. Biochemical and Biophysical Research Communications 78: 350-356.
  16. Cohen, E.L. and R.J. Wurtman (1976). Brain acetylcholine: control by dietary choline. Science 191: 561-562.
  17. Morel, N., Israel, M., et al (1977). Isolation of pure cholinergic nerve endings from Torpedo electric organ. Journal of Cell Biology 85: 43-55.
  18. Lank, K., Somers, L., et al (1974). Choline transport by neuroblastoma cells in tissue culture. Nature 252: 476-478.
  19. McGee, R., Jr. (1980). Choline uptake by the neuroblastoma x glioma hybrid, NG108-15. Journal of Neurochemistry 35: 829-837.
  20. Melega, W.P. and B.D. Howard (1981). Choline and acetylcholine metabolism in PC12 secretory cells. Biochemistry 20: 4477-4483.
  21. Szutowicz, A., Morrison, M.R. et al (1983). The enzymes of acetyl-CoA metabolism in differentiating cholinergic (S-20) and noncholinergic (NIE-115) neuroblastoma cells. Journal of Neurochemistry 40: 1664-1670.
  22. Casper, D., Davies, P. (1989). Stimulation of choline acetyltransferase activity by retinoic acid and sodium butyrate in a cultured human neuroblastoma. Brain Research 478: 74-84.
  23. Bielarczyk, H. and A. Szutowicz (1989). Evidence for the regulatory function of synaptoplasmic acetyl-CoA in synthesis in nerve endings. Biochemical Journal 262: 377-380.
  24. Casper, D. and P. Davies (1989). Mechanism of activation of choline acetyltransferase in a human neuroblastoma cell line. Brain Research 478: 85-94.
  25. Knusel, B. and F. Hefti (1988). Development of cholinergic pedunculopontine neurons in vitro: Comparison with cholinergic septal cells and response to nerve growth factor, ciliary neuronotrophic factor, and retinoic acid. Journal of Neuroscience Research 21: 365-375.
  26. Blusztajn, J.K. and R.J. Wurtman (1983). Choline and cholinergic neurons. Science 221: 614-620.
  27. Hefti, F., Hartikka, J., et al (1985). Nerve growth factor increases choline acetyltransferase but not survival or fiber outgrowth of cultured fetal septal cholinergic neurons. Neuroscience 14: 55-68.
  28. Vaca, K., Stewart., S.S. et al (1989). Identification of basic fibroblast growth factor as a cholinergic growth factor from human muscle. Journal of Neuroscience Research 23: 55-63.
  29. Saadat, S., Sendtner, M., et al (1989). Ciliary neurotrophic factor induces cholinergic differentiation of rat sympathetic neurons in culture. Journal of Cell Biology 108: 1807-1816.
  30. McManamam, J.L., Crawford, F.G., et al (1988). Purification of a skeletal muscle polypeptide which stimulates choline acetyltransferase activity in cultured spinal cord neurons. Journal of Biological Chemistry 263: 5890-5897.
  31. Fukada, K. (1985). Purification and partial characterization of a cholinergic neuronal differentiation factor. Proceedings of the National Academy of Science USA 82: 8795-8799.
  32. Yamamori, T., Fukada, K., et al (1989). The cholinergic neuronal differentiation factor from heart cells is identical to leukemia inhibitory factor. Science 246: 1412-1416.
  33. Adler, J.E., Schleifer, L.S., et al (1989). Partial purification and characterization of a membrane-derived factor regulating neurotransmitter phenotypic expression. Proceedings of the National Academy of Science USA 86: 1080-1083.
  34. Rao, M.S. and S.C. Landis (1990). Characterization of a target-derived neuronal cholinergic differentiation factor. Neuron 5: 899-910.
  35. Kamegai, K., Konishi, Y., et al (1990). Trophic effect of granulocyte-macrophage colony-stimulating factor on central cholinergic neurons in vitro. Brain Research 532: 323-325.
  36. Kamegai, M., Niijima, K., et al (1990). Interleukin-3 as a trophic factor for central cholinergic neurons in vitro and in vivo. Neuron 4: 429-436.
  37. Cashman, N.R., Boulet, S., et al. (1987). Clonal cell lines from neuroblastoma-spinal cord cell hybridization. Society of Neuroscience Abstracts 420: 7.
  38. Casper, D. and P. Davies (1988). Regulation of choline acetyltransferase activity by cell density in a cultured human neuroblastoma cell line. Developmental Neuroscience 10: 245-255.
  39. Blusztajn, J.K., Liscovitch, M. et al (1987). Phosphatidylcholine as a precursor of choline for acetylcholine synthesis. Journal of Neural Transmission [Suppl] 24: 247-259.


.E. Pfeiffer, PhD and A.E. Warrington, PhD
University of Connecticut School of Medicine
Department of Microbiology
Farmington, CT 06030


The oligodendrocyte (OL) is responsible for the formation of the myelin sheaths of the central nervous system. OLs are often damaged during the progression of inflammatory diseases of the brain, particularly multiple sclerosis. Understanding OL development in normal, demyelinating, and remyelinating conditions should provide insights into possible treatment and reversal of clinical demyelination.

OLs develop within the context of microenvironmental, hormonal, and histiotypic aspects present in the intact brain. To better study OL development within a complex, multicell type environment, we have laid the framework for a brain slice explant system. We have developed protocols to analyze OLs in thick tissue slices and have established baseline parameters of normal in vivo OL development with which to compare future in vitro results.


  1. A novel immunocytochemical staining protocol, employing slices of unfixed brain, was developed to analyze early stages of the OL lineage in situ identified by several specific anti-glycolipids (O4, R-mAb, 01);
  2. The thick slice staining protocol was extended to fix permeabilized brain slices to identify internal OL-specific antigens;
  3. The confocal microscope was utilized to non-invasively, optically thin-section of the brain, providing a new perspective of OL morphology in vivo. These techniques were applied to;
  4. Analyze the proliferation, migration, and differentiation of OL progenitors in situ in normal postnatal rat cerebellum;
  5. To assess the myelinogenic capacity of phenotypically defined populations of normal rat OL progenitors after transplantation into hypomyelinated mutant mice.

