Possible mechanisms underlying calcium signal changes in murine sensory neurons at experimental diabetes

Numerous investigations indicate thatdiabetes mellitus is associated with pronounced changes in calcium homeostasis, which in turn lead to substantial complications in most systemic functions. However, similar combined investigations of possible alterations in calcium signalling in the nerve cells are lacking, although pathological changes in nervous functions, including paresthesia and increased pain sensitivity, are common complications ofdiabetes mellitus in humans. Therefore, we studied the changes in calcium homeostasis in sensory neurons of mice with streptozotocin (STZ)-induced and genetically determineddiabetes mellitus.     

Effect of insulin and nimodipin on calcium signals in murine primary afferent neurons with experimental diabetes

The effect of insulin injections and long-term administration of the calcium channel blocker nimodipin on depolarization-induced calcium signals was studied in neurons isolated from the dorsal root ganglia of mice with streptozotocin-induced diabetes. Induction of diabetes in mice was followed by slowing of the calcium signal decay kinetics in small neurons (mainly related to the transmission of nociceptive signals). Subcutaneous insulin injections (1 U/kg) tended to normalize the parameters of calcium signals modified by diabetes. Preliminary 3-week-long peroral administration of nimodipin (40 mg/kg) increased the peak amplitude of depolarization-induced calcium signals in isolated neurons and caused spontaneous activity usually absent in the cells under control conditions. Kinetics of calcium transients in this case remained slow. It was concluded that hyperglycemia and related impairments of the surplus Ca²⁺ extrusion mechanisms play an essential role in genesis of the changes in calcium signals caused by streptozotocin-induced diabetes.Neirofiziologiya/Neurophysiology, Vol. 27, No. 5/6, pp. 342–349, September–December, 1995.     

Calcium transient evoked by nicotine in isolated rat vagal pulmonary sensory neurons

It has been shown that inhaled cigarette smoke activates vagal pulmonary C fibers and rapidly adapting receptors (RARs) in the airways and that nicotine contained in the smoke is primarily responsible. This study was carried out to determine whether nicotine alone can activate pulmonary sensory neurons isolated from rat vagal ganglia; the response of these neurons was determined by fura-2-based ratiometric Ca(2+) imaging. The results showed: 1) Nicotine (10(-4) M, 20 s) evoked a transient increase in intracellular Ca(2+) concentration ([Ca(2+)](i)) in 175 of the 522 neurons tested (Delta[Ca(2+)](i) = 142.2 +/- 12.3 nM); the response was reproducible, with a small reduction in peak amplitude in the same neurons when the challenge was repeated 20 min later. 2) A majority (59.7%) of these nicotine-sensitive neurons were also activated by capsaicin (10(-7) M). 3) 1,1-Dimethyl-4-phenylpiperazinium iodide (DMPP; 10(-4) M, 20 s), a selective agonist of the neuronal nicotinic acetylcholine receptors (NnAChRs), evoked a pattern of response similar to that of nicotine. 4) The responses to nicotine and DMPP were either totally abrogated or markedly attenuated by hexamethonium (10(-4) M). 5) In anesthetized rats, right atrial bolus injection of nicotine (75-200 mug/kg) evoked an immediate (latency <1-2 s) and intense burst of discharge in 47.8% of the pulmonary C-fiber endings and 28.6% of the RARs tested. In conclusion, nicotine exerts a direct stimulatory effect on vagal pulmonary sensory nerves, and the effect is probably mediated through an activation of the NnAChRs expressed on the membrane of these neurons.     

Inactivation of N-type calcium current in chick sensory neurons: calcium and voltage dependence

