Low Cell Density Cultures Research Articles

Photomechanical movements of cultured embryonic photoreceptors: regulation by exogenous neuromodulators and by a regulable source of endogenous dopamine

In the retina of nonmammalian vertebrates, light regulates photoreceptor morphology by causing rod photoreceptor elongation and cone photoreceptor contraction. The opposite photomechanical movements occur in the dark, and proceed with a circadian rhythm in many species in vivo. Using dissociated cultures of embryonic chick retina cells, we have recently demonstrated that photoreceptor cells that differentiate in vitro acquire the capacity of responding to light/dark cycles with photomechanical movements (Stenkamp and Adler, 1993). Here we report that the putative neuromodulators melatonin and dopamine can mimic the effects of darkness and light, respectively, on in vitro photomechanical movement. Pharmacological studies showed that dopamine appears to function by means of a D2-type receptor negatively coupled to adenylate cyclase. The effects of light on the cultured photoreceptors were inhibited by dopamine D2 receptor antagonists, and were attenuated by the dopaminergic neurotoxin 6-hydroxydopamine and by the dopamine synthesis inhibitor alpha-methyl-p-tyrosine. The possible existence of an endogenous source of dopamine in the cultures was also suggested by the presence of tyrosine hydroxylase-like immunoreactivity, and of an Na(+)-dependent mechanism for the accumulation of 3H-dopamine, which was predominantly associated with nonphotoreceptor cells. Additionally, 3H-dopamine release occurred in vitro through a Ca(2+)-dependent mechanism, as well as through reverse function of a nomifensine-sensitive dopamine transporter. Both of these putative release mechanisms appeared to be regulated by light and by melatonin, suggesting a mechanism whereby the putative dopaminergic cells may interact with other cells present in the cultures. These studies suggest that complex paracrine neuromodulatory mechanisms can differentiate in low-density embryonic cell culture, that dopaminergic activities exist in vitro, and that they are important for mediating photomechanical movements.

Polyamines promote neurite elongation of cultured rat hippocampal neurons

The effects of spermine, spermidine and putrescine on the survival and morphology of hippocampal neurons obtained from rat embryos were investigated using low cell density culture. The number of surviving neurons gradually decreased in the control cultures, while addition of spermine significantly promoted neuronal survival. The survival-promoting effect of spermine was seen maximally at a concentration of 10 −8 M. On the other hand, spermidine and putrescine showed no significant effect on neuronal survival. The axon length of neurons cultured in the presence of spermine was significantly longer than that in the medium with no added spermine, indicating that spermine promotes process elongation. The concentration-effect curve for spermine was bellshaped with the maximum effect at a concentration of 10 −8 M. However, the number of processes per soma, the elongation of dendrites and the process branching in axons were not significantly affected by the presence of spermine. Addition of spermidine and putrescine also promoted the axon elongation of cultured neurons, without affecting the branching number. The effects of spermidine and putrescine on neuronal growth were very similar to that of spermine in terms of the effective concentrations and the patterns of morphological changes. Since spermine specifically promoted neuronal survival while three polyamines affected the neuronal morphology in a similar manner, it is possible that the effects of polyamines on neuronal survival and neurite elongation are mediated by different mechanisms.

Reduction of neurite extension by piperine, examined on hippocampal and septal neurons in serum-free cultures.

The presence of piperine in a serum-free medium for 72 h markedly inhibited the neurite extension of cultured hippocampal and septal neurons under both high and low density cell culture conditions (initial cell density was 10(5) cells/cm2 and 2.5-5 x 10(3) cells/cm2, respectively). The average length of neurites decreased over a concentration range of piperine (12.5-100 microM), and the distribution of neurite population shifted towards the shorter neurite lengths whereas an increase in the soma size was observed only with high piperine concentrations (75 and 100 microM, in a high density cell culture). This indicates that the piperine effect was relatively selective to the neurite extension. The neurite extension in low density cell cultures was considerably more susceptible to piperine than that in high density cell cultures (hippocampus: EC50 = 39 versus 113 microM, septum: EC50 = 48 versus 101 microM, in the low and high density cell cultures, respectively). This difference may be due to a lack of neurotrophic supports from non-neuronal cells under the low density cell culture condition. These findings suggest that piperine, in addition to its cytotoxic effect on the neuronal survival, suppresses the neurite extension in developing neurons.