Several distinct phenotypic stages of the OL lineage were identified in vivo: morphologically simple O4+-mAb- blast cells localized at the leading edge of myelination; morphologically more complex, postmitotic R-mAb+O1- cells found more proximal to the white matter; and mature actively myelinating O1+ OLs residing within the white matter. A spatial comparison of these cell populations demonstrated a progressive wave of OL maturation beginning at the base of the cerebellum and moving toward the distal folia. At any given time in normal brain development, OL antigen expression, morphology, and proliferative capacity are related to the cells' position relative to the germinal zones.

The OL lineage prior to O1 recognition has been further subdivided. O4+O1- progenitors cells retain significant proliferative capacity in normal development, but the switch from the O4- to the O4+ phenotype results in a dramatic loss of migratory ability, consistent with the complex morphology of O4+ cells. Our transplantation experiments demonstrated that O4+O1- cells do not migrate significantly from the sites of injection, while phenotypically less mature O4- progenitors migrate and form significant amounts of myelin in the recipient environment.

In summary, we have significantly advanced the study of OL development in vivo, a necessary prelude to exploring OL development in a more experimentally accessible and biologically diverse explant slice system. An explant system that successfully mimics aspects of normal in vivo development will greatly increase the efficiency of animal use in the neurosciences.


Substantial information regarding OL progenitor development has been garnered from immunocytochemical studies both in vivo (1, 2, 3) and in model culture systems, especially those initiated from dissociated brain and optic nerve (4, 5, 6, 7). Nevertheless, important aspects of OL development in vivo are largely unexplored, especially as regards progenitor proliferation, migration, and differentiation as a function of the microenvironments encountered during maturation. OLs will quickly differentiate when cultured in the absence of other cell types (6), yet in vivo OL differentiation is regionally distinct and delayed (2, 3, 8). Thus, it appears that the default mechanism of OLs is to fully differentiate and is delayed only in the presence of neurons (9) and growth factors (10, 11, 12). Because of the demonstrated differences between developmental phenomena in the intact organism compared to dissociated cell culture, analyses in vivo become critical to our understanding of normal OL differentiation.

Transplantation of normal embryonic or neonatal CNS brain fragments or cell suspensions into lesioned (13, 14) or myelin-mutant animals (15) is an important tool for investigating the complex interactions of OLs in normal development and during remyelination. Some cells have the capacity to migrate sizeable distances and to myelinate bare axons in both newborn (16, 17, 18) and adult (19) hosts. Such transplantation procedures require numerous animals since each animal provides only one experimental time point.

We propose that a brain slice explant system may be substituted for whole animals to carry out studies of both normal in vivo development and following transplantation. Explants preserve many of the microenvironmental, hormonal, and histiotypic aspects of the whole brain, are readily observed and manipulated, and can be sliced so as to expose the brain areas of interest. An expected advantage is the more efficient use of laboratory animals; that is, each animal would now provide numerous samples limited only by the number of explant slices prepared from each tissue. In addition, chronic experimental manipulations are replaced by acute sacrifices.

Our research, in part funded by the Johns Hopkins CAAT, has resulted in three manuscripts in press (8, 17, 18).

Materials and Methods

In the progression of our research we have developed a novel immunocytochemical staining protocol to utilize the anti-glycolipid antibodies O4 (20, 21), R-mAb (21-22), and O1 (20, 21) to identify early stages of the OL lineage in situ. Previous reports used similar antibodies on fixed cryosections, but the resulting morphology was compromised due to poor antibody binding and specificity (23). We have stained unfixed slices of developing brain with anti-glycolipids (8) and expect the labeling of unfixed tissue slices to have general utility in many fields where fixation compromises immunocytochemical detail and specificity. The sensitivity was greatly increased over cryosections, the morphology was remarkable maintained, as was the specificity for external antigens. The slices were optically sectioned with a confocal microscope to visualize OLs in situ, providing a new degree of detail and appreciation of their intact structure. To our knowledge this was the first time OLs have been images with the confocal microscope in situ. We then further expanded the optical sectioning to fixed slices of brain tissue and staining for internal OL-specific antigens (17, 18).

Results and Discussion

We have investigated the development and proliferation of OLs and their progenitors recognized by the antibodies O4, R-mAb, and O1 (20, 21, 22) in unfixed cerebellar slices (Figure 1A). Several conclusions can be drawn from these data. First, in vivo the temporal order of antibody reactivity is O4, followed by R-mAb and then O1, as shown by both spatial comparisons of the cell populations identified by these antibodies in double label immunocytochemistry, and by the antigen expression in individual cells. Second, the spatial pattern of the cells detected with either of the three anti-glycolipids support a wave of OL maturation beginning near the cerebellar base and progressing out toward the peripheral folia. Third, O4 recognizes cells with distinct morphological characteristics that appear in a developmentally regulated sequence: initially, O4 stains cells with small perikarya and two to four thin radiating process which are predominantly the most distal O4+ cells from the cerebellar base; next, O4 stains morphologically more complex cells in the midst of myelination; finally, mature OLs packed tightly within the white matter are stained. Fourth, immature O4+ cells, but not the more mature cells, proliferate during normal development of in vivo, as determined in thick unfixed tissue slices.

Figure 1A

A confocal image of an OL identified by O1 antibody in a 300 µm thick slice of unfixed postnatal day 6 rat cerebellum (8). Galactosylcerebroside, present within the cell membrane, is stained by O1. Scale bar is 25 µm and the cell is at least 30 µm below the surface of the slice.

This study presents evidence that several developmental stages prior to the expression of galactosylcerebroside (GalC) can be identified in vivo, including the O4+CalC- proligodendroblasts (6). This was possible due to an immunocytochemical staining protocol utilizing anti-glycolipid antibodies on unfixed brain slices. Proligodendroblasts may be the first post-migratory stage of the OL lineage and the most mature stage capable of proliferation in vivo. This would lend further support to the importance of the initial O4+ stage in the local expansion of OL progenitors and their commitment to myelination.

