Ca2+ release dynamics in parotid and pancreatic exocrine acinar cells evoked by spatially limited flash photolysis

Intracellular calcium concentration ([Ca(2+)](i)) signals are central to the mechanisms underlying fluid and protein secretion in pancreatic and parotid acinar cells. Calcium release was studied in natively buffered cells following focal laser photolysis of caged molecules. Focal photolysis of caged-inositol 1,4,5 trisphosphate (InsP(3)) in the apical region resulted in Ca(2+) release from the apical trigger zone and, after a latent period, the initiation of an apical-to-basal Ca(2+) wave. The latency was longer and the wave speed significantly slower in pancreatic compared with parotid cells. Focal photolysis in basal regions evoked only limited Ca(2+) release at the photolysis site and never resulted in a propagating wave. Instead, an apical-to-basal wave was initiated following a latent period. Again, the latent period was significantly longer under all conditions in pancreas than parotid. Although slower in pancreas than parotid, once initiated, the apical-to-basal wave speed was constant in a particular cell type. Photo release of caged-Ca(2+) failed to evoke a propagating Ca(2+) wave in either cell type. However, the kinetics of the Ca(2+) signal evoked following photolysis of caged-InsP(3) were significantly dampened by ryanodine in parotid but not pancreas, indicating a more prominent functional role for ryanodine receptor (RyR) following InsP(3) receptor (InsP(3)R) activation. These data suggest that differing expression levels of InsP(3)R, RyR, and possibly cellular buffering capacity may contribute to the fast kinetics of Ca(2+) signals in parotid compared with pancreas. These properties may represent a specialization of the cell type to effectively stimulate Ca(2+)-dependent effectors important for the differing primary physiological role of each gland.     

InsP3- and Ca2(+)-induced Ca2+ release in single mouse oocytes

To better understand the mechanism of intracellular Ca2+ mobilization, mouse oocytes were micro-injected with 'caged'-inositol-1,4,5 triphosphate caged-InsP3) together with the Ca2+ indicator Fluo-3 to directly induce and monitor Ca2+ redistribution. Photo-released InsP3 elicits [Ca2+]i changes exhibiting several kinetic phases and threshold behaviour. Often Ca2+ oscillations were induced after a single InsP3 pulse. Autoregenerative Ca2+ transients could also be induced by injections of Ca2+ itself, demonstrating unequivocally the presence of a Ca2(+)-induced Ca2(+)-release mechanism in these cells.     

Role of extracellular Ca2+ in acetylcholine-induced repetitive Ca2+ release in submandibular gland acinar cells of the rat

Acetylcholine (ACh) caused repetitive transient Cl⁻ currents activated by intracellular Ca²⁺ in single rat submandibular grand acinar cells. As the concentration of ACh increased the amplitude and the frequency of the transient Cl⁻ currents increased. These responses occurred also in the absence of extracellular Ca²⁺ but disappeared after several minutes. Repetitive transient Cl⁻ currents were restored by readmission of Ca²⁺ to the extracellular solution. The higher the concentration of extracellular Ca²⁺ readmitted, the larger the amplitude of the transient Cl⁻ currents. Ca²⁺ entry through a store-coupled pathway was detected by application of Ca²⁺ to the extracellular solution during a brief cessation of stimulation with ACh. In these experiments too, the higher the concentration of Ca²⁺, the larger the transient Cl⁻ currents activated by Ca²⁺ released from the stores. The time course of decrease in total charge movements of repetitive transient responses to ACh with removal of extracellular Ca²⁺ depended on a decrease in charge movements of each transient event rather than a decrease in frequency of the repetitive events. The decrease of charge movements of each transient event was due to a decrease in its amplitude rather than its duration. The results suggest that in this cell type an amplitude-modulated mechanism is involved in repetitive Ca²⁺ release and that Ca²⁺ entry is essential to maintain the repetitive release of Ca²⁺. The results further suggest that the magnitude of Ca²⁺ entry determines the number of unitary stores filled with Ca²⁺ which can synchronously respond to ACh. © 1996 Wiley-Liss, Inc.     