We have studied the inactivation of high-voltage-activated (HVA), omega- conotoxin-sensitive, N-type Ca2+ current in embryonic chick dorsal root ganglion (DRG) neurons. Voltage steps from -80 to 0 mV produced inward Ca2+ currents that inactivated in a biphasic manner and were fit well with the sum of two exponentials (with time constants of approximately 100 ms and > 1 s). As reported previously, upon depolarization of the holding potential to -40 mV, N current amplitude was significantly reduced and the rapid phase of inactivation all but eliminated (Nowycky, M. C., A. P. Fox, and R. W. Tsien. 1985. Nature. 316:440-443; Fox, A. P., M. C. Nowycky, and R. W. Tsien. 1987a. Journal of Physiology. 394:149-172; Swandulla, D., and C. M. Armstrong. 1988. Journal of General Physiology. 92:197-218; Plummer, M. R., D. E. Logothetis, and P. Hess. 1989. Neuron. 2:1453-1463; Regan, L. J., D. W. Sah, and B. P. Bean. 1991. Neuron. 6:269-280; Cox, D. H., and K. Dunlap. 1992. Journal of Neuroscience. 12:906-914). Such kinetic properties might be explained by a model in which N channels inactivate by both fast and slow voltage-dependent processes. Alternatively, kinetic models of Ca-dependent inactivation suggest that the biphasic kinetics and holding-potential-dependence of N current inactivation could be due to a combination of Ca-dependent and slow voltage- dependent inactivation mechanisms. To distinguish between these possibilities we have performed several experiments to test for the presence of Ca-dependent inactivation. Three lines of evidence suggest that N channels inactivate in a Ca-dependent manner. (a) The total extent of inactivation increased 50%, and the ratio of rapid to slow inactivation increased approximately twofold when the concentration of the Ca2+ buffer, EGTA, in the patch pipette was reduced from 10 to 0.1 mM. (b) With low intracellular EGTA concentrations (0.1 mM), the ratio of rapid to slow inactivation was additionally increased when the extracellular Ca2+ concentration was raised from 0.5 to 5 mM. (c) Substituting Na+ for Ca2+ as the permeant ion eliminated the rapid phase of inactivation. Other results do not support the notion of current-dependent inactivation, however. Although high intracellular EGTA (10 mM) or BAPTA (5 mM) concentrations suppressed the rapid phase inactivation, they did not eliminate it. Increasing the extracellular Ca2+ from 0.5 to 5 mM had little effect on this residual fast inactivation, indicating that it is not appreciably sensitive to Ca2+ influx under these conditions.(ABSTRACT TRUNCATED AT 400 WORDS)     

Dynamics of calcium release and uptake by the internal calcium stores in rat sensory neurons

Calcium dynamics in the endoplasmic reticulum of dorsal root ganglion neurons of rats during Ca²⁺ release induced by caffeine and subsequent Ca²⁺ uptake were studied. Calcium release is shown to include two (a short transient and a prolonged slow) phases. We suggest that the transient phase reflects release of free Ca from the calcium store, while the slow phase reflects transition of Ca from a bound form to a free one. The process of Ca²⁺ uptake is characterized by exponential recovery of the calcium level in the store due to the SERCA activity. Neirofiziologiya/Neurophysiology, Vol. 38, No. 4, pp. 361–363, July–August, 2006.     

Modulation of calcium channels by norepinephrine in internally dialyzed avian sensory neurons

Modulation of voltage-dependent Ca channels by norepinephrine (NE) was studied in chick dorsal root ganglion cells using the whole-cell configuration of the patch-clamp technique. Cells dialyzed with K+ and 2-10 mM EGTA exhibited Ca action potentials that were reversibly decreased in duration and amplitude by NE. Ca channel currents were isolated from other channel contributions by using: (a) tetrodotoxin (TTX) to block gNa, (b) internal K channel impermeant ions (Cs or Na/N-methylglucamine mixtures) as K substitutes, (c) external tetraethylammonium (TEA) to block K channels, (d) internal EGTA to reduce possible current contribution from Ca-activated channels. A marked decline (rundown) of Ca conductance was observed during continual dialysis, which obscured reversible NE effects. The addition of 2-5 mM MgATP to the intracellular solutions greatly retarded Ca channel rundown and permitted a clear assessment of modulatory drug effects. The inclusion of an intracellular creatine phosphate/creatine phosphokinase nucleotide regeneration system further stabilized Ca channels, which permitted recording of Ca currents for up to 3 h. NE reversibly decreased both steady state Ca currents and Ca tail currents in Cs/EGTA/MgATP-dialyzed cells. A possible role of several putative intracellular second messengers in NE receptor-Ca channel coupling was investigated. Cyclic AMP or cyclic GMP added to the intracellular solutions at concentrations several orders of magnitude higher than the Kd for activation of cyclic nucleotide-dependent protein kinases did not block or mask the expression of the NE-mediated decrease in gCa. Addition of internal EGTA to a final concentration of 10 mM also did not affect the expression of the NE response. These results suggest that neither cyclic AMP nor cyclic GMP nor Ca is acting as a second messenger coupling the NE receptor to the down-modulated Ca channel population.     