The novel method developed in this study for immunocytochemistry with anti-glycolipid antibodies on unfixed brain slices has several advantages over a similar study of cryosections. The three-dimensional structure of the tissue is preserved, the entire brain can be stained and analyzed in a series of serial slices in a manner much less labor intensive than frozen sectioning, and the specificity of anti-glycolipids cells is maintained. The option of staining unfixed tissue slices is expected to have general use for the study of cell surface epitopes, following the successful application of other existing antibodies that present non-specific cross-reactions in cryosections.

In many cases, epifluorescence microscopy can provide sufficient detail of labeled structures at and below the surface of the tissue slice. However, in areas of densely packed labeled structures, as OLs and myelin in the adult brain, many crosscut processes and overlying axons obscure the view of underlying objects. In such instances, confocal microscopy offers improved imaging, a technique that was invaluable in this study for imaging the relatively weakly stained O4+ cells.

A thickly brain slice immuno-staining protocol has also been developed to identify cells and myelin positive for the major myelin basic protein (MBP) in the developing and post-myelinated CNS (Figure 1B). With this technique we have assessed the migratory and myelination capability of highly enriched sequential developmental stages of antigenically defined OL populations (A2B5+[24] O4-, O4+CalC-, and O1+) upon transplantation into the telencephalon of the hypomyelinating mouse, shiverer. The shiverer mouse neither expresses MBP nor makes normal myelin due to a large deletion in the gene for MBP. Thirty days after transplantation, serial 225 µm sections of the host brain were immuno-stained with antiserum to MBP and analyzed by confocal microscopy. The presence of MBP+ patches of myelin in the otherwise MBP- host brain allowed a retrospective analysis of the myelinogenic activity of the transplanted progenitors cells. Both the extent of MBP+ myelin and the location of MBP+ were highly dependent on the developmental stage of the transplanted cells. Specifically, A2B5+O4- OL progenitors migrated distances of ³600 µm and produced MBP+ patches in nearly every slice of the host brain. An average of over 250 separate patches were found per host brain, some of which had cross-sectional areas of >250,000 µm2 containing as many as 60 MBP+ OL cell bodies, and with densities of myelination rivaling that of normal brain. In marked contrast, transplantation of O4+GalC- cells produced only small (1,000-25,000 µm2), scattered (25-40 per brain) patches of MBP+ myelin containing one to five cell bodies, all of which were within 50 µm of the needle track or the nearest ventricular surface. GalC+ cells produced MBP+ myelin at a level similar to that of O4+ GalC- cells.

Figure 1B

A confocal image of a cell and myelin identified by a polyclonal antibody against myelin basic protein in a 225 µm thick slice of fixed shiverer brain. The animal recieved O4+GalC- oligodendrocyte progenitors 30 days before in a transplantation protocol (18). Scale bar is 10 µm.

These data suggest that the developmental transition of OL progenitors from the O4- to the O4+ phenotype is accompanied by a dramatic reduction in the innate capacity of the cells to migrate and survive in vivo. The use of developmentally identified, enriched populations of OL progenitor cells offers the opportunity for more precise analyses of transplantation and remyelination behavior, and relates to clinically relevant studies indicating that contaminate cell types can seriously interfere with the stable integration of donor tissue into the host.

Many of the milestones set forth at the inception of this project have been met. We have established basic parameters of in situ OL development with which to compare the biology resulting within an explant slice system. We have also developed the technology to identify and visualize OLs within thick brain slices and are now in a position to noninvasively study OL development in a complex tissue structure. Our ultimate goal is to study OL development in both normal explant tissue and that which has received transplanted cells over the course of a week. An explant system will allow the addition of growth factors, neuronal membrane fractions, and other natural ligands to understand how the OL utilizes external signals to promote and delay differentiation. Such a system will provide unparalleled insights into OL development and ultimately aid to slowing the progression of crippling demyelinating diseases.


This work was supported by grants from the Johns Hopkins Center for Alternatives to Animal Testing, the National Multiple Sclerosis Society (No. 2305), and from the National Institutes of Health (NS10861).