Glutamate triggers intracellular Ca2+ oscillations and nitric oxide release by inducing NAADP- and InsP3-dependent Ca2+ release in mouse brain endothelial cells

The neurotransmitter glutamate increases cerebral blood flow by activating postsynaptic neurons and presynaptic glial cells within the neurovascular unit. Glutamate does so by causing an increase in intracellular Ca²⁺ concentration ([Ca²⁺]_i) in the target cells, which activates the Ca²⁺/Calmodulin-dependent nitric oxide (NO) synthase to release NO. It is unclear whether brain endothelial cells also sense glutamate through an elevation in [Ca²⁺]_i and NO production. The current study assessed whether and how glutamate drives Ca²⁺-dependent NO release in bEND5 cells, an established model of brain endothelial cells. We found that glutamate induced a dose-dependent oscillatory increase in [Ca²⁺]_i, which was maximally activated at 200 μM and inhibited by α-methyl-4-carboxyphenylglycine, a selective blocker of Group 1 metabotropic glutamate receptors. Glutamate-induced intracellular Ca²⁺ oscillations were triggered by rhythmic endogenous Ca²⁺ mobilization and maintained over time by extracellular Ca²⁺ entry. Pharmacological manipulation revealed that glutamate-induced endogenous Ca²⁺ release was mediated by InsP₃-sensitive receptors and nicotinic acid adenine dinucleotide phosphate (NAADP) gated two-pore channel 1. Constitutive store-operated Ca²⁺ entry mediated Ca²⁺ entry during ongoing Ca²⁺ oscillations. Finally, glutamate evoked a robust, although delayed increase in NO levels, which was blocked by pharmacologically inhibition of the accompanying intracellular Ca²⁺ signals. Of note, glutamate induced Ca²⁺-dependent NO release also in hCMEC/D3 cells, an established model of human brain microvascular endothelial cells. This investigation demonstrates for the first time that metabotropic glutamate-induced intracellular Ca²⁺ oscillations and NO release have the potential to impact on neurovascular coupling in the brain.     

Swelling-induced Ca2+ release from intracellular calcium stores in rat submandibular gland acinar cells

The effects of osmotically-induced cell swelling on cytoplasmic free Ca2+ concentration ([Ca2+]i) were studied in acinar cells from rat submandibular gland using microspectrofluorimetry. Video-imaging techniques were also used to measure cell volume. Hypotonic stress (78% control tonicity) caused rapid cell swelling reaching a maximum relative volume of 1.78 +/- 0.05 (n = 5) compared to control. This swelling was followed by regulatory volume decrease, since relative cell volume decreased significantly to 1.61 +/- 0.08 (n = 5) after 10 min exposure to hypotonic medium. Osmotically induced cell swelling evoked by medium of either 78% or 66% tonicity caused a biphasic increase of [Ca2+]i. The rapid phase of this increase in [Ca2+]i was due to release of Ca2 + from intracellular stores, since it was also observed in cells bathed in Ca2+-free solution. The peak increase of [Ca2+]i induced by cell swelling was 3.40 +/- 0.49 (Fura-2 F340/F380 fluorescence ratio, n = 11) and 3.17 +/- 0.43 (n = 17) in the presence and the absence of extracellular Ca2+, respectively, corresponding to an absolute [Ca2+]i of around 1 microm. We found that around two-thirds of cells tested still showed some swelling-induced Ca2+ release (SICR) even after maximal concentrations (10(-5) M - 10(-4) M) of carbachol had been applied to empty agonist-sensitive intracellular Ca2+ stores. This result was confirmed and extended using thapsigargin to deplete intracellular Ca2+ pools. Hypotonic shock still raised [Ca2+]i in cells pretreated with thapsigargin, confirming that at least some SICR occurred from agonist-insensitive stores. Furthermore, SICR was largely inhibited by pretreatment of cells with carbonyl cyanide m-cholorophenyl hydrazone (CCCP) or ruthenium red, inhibitors of mitochondrial Ca2+ uptake. Our results suggest that the increase in [Ca2+]i, which underlies regulatory volume decrease in submandibular acinar cells, results from release of Ca2+ from both agonist-sensitive and mitochondrial Ca2+ stores.     

Kinetics of the secretory response in bovine chromaffin cells following flash photolysis of caged Ca2+.