Mechanisms for calcium sensing receptor-regulated stomatal closure in response to the extracellular calcium signal

In Arabidopsis, extracellular calcium (Ca²⁺_o) promotes intracellular calcium (Ca²⁺_i) transients and stomatal closure, which has been found to be regulated by the calcium sensing receptor (CAS). However, the detailed pathways for transducting the Ca²⁺_o signal by CAS are still unclear. We found that nitric oxide (NO) and the hydrogen peroxide (H₂O₂) accumulated in the guard cell chloroplast were the two elements that act downstream of the CAS signaling and trigger the stomatal closure by prolonging Ca²⁺_i transients.¹ Here we provide more commentary on CAS-regulated H₂O₂ generation from chloroplast and Ca²⁺_i transients in response to Ca²⁺_o, as well as other potential mechanisms that may be involved in the CAS signaling pathway.     

Modulation of action potential and calcium signaling by levetiracetam in rat sensory neurons

Levetiracetam (LEV), a new anticonvulsant agent primarily used to treat epilepsy, has been used in pain treatment but the cellular mechanism of this action remains unclear. This study aimed to investigate effects of LEV on the excitability and membrane depolarization-induced calcium signaling in isolated rat sensory neurons using the whole-cell patch clamp and fura 2-based ratiometric Ca(2+)-imaging techniques. Dorsal root ganglia (DRG) were excised from neonatal rats, and cultured following enzymatic and mechanical dissociation. Under current clamp conditions, acute application of LEV (30 μM, 100 μM and 300 μM) significantly increased input resistance and caused the membrane to hyperpolarize from resting membrane potential in a dose-dependent manner. Reversal potentials of action potential (AP) after hyperpolarising amplitudes were shifted to more negative, toward to potassium equilibrium potentials, after application of LEV. It also caused a decrease in number of APs in neurons fired multiple APs in response to prolonged depolarization. Fura-2 fluorescence Ca(2+) imaging protocols revealed that HiK(+) (30 mM)-induced intracellular free Ca(2+) ([Ca(2+)](i)) was inhibited to 97.8 ± 4.6% (n = 17), 92.6 ± 4.8% (n = 17, p < 0.01) and 89.1 ± 5.1% (n = 18, p < 0.01) after application of 30 μM, 100 μM and 300 μM LEV (respectively), without any significant effect on basal levels of [Ca(2+)](i). This is the first evidence for the effect of LEV on the excitability of rat sensory neurons through an effect which might involve activation of potassium channels and inhibition of entry of Ca(2+), providing new insights for cellular mechanism(s) of LEV in pain treatment modalities.     

Modulation of action potential and calcium signaling by levetiracetam in rat sensory neurons

Levetiracetam (LEV), a new anticonvulsant agent primarily used to treat epilepsy, has been used in pain treatment but the cellular mechanism of this action remains unclear. This study aimed to investigate effects of LEV on the excitability and membrane depolarization-induced calcium signaling in isolated rat sensory neurons using the whole-cell patch clamp and fura 2–based ratiometric Ca²⁺-imaging techniques. Dorsal root ganglia (DRG) were excised from neonatal rats, and cultured following enzymatic and mechanical dissociation. Under current clamp conditions, acute application of LEV (30 µM, 100 µM and 300 µM) significantly increased input resistance and caused the membrane to hyperpolarize from resting membrane potential in a dose-dependent manner. Reversal potentials of action potential (AP) after hyperpolarising amplitudes were shifted to more negative, toward to potassium equilibrium potentials, after application of LEV. It also caused a decrease in number of APs in neurons fired multiple APs in response to prolonged depolarization. Fura-2 fluorescence Ca²⁺ imaging protocols revealed that HiK⁺ (30 mM)-induced intracellular free Ca²⁺ ([Ca²⁺]_i) was inhibited to 97.8 ± 4.6% (n = 17), 92.6 ± 4.8% (n = 17, p < 0.01) and 89.1 ± 5.1% (n = 18, p < 0.01) after application of 30 µM, 100 µM and 300 µM LEV (respectively), without any significant effect on basal levels of [Ca²⁺]_i. This is the first evidence for the effect of LEV on the excitability of rat sensory neurons through an effect which might involve activation of potassium channels and inhibition of entry of Ca²⁺, providing new insights for cellular mechanism(s) of LEV in pain treatment modalities.     