  1. LeVine, S.M. and J.E. Goldman (1988). Embryonic divergence of oligodendrocyte and astrocyte lineages in the developing rat cerebrum. Journal of Neuroscience 8: 3992-4006.
  2. Reynolds, R. and G.P. Wilkin (1988). Development of macroglial cells in rat cerebellum II. An in situ immunohistochemical study of oligodendroglial lineage from precursor to mature myelinating cell. Development 102: 409-425.
  3. Hardy, R. and R. Reynolds (1991). Proliferation and differentiation potential of rat forebrain oligodendroglial progenitors both in vitro and in vivo. Development 111: 1061-1080.
  4. Raff, M.C., Miller, R.H., and M. Noble (1983). A glial progenitor cell that develops in vitro into an astrocyte or an oligodendrocyte depending on the culture medium. Nature 303: 389-396.
  5. Pfeiffer, S.E. and A.L. Gard (1988). Biochemical, immunological, and molecular cell-type specific markers of the central nervous system. In Advances in Neuro-oncology (eds. Kornblith, P. and M. Walker) Mount Kisco, NY: Futura Publishing Co Inc.: 3-40.
  6. Gard, A.L. and S.E. Pfeiffer (1989). Oligodendrocyte progenitors isolated directly from developing telencephalon at a specific phenotypic stage: myelinogenic potential in a defined environment. Development 106: 119-132.
  7. Bansal, R., Stefansson, K., and S.E. Pfeiffer (1992). Proligodendroblast antigen, a developmental antigen expressed by A007/O4- positive oligodendrocyte progenitors prior to the appearance of sulfatide and galactocerebroside. Journal of Neurochemistry 58: 2221-2229.
  8. Warrington, A.E. and S.E. Pfeiffer (1992). Differentiation and proliferation in postnatal rat cerebellum: analysis in unfixed tissue slices using anti-glycolipid antibodies. Journal of Neuroscience Research (in press).
  9. Levine, J.M., (1989). Neuronal influences on glial progenitor cell development. Neuron 3: 103-113.
  10. Noble, M., Murray, Stroobant, P., Waterfield, M.D., and P. Riddle (1988). Platelet-derived growth factor promotes division and motility and inhibits differentiation of the oligodendrocyte/type-2 astrocyte progenitor cell. Nature 333: 560-562.
  11. McKinnon, R.D., Toshimitsu, M., Dubois-Dalcq, M., and S.A. Aaronson (1990). FGF modulates the PDGF-driven pathway of oligodendrocyte development. Neuron 5: 603-614.
  12. Gard, A.L. and S.E. Pfeiffer (1990) FGF and PDGF differentially regulate the developmental course of oligodendrocyte progenitors placed in culture. Transactions of the Society of Neuroscience 16: 482.
  13. Blakemore, W.F. and A.J. Crang (189). The relationship between type-1 astrocytes, Schwann cells and oligodendrocytes following transplantation of glial cell cultures into demyelinating lesions in the adult rat spinal cord. Journal of Neurocytology 18: 519-528.
  14. Rosenbluth, J., Hasegawa, M., and R. Schiff (1989). Myelin-formation in myelin-deficient rat spinal cord following transplantation of normal fetal spinal cord. Neuroscience Letters 97: 35-40.
  15. Gumpel, M., Baumann, N., Raoul, M., and C. Jacque (1983). Survival and differentiation of oligodendrocytes from neural tissue transplanted into newborn shiverer brain. Neuroscience Letters 37: 307-312.
  16. Baulac, M., Lachapelle, F., Gout, O., Berger, B., Baumann, N., and M. Gumpel (1987). Transplantation of oligodendrocytes in the mouse brain: anatomical studies on the sites of myelination by transplanted myelin-forming cells. Brain Research 420: 39-47.
  17. Warrington, A.E., Barbarese, E., and S.E. Pfeiffer (1992). Stage specific, (O4+GalC-) isolated oligodendrocyte progenitors produce MBP+ myelin in vivo. Developmental Neuroscience (in press).
  18. Warrington, A.E., Barbarese, E. and S.E. Pfeiffer (1993). Differential myelinogenic capacity of specific developmental stages of the oligodendrocyte lineage upon transplantation into hypomyelinating hosts. Journal of Neuroscience Research (in press).
  19. Kohsaka, S., Yoshida, K., et al (1986). Transplantation of bulk separated oligodendrocytes into the brain of shiverer mutant mice: immunohistochemical and electron microscopic studies on myelination. Brain Research 372: 137-42.
  20. Sommer, I. and M. Schachner (1981). Monoclonal antibodies (O1 to O4) to oligodendrocyte cell surfaces: an immunocytochemical study in the central nervous system. Developmental Biology 83: 311-327.
  21. Bansal, R., Warrington, A.E., Gard, A.L., and S.E. Pfeiffer (1989). Multiple and novel specificities of monoclonal antibodies O1, O4 and R-mAb used in the analysis of oligodendrocyte development. Journal of Neuroscience Research 24: 548-557.
  22. Ranscht, B., Clapshaw, P.A., Noble, M., and W. Seifert (1982). Development of oligodendrocytes and Schwann cells studied with a monoclonal antibody against galactocerebroside. Proceedings of the National Academy of Science USA 79: 2709-2713.
  23. Curtis, R., Cohen, J., et al (1988). Development of macroglial cells in rat cerebellum. I. Use of antibodies to follow early in vitro development and migration of oligodendrocytes. Journal of Neurocytology 17: 43-54.
  24. Eisenbarth, G.S., Walsh, F.S., and M. Nirenberg (1979). Monoclonal antibody to a plasma membrane antigen of neurons. Proceedings of the National Academy of Science USA 76: 4913-4917.


Robert M. Bigsby and Peter C.M. Young
Dept. of Obstetrics & Gynecology
Indiana University School of Medicine
1001 Walnut Street, MF104
Indianapolis, IN 46223


The classical bioassays currently in use for the development of progestin and anti-progestin steroid analogues depend upon in vivo progestin effects. The purpose of the project was to develop a cell culture system that could be used as alternative bioassays. Addition of progesterone to rabbit uterine stromal cells in primary culture stimulates synthesis of a 42 kilodalton (42 kD) protein that appears in a the medium. The 42 kD protein response is specific for a progesterone receptor mediated event, and we have shown that it is blocked by known synthetic anti-progestins, RU486, ZK98, 299, and SK98, 734. Using this culture system as a bioassay, one rabbit yields enough cells to measure the effect of 30 test doses in triplicate. Thus, a culture system has been described that could substitute for current in vivo bioassays. Once fully developed, this culture bioassay would allow large scale screening for potential progestin or anti-progestin activity.

We have attempted to enhance the applicability of the bioassay in three ways:

  1. Establish a permanent cell line that remains responsive to progesterone;
  2. Purify the 42 kD protein to be used in an immunoassay;
  3. Produce a monoclonal antibody or polyclonal antiserum for development of an immunoassay.

To date we have not been successful in any of these three endeavors. Passaged cells have failed to respond to progesterone. Immortalization of cells with a temperature sensitive mutant of SV40 was also unsuccessful in establishing a progesterone-responsive cell line. Attempts at purification of the protein have not been successful but are continuing. An attempt to raise monoclonal antibodies was also successful. We are presently continuing with the purification of the protein in the hope that it can be used to raise antibodies or to obtain an amino acid sequence.


Progesterone is a key regulatory hormone in the female reproductive tract. It blocks estrogen-induced proliferation of uterine epithelial cells and promotes cellular differentiation of endometrial cells that is essential for establishment of pregnancy. Synthetic analogs of progesterone have long been used in oral contraceptive pills and as adjuvant therapy in recurring endometrial cancer (1, 2, 3). Recent observations concerning the deleterious effects of progestins on lipid metabolism has prompted continued development of new progestin analogs to be used therapeutically (4). In addition, a new class of steroids has been recently developed that exhibits effects that are antagonistic to progesterone; these compounds have varying degrees of anti-glucocorticoid activity as well (5, 6, 7, 8). The investigations carried out in the project funded by CAAT were proposed to develop a new cell culture bioassay system that could be used to screen potential progestins and anti-progestins.