The kinetics of the secretory response in bovine chromaffin cells following flash photolysis of caged Ca2+ were studied by capacitance (Cm) measurements with millisecond time resolution. After elevation of the internal Ca2+ concentration ([Ca2+]i), Cm rises rapidly with one or more exponentials. The time constant of the fastest component decreases for higher [Ca2+]i (range 3-600 microM) over three orders of magnitude before it saturates at approximately 1 ms. The corresponding maximal rates of secretion can be as fast as 100,000 fF/s or 40,000 vesicles/s. There is a Ca(2+)-dependent delay before Cm rises, which may reflect the kinetics of multiple Ca2+ ions binding to the secretory apparatus. The initial rise in Cm is described by models containing a sequence of two to four single Ca(2+)-binding steps followed by a rate-limiting exocytosis step. The predicted Ca2+ dissociation constant (Kd) of a single Ca(2+)-binding site is between 7 and 21 microM. At [Ca2+]i > 30 microM clear indications of a fast endocytotic process complicate the analysis of the secretory response.     

Millisecond studies of secretion in single rat pituitary cells stimulated by flash photolysis of caged Ca2+.

To study the final steps in the secretory pathway of rat pituitary melanotrophs, we have monitored changes in cell surface area due to exocytosis after flash photolysis of caged Ca2+. A step increase in cytosolic [Ca2+] to 45-125 microM triggers three phases of exocytic secretion. A small cohort of a few hundred vesicles is exocytosed in 40 ms in a secretory burst with a peak rate of 17,000 vesicles/s. Next, 2700 more vesicles are released in a slower phase that is complete within 400-1000 ms. Finally, vesicles continue to be released slowly (500 vesicles/s) for > 8s. The approach described provides a way to identify and monitor the final steps in the secretory pathway at millisecond resolution. That a small portion of secretory vesicles can be released much faster than all others suggests that these vesicles are functionally equivalent to those at the presynaptic active zone of a neuron. Their release would be fast enough to be temporally correlated with single action potentials.     

InsP3 receptor type 2 and oscillatory and monophasic Ca2+ transients in rat adrenal chromaffin cells

Muscarinic receptor stimulation induced oscillatory and monophasic Ca(2+) transients in rat adrenal chromaffin cells in the absence of external Ca(2+). As this Ca(2+) mobilization may be mediated by InsP(3), we first explored types of InsP(3) receptors and their intracellular distribution in chromaffin cells. The InsP(3) receptor type 1 was not immunodetected in precipitates of adrenal medulla homogenates and in dissociated adrenal chromaffin cells, whereas an anti-type 3 mAb recognized a faint band with about 250 kDa, but no significant immunoreaction was visible in chromaffin cells. The anti-type 2 mAb strongly detected a band with about 220 kDa and the immunoreaction was observed perinuclearly and at the cell periphery. These results indicate that InsP(3) receptor type 2 is predominant in chromaffin cells. The oscillatory and monophasic Ca(2+) transients were reproduced in simulation based on a three-state kinetic model (shut, open, and inactivated states). Ca(2+) ions were found experimentally and theoretically to turn over rapidly between stores and the cytosol during stimulation. The results suggest that InsP(3) receptor type 2 is responsible for both oscillatory and monophasic Ca(2+) transients and that change in mode of Ca(2+) responses may be accounted for by the kinetic property of the type 2 receptor.     

Modulation of Ca2+ mobilization by protein kinase C in rat submandibular acinar cells

The effects of protein kinase C (PKC) activation and inhibition on the inositol 1,4,5-trisphosphate (IP3) and cytosolic Ca2+ ([Ca2+]i) responses of rat submandibular acinar cells were investigated. IP3 formation in response to acetylcholine (ACh) was not affected by the PKC activator phorbol 12-myristate 13-acetate (PMA), nor by the PKC inhibitor calphostin C (CaC). The ACh-elicited initial increase in [Ca2+]i in the absence of extracellular Ca2+ was not changed by short-term (0.5 min) exposure to PMA, but significantly reduced by long-term (30 min) exposure to PMA, and also by pre-exposure to the PKC inhibitors CaC and chelerythrine chloride (ChC). After ACh stimulation, subsequent exposure to ionomycin caused a significantly (258%) larger [Ca2+]i increase in CaC-treated cells than in control cells. However, pre-exposure to CaC for 30 min did not alter the Ca2+ release induced by ionomycin alone. These results suggest that the reduction of the initial [Ca2+]i increase is due to an inhibition of the Ca2+ release mechanism and not to store shrinkage. The thapsigargin (TG)-induced increase in [Ca2+]i was significantly reduced by short-term (0.5 min), but not by long-term (30 min) exposure to PMA, nor by pre-exposure to ChC or CaC. Subsequent exposure to ionomycin after TG resulted in a significantly (70%) larger [Ca2+]i increase in PMA-treated cells than in control cells, suggesting that activation of PKC slows down the Ca2+ efflux or passive leak seen in the presence of TG. Taken together, these results indicate that inhibition of PKC reduces the IP3-induced Ca2+ release and activation of PKC reduces the Ca2+ efflux seen after inhibition of the endoplasmic Ca2+-ATPase in submandibular acinar cells.     