Influence of external pH on two types of low-voltage-activated calcium currents in primary sensory neurons of rats

The influence of extracellular pH (pH(o)) on low-voltage-activated calcium channels of acutely isolated DRG neurons of rats was examined using the whole cell patch-clamp technique. It has been found that in the neurons of middle size with capacitance C=60+/-4.8 pF (mean+/-S.E., n=8) extracellular acidification from pH(o) 7.35 to pH(o) 6.0 significantly and reversibly decreased LVA calcium current densities by 75+/-3.7%, shifted potential for half-maximal activation to more positive voltages by 18.7+/-0.6 mV with significant reduction of its voltage dependence. The half-maximal potential of steady-state inactivation shifted to more positive voltages by 12.1+/-1.7 mV (n=8) and also became less voltage dependent. Dose-response curves for the dependence of maximum values of LVA currents on external pH in neurons of middle size have midpoint pK(a)=6.6+/-0.02 and hill coefficient h=0.94+/-0.04 (n=5). In small cells with capacitance C=26+/-3.6 pF (n=5), acidosis decreased LVA calcium current densities only by 15.3+/-1.3% and shifted potential for half-maximal activation by 5.5+/-1.0 mV with reduction of its voltage dependence. Half-maximal potential of steady-state inactivation shifted to more positive voltages by 10+/-1.6 mV (n=4) and also became less voltage dependent. Dose-response curves for the dependence of maximum values of LVA currents on external pH in neurons of small size have midpoint pK(a)=7.9+/-0.04 and hill coefficient h=0.25+/-0.1 (n=4). These two identified types of LVA currents besides different pH sensitivity demonstrated different kinetic properties. The deactivation of LVA currents with weak pH sensitivity after switching off depolarization to -30 mV had substantially longer decay time than do currents with strong pH sensitivity (tau(d) approximately 5 ms vs. 2 ms respectively). It was found that the prolongation of depolarization steps slows the subsequent deactivation of T-type currents in small DRG neurons. Deactivation traces in these neurons were better described by the sum of two exponentials. Thus, we suppose that T-type channels in small DRG neurons are presented mostly by alpha1I subunit. We suggest that these two types of LVA calcium channels with different sensitivity to external pH can be differently involved in the origin of neuropathic changes.     

Thiamine monophosphatase: a genuine marker for transganglionic regulation of primary sensory neurons

Thiamine monophosphatase (TMPase) has been selectively localized in small dorsal root ganglion cells and in their central and peripheral terminals. Light microscopic localization of TMPase, and its alterations due to transganglionic effects, are identical with those of fluoride-resistant acid phosphatase (FRAP), but are not contaminated by the ubiquitous lysosomal reaction inevitable in trivial acid phosphatase-stained sections. TMPase is inhibited by 0.1 mM NaF, which is slightly less than the concentration needed to inhibit FRAP (0.2-0.4 mM). It is assumed that TMPase and FRAP are identical enzymes. In the perikaryon of small dorsal root ganglion cells, TMPase is located in the cisterns of the endoplasmic reticulum and in the Golgi apparatus. The central terminals of these cells are scalloped (sinusoid) axon terminals, surrounded by membrane-bound TMPase activity. Central terminals outline substantia gelatinosa Rolandi throughout the spinal cord, as well as the analogous nucleus spinalis trigemini in the medulla. TMPase-active central terminals outline "faisceau de la corne postérieure" in the sacral cord, as well as Lissauer's tract in the thoracic, upper lumbar, and sacral segments, and the paratrigeminal nucleus and the terminal (sensory) nucleus of the ala cinerea in the brainstem. Peripheral terminals displaying TMPase activity are fine nerve plexuses of C fibers. The TMPase activity of the central terminals disappears after dorsal rhizotomy in the course of Wallerian degeneration, and is depleted in the course of transganglionic degenerative atrophy (after transection of the related peripheral sensory nerve). TMPase is an outstanding genuine marker for the study of transganglionic regulation in Muridae.     

Mechanisms of up-regulation of single calcium channels by serotonin in Helix pomatia neurons

Action of serotonin (5-HT) on single Ca(2+) channel activity was studied in identified neurons of snail Helix pomatia. Only one type of Ca(2+) channels of 5 pS unitary conductance was determined under patch-clamp cell-attached mode. Kinetic analysis have shown a monotonically declining distribution of channel open times (OT) with mean time constant of 0.2 ms. The distribution of channel closed times (CT) could be fitted by double-exponential curve with time constants 1 and 12 ms. We established that 5-HT acts on Ca(2+) channel activity indirectly via cytoplasm. 5-HT prolonged the OT (up to 0.3 ms) and shortened the CT proportionally for both constants to 0.4 and 6 ms correspondingly. A conclusion is made that enhancement of Ca(2+) macro-current by 5-HT is determined by kinetic changes, increase of the number of active channels, and increase of the probability of OT. At the same time the transmitter did not affect the unitary channel conductance.     