The work performed has resulted in two publications to date (9, 10). These reports describe the use of primary cultures of cells obtained from rabbit uteri to test for the effects of progestins and anti-progestins. The bioassay used is based on the ability of the rabbit uterine cells to synthesis and secrete a protein of 42 kilodaltons (42 kD). Newly synthesized proteins are radiolabeled during stimulation with test compounds; the resulting radiolabeled medium proteins are separated by SDS-PAGE and analyzed densitometrically. Additional support was granted to refine this bioassay by developing a progesterone-responsive cell line and an immunoassay for detection of the 42 kD protein.

Materials and Methods

Primary Cell Cultures and 42 kD Protein Synthesis

Rabbit uterine stromal cells were enzymatically dissociated and cultured as described (11). Briefly, uteri from mature (3-3.5 kg) New Zealand rabbits were everted and incubated for 20 minuntes at 37°C with shaking in a buffer solution containing trypsin and collagenase. Dissociated cells were collected in the enzyme solution and fresh enzymes were added back to the tissue for an additional 20 minutes incubation. After dissociation, trypsin activity was inhibited by addition of soybean trypsin inhibitor and the cells were centrifuged out of suspension. The first dissociation contained mainly epithelial cells; ensuing dissociations were collected and suspended in culture medium (Ham's F12 plus 0.05% BSA, 10 µg/ml insulin, 5 µg/ml transferrin, and a solution of penicillin, streptomycin, and neomycin).

To test for the responsiveness of the cells to progesterone or other steroid analogs, the culture medium was changed on the third or fourth day of culture and 35S-methionine was added (50 µCi/ml). Steroid was added to a final concentration of 10-9 to 10-6 M and culture was continued for an additional 24 hours. The media protein was analyzed by electrophoresis through polyacrylamide (SDS-PAGE), followed with fluorography and densitometry. The optical density of the 42 kD band on the fluorogram was expressed as a percentage of the total radiolabeled proteins, i.e. the densitometric area under the curve generated by the 42 kD band was related to the area under the curves produced by all bands.

Cell Lines

Cells were grown in primary culture for five days after which they were removed from the plastic dishes and replated in a 3:1 split. Three to four days after replating, the cells were exposed to 35S-methionine and hormone as above.

In an attempt to produce an immortalized cell line that was progesterone-responsive, cells were infected with a temperature-sensitive mutant of SV40, tsSV40A255 (a gift from Dr. Janice Chou, NICHD). In preliminary tests on a monkey cell line, CV-1, we found that a high percentage of the stock virus were wild-type revertants, i.e. approximately 50% of the plaque forming units were not sensitive to culture at high temperature (39°C). However, we decided to proceed by infecting primary cultures of rabbit cells, isolated cells in foci formed during culture at 32°C and testing clones of these cells for temperature-sensitive focus forming capacity. We could expect that one half of the immortalized cell lines thus produced would be temperature-sensitive. Cells were grown on plastic dishes as usual and an aliquot of the tsSV40 virus preparation was added to the culture. Cells were switched to the permissive temperature, 32°C, for continued growth.

Protein Purification and Antibody Production

Since 2-D gel electrophoresis showed that the 42 kD produced a series of isoelectric variants (11), it may be a glycoprotein and, if so, it might be possible to isolate the protein by lectin affinity chromatography. To this end, various lectins were tested for their ability to specifically bind to the 42 kD protein. Medium from cells cultured with progesterone was electrophoresed and transblotted onto nitrocellulose. Strips of the nitrocellulose were incubated in buffer solutions containing the following biotinylated lectins: soybean, asparagus, conconavalin A, wheat germ (all from Sigma). Each of these lectins bound to the 42 kD band plus 1 or more other protein bands on the transblotted strips; wheat germ lectin bound to the 42 kD plus another band at about 90 kD. Thus, wheat germ lectin might be useful as a tool for affinity chromatography. Wheat germ lectin covalently linked to a sepharose (Sigma) was placed in a 3 ml syringe to form a column. After extensive washing of the column with tris buffered saline containing 0.1% Tween-20 (TTBS), a sample of concentrated medium wasapplied to the column. The column was washed with TTBS and eluted with N-acetyl-D-glucosamine; 1 ml fractions were collected. Samples of each fraction were analyzed by SDS-PAGE followed by Coomasie staining.

For antibody production, media samples from cultures under progesterone stimulation were concentrated and injected ip. into Balb/c mice. The injection was repeated two weeks later. Concentrated media samples were then transblotted onto nitrocellulose. The blot was stained with Fast green and the band corresponding to 42 kD was cut out and applied to the mice by subdural incision. After an additional four weeks, the mice were given a sc. injection of concentrated medium.

During the course of immunization, the mouse serum was tested for antibody production by Western blot immunoassay. Electrophoretically separated media proteins were transblotted onto nitrocellulose. Serum was diluted 1:10 with buffer and applied to strips of the transblotted nitrocellulose; antibody binding was detected with a standard biotinylated secondary antibody and the avidin-biotinylated peroxidase method using a kit (Vectastain). When a positive stain of the 42 kD band was obtained, the mice were killed and their spleen cells fused with SP2 cells. Media samples from the resultant hybridomas were tested for antibody production in an ELISA using concentrated media proteins to coat the test wells.

Results and Discussion

We have established that the bioassay as originally described, using 35S-methionine incorporation into the newly synthesized 42 kD protein band on a SDS-PAGE as the endpoint, is capable of detecting both progestin and anti-progestin activity (7, 8). Comparison with other possible in vitro assays for steroid action suggests that this bioassay could be superior if developed fully (7).

Our attempts to immortalize the cells while maintaining progesterone responsiveness has not proved successful to date. Merely passaging the cells leads to loss of progesterone responsiveness within the first passage. Infecting the cells with tsSV40 has not proved successful as the cultures in our two attempts became contaminated before immortalization could be established in any cloned lines. We are continuing with this approach, but as yet, we do not have a temperature sensitive clone for testing.

Purification attempts are along two lines: The culture medium is being concentrated by lyophilization and the concentrate is applied to preparative SDS-PAGE or to a wheat germ lectin-sepharose column. We found that elution of the wheat germ lectin column with increasing concentrations of N-acetyl-glucosamine allowed for differential elution of the 42 kD protein and the protein that migrates at 90 kD. Thus, the wheat germ lectin column should allow us to purify the protein to a large extent and to concentrate it at the same time. However, in scaling up the column purification step, we have not been able to concentrate the protein to a satisfactory extent. Likewise, using the preparative SDS-PAGE, we have not eluted sufficient protein to verify purification. This work is continuing using the remaining CAAT funds from the third year grant period.