CIB1, a ubiquitously expressed Ca2+-binding protein ligand of the InsP3 receptor Ca2+ release channel

A family of Ca(2+)-binding proteins (CaBPs) was shown to bind to the inositol 1,4,5-trisphosphate receptor (InsP(3)R) Ca(2+) release channel and gate it in the absence of InsP(3), establishing them as protein ligands (Yang, J., McBride, S., Mak, D.-O. D., Vardi, N., Palczewski, K., Haeseleer, F., and Foskett, J. K. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 7711-7716). However, the neuronally restricted expression of CaBP and its inhibition of InsP(3)R-mediated Ca(2+) signaling when overexpressed (Kasri, N. N., Holmes, A. M., Bultynck, G., Parys, J. B., Bootman, M. D., Rietdorf, K., Missiaen, L., McDonald, F., De Smedt, H., Conway, S. J., Holmes, A. B., Berridge, M. J., and Roderick, H. L. (2004) EMBO J. 23, 312-321; Haynes, L. P., Tepikin, A. V., and Burgoyne, R. D. (2004) J. Biol. Chem. 279, 547-555) have raised questions regarding the functional implications of this regulation. We have discovered the Ca(2+)-binding protein CIB1 (calmyrin) as a ubiquitously expressed ligand of the InsP(3)R. CIB1 binds to all mammalian InsP(3)R isoforms in a Ca(2+)-sensitive manner dependent on its two functional EF-hands and activates InsP(3)R channel gating in the absence of InsP(3). In contrast, overexpression of CIB1 or CaBP1 attenuated InsP(3)R-dependent Ca(2+) signaling, and in vitro pre-exposure to CIB1 reduced the number of channels available for subsequent stimulation by InsP(3). These results establish CIB1 as a ubiquitously expressed activating and inhibiting protein ligand of the InsP(3)R.     

Ca2+ oscillations in hepatocytes do not require the modulation of InsP3 3-kinase activity by Ca2+

Vacuole fusion requires Sec18p-dependent acylation of the armadillo-repeat protein Vac8p that has been isolated with cis-SNARE complexes. To gain more insight into the mechanism of acylation, we analyzed the palmitoylation reaction on isolated vacuoles or in vacuolar detergent extracts. Recombinant Vac8p is palmitoylated when added to vacuoles and is anchored to membranes after modification. The palmitoyl acyltransferase (PAT) extracted from vacuolar membranes is functional in detergent extracts and shows all characteristics of an enzymatic activity: It modifies exogenous Vac8p in a temperature-, dose- and time-dependent manner, and is sensitive to bromo-palmitate, a known inhibitor of protein palmitoylation in vivo. Importantly, PAT is specific for palmitoyl-CoA, since myristoyl- and stearyl-CoA can compete with labeled Pal-CoA only at 10-fold higher amounts.     

Activation of ryanodine receptors by flash photolysis of caged Ca2+.

Flash photolysis of DM-nitrophen generates an extremely large [Ca2+] transient ("Ca2+ spike") at the start of each Ca2+ "step." The Ca2+ spike greatly increases the speed of activation of the ryanodine receptor channel ("supercharging") and could be responsible for apparent channel adaptation.     