Synapse formation by sensory neurons after cross-species transplantation in crickets: the role of positional information

The role of positional information in synapse formation was studied in the cricket cercal sensory system by transplanting epidermis from one species of cricket to another. Strips of cercal epidermis containing identified sensory neurons were transplanted from a black donor species to a tan host species; the color difference was used to distinguish between donor and host tissue in adults. Transplanted sensory neurons regenerated axons into the host terminal abdominal ganglion where they formed functional chimeric synapses. These methods were used to test the role of positional information in central synapse formation. Newly generated sensory neurons, formed by the donor tissue at the border between graft and host, were examined to test the idea that their position would determine their structure, function, and projection pattern. These "intercalated" sensory neurons support the positional information hypothesis. First, they had directional sensitivities which were appropriate to their location on the cercus; receptors of this directionality would never be made by the donor tissue if left in its original position. Second, these sensory neurons projected to regions of the CNS known to be appropriate for their directionality. Finally, simultaneous recordings from these ectopic sensory neurons and host interneurons demonstrated the expected synaptic connection, based on the overlap of pre- and postsynaptic cells. Thus three aspects of receptor function, directionality, afferent projection, and choice of synaptic partners, appeared to be controlled by positional information.     

Mutual influence of plasma membrane structures and intracellular calcium stores on the formation of calcium signals in primary nociceptive neurons

Participation of different calcium-regulating mechanisms in the formation of intracellular calcium signals in rat primary sensory neurons was studied using two-wavelength fluorescent microscopy. Mitochondria were shown to be the most powerful intracellular calcium-regulating structures in the investigated neurons. These organelles were involved in the modulation of calcium signals induced either by Ca²⁺ entry from the extracellular medium or by Ca²⁺ release from endoplasmic reticulum (ER). Analysis of the mitochondrial calcium exchange showed that the efficiency of mitochondria depended on whether calcium entered the cytosol from ER or from the extracellular solution. Depletion of ER by activation of ryanodine-sensitive, inositol-3-phosphate-sensitive receptors of ER or by activation of the leak channels via the block of ATPases in ER activated the store-operated calcium entry from the extracellular medium to cytosol. The kinetics of the rising phase of these Ca²⁺ transients depended on the way of ER depletion. This allows suggesting the existence of different activation mechanisms for the studied signals. The block of the mitochondrial calcium uniporter resulted in a rapid recovery of the intracellular calcium concentration after the Ca²⁺ transient induced by store-operated calcium influx. We conclude that mitochondrial calcium uptake can prevent calcium-dependent inactivation of store-operated calcium channels.     

Wheat germ agglutinin binding in rat primary sensory neurons: a histochemical study

Summary The binding of wheat germ agglutinin (WGA) to L5 dorsal root ganglion (DRG) cells in the rat was studied with the WGA-FITC conjugate. These cells were also examined with regard to overlap between WGA staining and choleragenoid-like immunoreactivity. The DRG cells showed varying intensity of the staining, which was confined to the cytoplasm. The majority of the small cells were heavily stained, whereas the large cells showed less or occasionally no staining. Lectin binding was observed along nerve fibres in the ganglion, and appeared to be localized to the Schwann cells and to the nodes of Ranvier. Strong staining was also observed in the area surrounding the ganglion cells and seemed to be confirned to the satellite cells. A subpopulation of the ganglion cells showed both WGA-staining and choleragenoid-like immunoreactivity. These results indicate a nonuniform affinity of WGA for different subpopulations of DRG neurons.     