In summary, CAAT funding has allowed further development of the in vitro bioassay for progesterone/anti-progesterone activity. Two publications have resulted from this work. Further work is continuing on the purification of the protein for eventual immunoassay development. Once we have picomolar quantities of the protein in hand, it may also become possible to obtain an amino acid sequence. Work is also continuing on the production of a temperature-sensitive cell line that will obviate the use of animals in these studies.


  1. Dickey, R.P. and S.C. Stone (1976). Progestational potency or oral contraceptives. American Journal of Obstetrics and Gynecology 47: 106-111.
  2. Phillips, A., Hahn, D.W., et al (1986). Comprehensive comparison of the potencies and activities of progestagens used in contraceptives. Contraceptives 36: 181-192.
  3. Gambrel, R.D. (1987). Use of progestogen therapy. American Journal of Obstetrics and Gynecology 156: 1304-1313.
  4. Hirvonen, E., Malkonen, M., and V. Manninen (1981). Effects of different progestogens on lipoproteins during postmenopausal replacement therapy. New England Journal of Medicine 304: 560-563.
  5. Baulieu, E. (1986). Fertility control in women: results with RU486 by the end of 1985. Journal of Steroid Biochemistry 25: 847-851.
  6. Baulieu, E.E. (1989). Contragestion and other clinical applications of RU486, an antiprogesterone at the receptor. Science 245: 1351-1357.
  7. Neef, G., Beier, S., et al (1984). New steroids with antiprogestational and antiglucocorticoid activities. Steroids 44: 349-372.
  8. Elger, W., Beier, S., et al (1986). Studies on the mechanisms of action of progesterone antagonists. Journal of Steroid Biochemstry 25: 835-845.
  9. Bigsby, R.M. (1990). Progestins and antiprogestins: A review of their role in medicine and bioassays used in their development. ATLA 17: 301-311.
  10. Bigsby, R.M. and L.M. Everett (1991). Effects of progestin antagonists, glucocorticoids, and estrogen on progesterone-induced protein secreted by rabbit endometrial stromal cells in culture. Journal of Steroid Biochemistry and Molecular Biology 39: 27-32.
  11. Bigsby, R.M. (1986). Progesterone induces a secretory protein in cultured rabbit endometrial stromal cells. Journal of Steroid Biochemistry 25: 937-942.


Alice P. Pentland, MD
Division of Dermatology
Department of Medicine
Washington University School of Medicine
606 South Euclid Avenus
St. Louis, MO 63110


Ultraviolet light exposure is known to induce erythema in sufficient doses, a response which is due in part to synthesis of prostaglandins. However, the mechanism of this increase is poorly understood. Human keratinocyte cultures exposed to 30 mj/cm2 UVB were studied 6 hours after injury. The relative activity of phospholipase versus cyclooxygenase examined using stable isotope mass measurements of PGE2. By this method, prior irradiation increased bradykinin-stimulated phospholipase activity 3.5 fold, while no change in total cellular cyclooxygenase activity was observed. The effects of irradiation on phospholipase activity were then assessed in more detail. The activities of phospholipase A2, arachidonyl CoA synthase, and arachidonyl CoA lysophosphatide acyltransferase in cell homogenates were determined. No effect of UV exposure on the activity of these enzymes was observed. These results suggest that the increase in prostaglandin synthesis produced after UV irradiation is due to increased phospholipase activity, secondary to changes in signal transduction or that the specific phospholipase responsible represents a small percentage of the total.

The role of lipid peroxidation in enhancing phospholipase activity was investigated directly, however, peroxidative changes were difficult to detect. The capacity of vitamin E, the main membrane antioxidant to suppress prostaglandin synthesis was therefore investigated. Vitamin E and 2,2,5,7,8-pentamethyl-6-hydroxychromane (PMC), a derivative in which the phytol chain is replaced by a methyl group were studied for their effect on PGE2 synthesis. PMC inhibited basal, bradykinin (BK)- and A23187- stimulated PGE2 synthesis with an apparent Ki of 1.3 µM, while no effect of vitamin E was observed. Because cultures are not vitamin E deficient, it may not be possible to detect its effect in standard culture conditions. If vitamin E acts similarly, its loss during oxidative injury may increase quantity of arachidonic acid available for eicosanoid synthesis, thereby initiating inflammation.


One of the most common manifestations of ultraviolet light B (UV) irradiation of the skin is erythema, a change which is mediated by prostaglandin synthesis (1, 2). However the mechanisms resulting in increased prostaglandin synthesis after UV injury are not well understood. Because of the role of UV light in the induction of skin cancer, a better understanding of its proinflammatory effects are needed. The mechanism of UV induced inflammation may involve the capacity of UV light to produce free radicals, oxidizing membrane lipids. This membrane damage is repaired rapidly, making it difficult to detect, but membrane based antioxidants such as vitamin E are known to be depleted (3, 4). In addition, evidence also suggests vitamin E can act as an inhibitor of cyclooxygenase and phospholipase, although high concentrations are required (5). Vitamin E and its short chain analog, 2,2,5,7,8-pentamethyl-6-hydroxychromane (PMC) were therefore examined for their capacity to modify cellular arachidonic acid metabolism.

The results of these studies have been published in American Journal of Physiology (in press) and Journal of Biological Chemistry 267: 15578-15584, 1992 (6, 7).

Materials and Methods


Octadeuterated arachidonic acid (D8-AA) was obtained from Cayman Chemical. 2,2,,5,7,8-pentamethyl-6-hydroxychromane (PMC) and vitamin E were the gift of the Esai Company and Dr. Lester Packer, respectively.

Preparation and Irradiation of Keratinocyte Cultures

Keratinocyte cultures were prepared and irradiated as described previously using 30 mj/cm2 UVB (8).