The type 2 inositol (1,4,5)-trisphosphate (InsP3) receptor determines the sensitivity of InsP3-induced Ca2+ release to ATP in pancreatic acinar cells

Calcium release through inositol (1,4,5)-trisphosphate receptors (InsP(3)R) is the primary signal driving digestive enzyme and fluid secretion from pancreatic acinar cells. The type 2 (InsP(3)R2) and type 3 (InsP(3)R3) InsP(3)R are the predominant isoforms expressed in acinar cells and are required for proper exocrine gland function. Both InsP(3)R2 and InsP(3)R3 are positively regulated by cytosolic ATP, but InsP(3)R2 is 10-fold more sensitive than InsP(3)R3 to this form of modulation. In this study, we examined the role of InsP(3)R2 in setting the sensitivity of InsP(3)-induced Ca(2+) release (IICR) to ATP in pancreatic acinar cells. IICR was measured in permeabilized acinar cells from wild-type (WT) and InsP(3)R2 knock-out (KO) mice. ATP augmented IICR from WT pancreatic cells with an EC(50) of 38 mum. However, the EC(50) was 10-fold higher in acinar cells isolated from InsP(3)R2-KO mice, indicating a role for InsP(3)R2 in setting the sensitivity of IICR to ATP. Consistent with this idea, heterologous expression of InsP(3)R2 in RinM5F cells, which natively express predominately InsP(3)R3, increased the sensitivity of IICR to ATP. Depletion of ATP attenuated agonist-induced Ca(2+) signaling in WT pancreatic acinar cells. This effect was more profound in acinar cells prepared from InsP(3)R2-KO mice. These data suggest that the sensitivity of IICR to ATP depletion is regulated by the particular complement of InsP(3)R expressed in an individual cell. The effects of metabolic stress on intracellular Ca(2+) signals can therefore be determined by the relative amount of InsP(3)R2 expressed in cells.     

Regulation of InsP3 receptor activity by neuronal Ca2+-binding proteins

Inositol 1,4,5-trisphosphate receptors (InsP(3)Rs) were recently demonstrated to be activated independently of InsP(3) by a family of calmodulin (CaM)-like neuronal Ca(2+)-binding proteins (CaBPs). We investigated the interaction of both naturally occurring long and short CaBP1 isoforms with InsP(3)Rs, and their functional effects on InsP(3)R-evoked Ca(2+) signals. Using several experimental paradigms, including transient expression in COS cells, acute injection of recombinant protein into Xenopus oocytes and (45)Ca(2+) flux from permeabilised COS cells, we demonstrated that CaBPs decrease the sensitivity of InsP(3)-induced Ca(2+) release (IICR). In addition, we found a Ca(2+)-independent interaction between CaBP1 and the NH(2)-terminal 159 amino acids of the type 1 InsP(3)R. This interaction resulted in decreased InsP(3) binding to the receptor reminiscent of that observed for CaM. Unlike CaM, however, CaBPs do not inhibit ryanodine receptors, have a higher affinity for InsP(3)Rs and more potently inhibited IICR. We also show that phosphorylation of CaBP1 at a casein kinase 2 consensus site regulates its inhibition of IICR. Our data suggest that CaBPs are endogenous regulators of InsP(3)Rs tuning the sensitivity of cells to InsP(3).     

Kinetics of Ca2+ release and contraction induced by photolysis of caged D-myo-inositol 1,4,5-trisphosphate in smooth muscle. The effects of heparin, procaine, and adenine nucleotides.

The kinetics of Ca2+ release and contraction induced by photolytic release of inositol 1,4,5-trisphosphate (InsP3) were determined in permeabilized smooth muscle. The rate of Ca2+ release was half-maximal at 1 microM InsP3. The concentration-dependent delay of Ca2+ release at saturating InsP3 concentration was approximately 10 ms and within the uncertainty of the measurements. The relationship between the delay and InsP3 concentration showed no evidence of a high level (n = 4 or higher) of cooperativity but could not distinguish between no cooperativity (n = 1) or a low level (n = 2) of cooperativity. Submaximal [InsP3] caused only partial Ca2+ release from the InsP3-sensitive stores. InsP3-induced Ca2+ release was markedly potentiated by ATP or by adenosine 5'-(beta,gamma-methylene-triphosphate), but neither the rate nor the amplitude of release was significantly affected by procaine (2-5 mM). Heparin increased the delay between photolysis and Ca2+ release, indicating that the off rate of inert ligand(s) bound to InsP3 receptors may contribute to the physiological delay in Ca2+ release. There was a much longer (370 ms +/- 45 S.E.) delay between the rise of Ca2+ and force development, presumably reflecting events preceding and associated with myosin light chain phosphorylation.     