Comparison of currents activated by noxious heat in rat and chicken primary sensory neurons

The vanilloid receptor 1 (VR1) gene is responsible for both capsaicin-, and low threshold (LT) noxious heat-sensitivity in mammalian primary sensory neurons. Although, birds lack capsaicin-sensitivity they express LT noxious heat-sensitivity. Here, we compared LT noxious heat-activated whole-cell currents produced by rat and chicken cultured dorsal root ganglion neurons in order to find out the similarities and differences in the LT noxious heat transduction mechanisms between the two species. No significant differences between rat and chicken neurons were found in the mean cell diameter of the LT noxious heat-sensitive cells (20.4+/-0.8 microm, n=19 and 20.6+/-0.6 microm, n=11, respectively) and the average threshold (45.7+/-0.5 degrees C, n=19 and 46.1+/-0.7 degrees C, n=11, respectively) and peak amplitude (-2.9+/-0.6 nA, n=19 and -2.1+/-0.6 nA, n=11, respectively) of the heat-evoked responses. The current-voltage curves of the responses both in rat and chicken cells reversed at the same range (-19.5+/-3.8 mV, n=4 and -15.5+/-1. 2 mV, n=3, respectively) and showed strong outward rectification at negative membrane potentials. While all LT noxious heat-sensitive rat cells responded to capsaicin, none of the chicken neurons produced detectable response to it. These findings suggest that a VR1 homologue which lacks to sequence for capsaicin-sensitivity is possibly the LT noxious heat transducer in chicken.     

Temperature sensing by plants: the primary characteristics of signal perception and calcium response

Cold elicits an immediate rise in the cytosolic free calcium concentration ([Ca2+]c) of plant cells. We have studied the concerted action of the three underlying mechanisms, namely sensing, sensitisation and desensitisation, which become important when plants in the field are subjected to changes in temperature. We applied different regimes of temperature changes with well-defined cooling rates to intact roots of Arabidopsis thaliana expressing the calcium-indicator, aequorin. Our results indicate that temperature sensing is mainly dependent on the cooling rate, dT/dt, whereas the absolute temperature T is of less importance. Arabidopsis roots were found to be sensitive to cooling rates of less than dT/dt = 0.01 degrees C/s. However, at cooling rates below 0.003 degrees C/s (i.e. cooling 10 degrees C in 1 h) there is no detectable [Ca2+]c response at all. At low temperature, the sensitivity of the plant cold-detection system is increased. This in turn produces greater cooling-induced [Ca2+]c elevations. Prolonged or repeated cold treatment attenuates the [Ca2+]c responses to subsequent episodes of cooling.     

Spike-timing in primary sensory neurons: a model of somatosensory transduction in the rat

In previous work, we constructed a simple electro-mechanical model of transduction in the rat mystacial follicle that was able to replicate primary afferent response profiles to a variety of whisker deflection stimuli. Here, we update that model to fit newly available spike-timing response data, and demonstrate that the new model produces appropriate responses to richer stimuli, including pseudo white noise and natural textures, at a spike-timing level of detail. Additionally, we demonstrate reliable distributed encoding of multi-component oscillatory signals. No modifications were necessary to the mechanical model of the physical components of the follicle–sinus complex, supporting its generality. We conclude that this model, and its continued development, will aid the understanding both of somatosensory systems in general, and of physiological results from higher (e.g. thalamocortical) systems by accurately characterising the signals on which they operate.     

Depletion of 43-kD growth-associated protein in primary sensory neurons leads to diminished formation and spreading of growth cones

Epithelial-mesenchymal interactions control epidermal growth and differentiation, but little is known about the mechanisms of this interaction. We have examined the effects of human dermal microvascular endothelial cells (DMEC) and fibroblasts on keratinocytes in conventional (feeder layer) and organotypic cocultures (lifted collagen gels) and demonstrated the induction of paracrine growth factor gene expression. Clonal keratinocyte growth was similarly stimulated in cocultures with irradiated DMEC and fibroblasts as feeder cells. This effect is most probably caused by induction of growth factor expression in cocultured dermal cells. Keratinocytes stimulated mRNA levels for KGF and IL-6 in both mesenchymal cell types and GM-CSF in fibroblasts. The feeder effect could not be replaced by conditioned media or addition of isolated growth factors. In organotypic cocultures with keratinocytes growing on collagen gels (repopulated with dermal cells), a virtually normal epidermis was formed within 7 to 10 d. Keratinocyte proliferation was drastically stimulated by dermal cells (histone 3 mRNA expression and BrdU labeling) which continued to proliferate as well in the gel. Expression of all typical differentiation markers was provoked in the reconstituted epithelium, though with different localization as compared to normal epidermis. Keratins K1 and K10 appeared coexpressed but delayed, reflecting conditions in epidermal hyperplasia. Keratin localization and proliferation were normalized under in vivo conditions, i.e., in surface transplants on nude mice. From these data it is concluded that epidermal homeostasis is in part controlled by complex reciprocally induced paracrine acting factors in concert with cell-cell interactions and extracellular matrix influences.