Radioimmunoassays and Protein Determinations

PGE2 was determined in the supernatants from cultured cells by RIA as previously described (8). Protein was determined by the Bicinchonic acid (BCA) Protein Assay method (7).

Assay of Phospholipase A2, Arachidonyl - CoA Synthase and Arachidonyl-CoA Lysophosphatide Acyltransferase

Homogenates were prepared by dounce homogenizer in buffer containing 10 mM Tris pH 7.6, 2 mM EDTA, 10 µM leupeptin, 1 mM PMSF, and 1 mg/ml soybean trypsin inhibitor. 100,000 x g supernates were prepared. The assays were performed as described (8).

Stable Isotope Measurements

Six hours after UV irradiation, or at appropriate times after treatment with vitamin E or its analog, keratinocyte cultures were rinsed and refed serum-free DMEM containing 30 µM octadeuterated arachidonic acid (D8-AA) with or without 1 µM bradykinin for 15 minutes. The supernatants were harvested and 10 ng of tetradeuterated PGE2 (D4-PGE2) was added to each sample as an internal standard (9). The samples were applied to 1 ml Baker octadecyl columns, then derivatized with 25 µl of 3% methoxamine hydrochloride in pyridine. The pentafluorobenzyl derivatives were prepared then derivatized to the trimethylsilyl ether of PGE2, was measured by mass ion detection using a Nermag 1010H mass spectrometer interfaced with a Delsi gas chromatograph. The ions monitored were m/z 531 (D7-PGE2, m/z 528 (D4-PGE2), and m/z 524 (D0-PGE2).


Enhanced Prostaglandin Synthesis is Due to Increased Phospholipase Activity

To examine the effect of irradiation on both cellular phospholipase and cyclooxygenase activity, stable isotope mass spectroscopy measurements of PGE2 formation were made (Figure 1) (9). By selective ion monitoring, it was possible to distinguish prostaglandin synthesized from exogenous D8 arachidonate (D7-PGE2) which reflects cyclooxygenase activity, from that synthesized using endogenous fatty acid released by phospholipase activation (D0-PGE2). Cultures were incubated with 30 µM octadeuterated arachidonic acid (D8-AA) during BK stimulation. The basal rate of D0-PGE2 synthesis (formed by phospholipase activation) was increased three-fold by prior UV exposure. In control cultures, bradykinin-stimulated formation of D0-PGE2 was increased three-fold, while in irradiated cultures it increased six-fold. No difference in synthesis of D7-PGE2 was observed, indicating that UV exposure did not increase PGE2 synthesis by increasing cyclooxygenase activity.

Figure 1

Effects of UV irradiation on PGE2 synthesis derived from endogenous arachidonic (D0-PGE2) or exogenous arachidonic acid (D7-PGE2) as measured by stable isotope mass spectroscopy. Cultures were incubated with 30 µM deuterium-arachidonic acid (D8-AA) during bradykinin stimulation (1 µM). 30 mj/cm2 UVB 6 hours prior to study (UV); Bradykinin-stimulated (BK) D0-PGE2 (), D7-PGE2 (). Mean ± SEM for 3 experiments.

UV Irradiation Does Not Change Phospholipase Activity Detected in Homogenates

To sort out the mechanism by which UV irradiation increases the availability of endogenous arachidonic acid, keratinocyte homogenates were prepared. Controls or cultures irradiated six hours previously were homogenized and activity of phospholipase determined (Figure 2). Release of 1-[14C]-arachidonic acid from the sn-2 position of phosphatidyl choline was assayed using 100,000 x g supernatant and pellet fractions in the presence and absence of calcium. Calcium-dependent hydrolysis of labeled arachidonic acid was unchanged by prior irradiation in both membrane and cytosolic fractions, while no hydrolysis of substrate was observed in the absence of calcium. In some experiments, homogenates were prepared 10 minutes after bradykinin stimulation, and phospholipase activity determined to detect specific phospholipase activity which might have become associated with the membrane after stimulation. Again, no effect of prior irradiation was observed.

Figure 2

Effect of UV irradiation on phospholipase activity detected in the 100,000 x g supernatant fraction from control () or irradiated () cultures. The data shown represent the mean value ± SEM of 2 experiments.

UV Irradiation Does Not Change the Activity of Reacylating Enzymes in Keratinocyte Homogenates

To investigate the possibility that the increased release of endogenous AA by irradiated cells could be due to differences in the rates of re-incorporation of released arachidonate into phospholipids, we examined the activities of arachidonyl CoA synthase and arachidonyl-CoA lysophosphatide acyltransferase. The activity of arachidonyl CoA synthase during a 10 minute incubation was the same in homogenates from control and irradiated cultures. Similarly, activity of arachidonyl CoA lysophosphatide acyltransferase was not influenced by prior irradiation of the keratinocyte culture (Figure 3).

Figure 3

Effect of UV irradiation on arachidonyl-CoA synthase and arachidonoyl-CoA lysophosphatide acyltransferase activities. Control (), UV irradiated (). The values shown are mean obtained from 2 experiments.

Effects of Tocopherol Analogues on Cultured Keratinocyte Prostaglandin Synthesis

Cultured keratinocytes are known to contain vitamin E as their major membrane-based antioxidant (4). Because vitamin E is known to be depleted by UV, the effect on PGE2 synthesis of tocopherol and a less hydrophobic analog of tocopherol, 2,2,5,7,8-penta methyl-6-hydroxy-chromane (PMC) was studied (3). Tocopherol or its analogue were added to confluent cultures in concentrations between 0.25 and 25 µM and the cumulative PGE2 synthesis over the next 24 hours determined by immunoassay. A dose-dependent decrease in the quantity of prostaglandin synthesized by the cultures over the 24 hours after addition occurred in cultures treated with the PMC (Figure 4). Little effect was noted in cultures treated with a-tocopherol itself.

Figure 4

Effect of a-tocopherol () and PMC (chromanol C1) () on cumulative prostaglandin synthesis by confluent keratinocyte cultures. Compounds were added at time 0 and the cumulative PGE2 synthesis by the cultures determined after 24 hours. Mean ± SEM of 3 experiments.