Cortical granule translocation during maturation of starfish oocytes requires cytoskeletal rearrangement triggered by InsP3-mediated Ca2+ release

Cortical granules (secretory vesicles located under the cortex of mature oocytes) release their contents to the medium at fertilization. Their exocytosis modifies the extracellular environment, blocking the penetration of additional sperm. The granules translocate to the surface during the maturation process, and it has been suggested that they move to the cortex via cytoskeletal elements. In this paper we show that the increase in intracellular Ca2+, which the maturing hormone 1-methyladenine (1-MA) induces in starfish through the activation of inositol 1,4, 5-trisphosphate (InsP3) receptors, triggers changes in filamentous actin, which then direct the correct movement and reorientation of the cortical granules and the elevation of the fertilization envelope.     

Characterization of two different Ca2+ uptake and IP3-sensitive Ca2+ release mechanisms in microsomal Ca2+ pools of rat pancreatic acinar cells

We have examined the effect of the Ca²⁺ (Mg²⁺)-ATPase inhibitors thapsigargin (TG) and vanadate on ATP-dependent ⁴⁵Ca²⁺ uptake into IP₃-sensitive Ca²⁺ pools in isolated microsomes from rat pancreatic acinar cells. The inhibitory effect of TG was biphasic. About 40–50% of total Ca²⁺ uptake was inhibited by TG up to 10 nm (apparent K_i4.2 nm, Ca²⁺ pool I). An additional increase of inhibition up to 85–90% of total Ca²⁺ uptake could be achieved at 15 to 20 nm of TG (apparent K_i12.1 nm, Ca²⁺ pool II). The rest was due to TG-insensitive contaminating plasma membranes and could be inhibited by vanadate (apparent K_i10 m). In the absence of TG, increasing concentrations of vanadate also showed two phases of inhibition of microsomal Ca²⁺ uptake. About 30–40% of total Ca²⁺ uptake was inhibited by 100 m of vanadate (apparent K_i18 m, Ca²⁺ pool II). The remaining 60–70% could be inhibited either by vanadate at concentrations up to 1 mm (apparent K_i300 m) or by TG up to 10 nm (Ca²⁺ pool I). The amount of IP₃-induced Ca²⁺ release was constant at 25% over a wide range of Ca²⁺ filling. About 10–20% remained unreleasable by IP₃. Reduction of IP₃ releasable Ca²⁺ in the presence of inhibitors showed similar dose-response curves as Ca²⁺ uptake (apparent K_i 3.0 nm for IP₃-induced Ca²⁺ release as compared to 4.2 nm for Ca²⁺ uptake at TG up to 10 nm) indicating that the highly TG-sensitive Ca²⁺ pump fills the IP₃-sensitive Ca²⁺ pool I. At TG concentrations >10 nm which blocked Ca²⁺ pool II the apparent K_i values were 11.3 and 12.1 nm, respectively. For inhibition by vanadate up to 100 m the apparent K_i values were 18 m for Ca²⁺ uptake and 7 m for Ca²⁺ release (Ca²⁺ pool II). At vanadate concentrations up to 1 mm the apparent K_i values were 300 and 200 m, respectively (Ca²⁺ pool I). Both Ca²⁺ pools I and II also showed different sensitivities to IP₃. Dose-response curves for IP₃ in the absence of inhibitors (control) showed an apparent K_m value for IP₃ at 0.6 m. In the presence of TG (inhibition of Ca²⁺ pool I) the curve was shifted to the left with an apparent K_m for IP₃ at 0.08 m. In the presence of vanadate (inhibition of Ca²⁺ pool II), the apparent K_m for IP₃ was 2.1 m. These data allow the conclusion that there are at least three different Ca²⁺ uptake mechanisms present in pancreatic acinar cells: TG- and IP₃ insensitive but highly vanadate-sensitive Ca²⁺ uptake occurs into membrane vesicles derived from plasma membranes. Two Ca²⁺ pools with different TG-, vanadate- and IP₃-sensitivities are most likely located in the endoplasmic reticulum at different cell sites, which could have functional implications for hormonal stimulation of pancreatic acinar cells.This work was supported by the Deutsche Forschungsgemeinschaft, Sonderforschungsbereich 246. The authors wish to thank Dr. KlausDieter Preuß for valuable discussions and Mrs. Gabriele Mörschbächer for excellent secretarial help.