PMC Inhibits Phospholipase But Not Cyclooxygenase

To examine the effect of PMC on both cellular phospholipase and cyclooxygenase activity, stable isotope mass spectroscopy measurements of PGE2 formation were made (Figure 5). Cultures were incubated with 30 µM octadeuterated arachidonic acid (D8-AA) during BK stimulation. BK-stimulated formation of D0-PGE2, formed by PL activation, was decreased 70% by 5 µM PMC, and eliminated by 50 µM PMC. No difference in D7-PGE2 was observed, again indicating that PMC did not inhibit PGE2 synthesis by inhibiting cyclooxygenase activity.

Figure 5

PMC effects on PGE2 synthesis derived from endogenous arachidonic (D0-PGE2) or exogenous arachidonic acid (D7-PGE2) as measured by stable isotope mass spectroscopy, Cultures were incubated with 30 µM deuterium-arachidonic acid (D8-AA) during bradykinin stimulation. The ions monitored were m/z 531 (D7-PGE2), m/z 524 (D0-PGE2) and m/z 528 (D4-PGE2, an internal standard). Unstimulated (); Bradykinin-stimulated (). Mean ± SEM for 3 experiments.

PMC Decrease Hydrolysis of [14C-AA] from the sn-2 Position of Phosphatidylcholine

To document the efficacy of PMC as an inhibitor of phospholipase activity, the effect of PMC on bovine pancreatic phospholipase was examined in vitro. Calcium-dependent hydrolysis of arachidonic acid from the sn-2 position of phosphatidylcholine in small unilamellar vesicles formed in the presence or absence of 0.5, 5, or 50 µM PMC was studied. Substrate and enzyme were preincubated together before initiating the reaction by the addition of calcium. In contrast to the data using prelabeling methods, dose-dependent inhibition of hydrolysis was produced by PMC (Figure 6), documenting the effectiveness of PMC as an inhibitor of phospholipase.

Figure 6

Effect of PMC on hydrolysis of [14]-arachidonic acid from the sn-2 position of phosphatidylcholine. Small unilamellar vesicles containing 0.5 to 50 µM PMC were incubated with 0.5 units bovine pancreatic phospholipase A2 for 15 minutes and the release of label determined after organic extraction and thin layer chromatography. The mean ± SEM of 4 experiments if shown.


The mechanism by which UV light exposure increases the synthesis of prostaglandins in irradiated cultures has been examined in detail. At the time points examined, the increased release of PGE2 observed appears due to an increase in phospholipase activity in studies done using intact cells. No change in activity of phospholipase could be detected in homogenate preparations, however. A detailed examination of fatty acid uptake into membrane lipids was also done. Results using homogenates prepared from previously irradiated keratinocytes revealed no change in activity of phospholipase, arachidonyl-CoA lysophosphotide acyltransferase or arachidonyl CoA synthase. Collectively, the data examining the effect of UV on irradiated cultures suggests that phospholipase activation is increased by changes in signal transduction reacylase which is specifically regulated in response to irradiation, but is present in low relative abundance.

One of the possible underlying mechanisms for this increase in phospholipase activity is membrane peroxidation, as perturbed membrane structure is known to increase the capacity of many phospholipases to hydrolyze substrate (10). Initial efforts to detect peroxidation of intact cellular membranes were suggestive, but not conclusive. Because the chief membrane antioxidant is vitamin E, the effects of the vitamin E and PMC were examined. Although there was no effect of it naturally in the membrane (3). In contrast, its analogue was very effective in suppression of PGE2 synthesis. If endogenous vitamin E is capable of producing the same effects in vivo, its loss during UV injury may explain the increased release of PGE2 observed after irradiation.

Milestones: Initial studies to identify products of peroxidation were done. Using the thiobarbituric acid method, no consistent increase in peroxidated lipid was observed. Detailed analysis of the fatty acid composition of the phospholipid groups present in keratinocytes was done, as presented in the paper currently in press in American Journal of Physiology (appended). No major changes induced by irradiation were found at any of the time points examined. In lieu of documenting this effect, the possible role of decreased cellular antioxidant capacity was assessed by studying the effect of well characterized antioxidants on cellular prostaglandin synthesis. All the studies described were completed within three months of the termination of the grant.


  1. Black, A.K., Greaves, M.W., et al (1978). The effects of indomethacin on arachidonic acid and prostaglandins E2 and F2a in human skin 24 hours after UVB and UVC irradiation. British Journal of Clinical Pharmacology 6: 261-266.
  2. Synder, D.S. and W. Eaglestien (1974). Intradermal anti-prostaglandin agents and sunburn. Journal of Investigative Dermatology 62: 47-50.
  3. Kagan, V., Packer, L., et al (1990). Biological Oxidation Systems (ed. C. Reddy) NY: Academic Press.
  4. Fuchs, J., Huflejit, M.E., et al (1989). Impairment of enzymatic and nonenzymatic antioxidants in skin by UVB irradiation. Journal of Investigative Dermatology 93: 769-773.
  5. Panganamala, R.V. and D.G. Cornwell (1982). The effects of vitamin E on arachidonic acid metabolism. New York Academy of Science 393: 376-393.
  6. Kang-Rotondo, C., Miller, C.C., et al (1992). Enhanced prostaglandin synthesis after ultraviolet injury is due to increased phospholipase activity. American Journal of Physiology (in press).
  7. Pentland, A.P., Morrison, A.R., et al (1992). Tocopherol analogs suppress arachidonic acid metabolism via phospholipase inhibition. Journal of Biological Chemistry 267: 15578-15584.
  8. Fuse, I., Tomoaki, I. et al (1988). Phorbol ester, 1,2,diacylglycerol, and collagen induce inhibition of arachidonic acid incorporation into phospholipids in human platelets. Journal of Biological Chemistry 264: 3890-3895.
  9. Coyne, D.W., Nickols, M., et al (1992). Regulation of mesangial cells cyclooxygenase synthesis by cytokines and glucocorticoids. American Journal of Physiology (in press).
  10. Beckman, J.K., Borowitz, S.M., et al (1987). The role of phospholipase A activity in rat liver, microsomal lipids peroxidation. Journal of Biological Chemistry 262: 1479-1484.