Malonylginsenoside Rb1 potentiates nerve growth factor (NGF)-induced neurite outgrowth of cultured chick embryonic dorsal root ganglia.

Effects of malonylginsenoside Rb1 (GRb1-m) isolated from dried root of Panax ginseng C. A. Meyer (Araliaceae) on neuronal survival and neurite outgrowth were compared with those of ginsenoside Rb1 (GRb1) using organ culture of chick embryonic dorsal root ganglia (DRG) and cell culture of DRG neurons. In the organ culture, nerve growth factor (NGF) showed neurite outgrowth promoting effect. GRb1-m (30 microM) significantly potentiated this NGF-induced neurite outgrowth with similar potency to that of GRb1 (30 microM). Low density cell culture of DRG neurons was employed to minimize the contribution of contaminating non-neuronal cells. NGF (10 ng/ml) prolonged duration of neuronal survival. Neither GRb1-m nor GRb1 (1-30 microM) changed the prolongation effect of NGF, nor did NGF show a significant effect on the neurite elongation and re-elongation after axotomy by laser beam irradiation. However, GRb1-m (10 microM) potentiated initial neurite elongation when co-treated with NGF. In the process of re-elongation of neurites, GRb1-m (1, 30 microM) also promoted it in the presence of NGF. These results suggest, first, that GRb1-m potentiates NGF-induced neurite outgrowth of chick embryonic DRGs and DRG neurons, but behaves in a slightly different manner from GRb1, and, second, that the effects of the two saponins may work primarily on neurons causing the potentiation of NGF-induced neurite outgrowth.

Piperine, a pungent alkaloid, is cytotoxic to cultured neurons from the embryonic rat brain.

The present study reveals that piperine, a pungent alkaloid present in various Piper species, is cytotoxic to cultured brain neurons. Exposure to piperine (12.5-100 microM) for 72 h caused a concentration-dependent reduction in the survival of primary cultured neurons from various regions of the embryonic rat brain under high density cell culture conditions. There were relative regional differences in the susceptibility to cytotoxic effects of piperine in which septum and hippocampus showed higher vulnerability among the eight regions. The primary cultures of septal and hippocampal neurons under low density cell culture condition were performed to evaluate the contribution of non-neuronal cells. The concentration-response profiles in both high and low density cell culture conditions were comparable (septum: EC50 = 43 and 27 microM, hippocampus: EC50 = 50 and 44 microM, under high and low density cell culture conditions, respectively) suggesting a minor role of non-neuronal cells on cytotoxicity of piperine to developing neurons.

Interleukin-2 promotes survival and neurite extension of cultured neurons from fetal rat brain

We investigated the effects of interleukin-2 (IL-2) on the survival and morphology of primary cultured neurons from fetal rat brain. Addition of recombinant human IL-2 significantly supported the survival of brain neurons in high cell density culture, but did not show any effect on the neuronal survival in low cell density culture. In addition, IL-2 significantly promoted the neurite elongation and branching of hippocampal neurons in low cell density culture. These results suggest that IL-2 supports neuronal survival indirectly and promotes neuritogenesis by directly acting on brain neurons.

Calcium-destabilizing and neurodegenerative effects of aggregated β-amyloid peptide are attenuated by basic FGF

The mechanisms that contribute to neuronal degeneration in Alzheimer's disease (AD) are not understood. Abnormal accumulations of β-amyloid peptide (βAP) are thought to be involved in the neurodegenerative process, and recent studies have demonstrated neurotoxic actions of βAPs. We now report that the mechanism of βAP-mediated neurotoxicity in hippocampal cell culture involves a destabilization of neuronal calcium homeostasis resulting in elevations in intracellular calcium levels ([Ca 2+] i) that occur during exposure periods of 6 hr to several days. Both the elevations of [Ca 2+] i and neurotoxicity were directly correlated with aggregation of the peptide as assessed by βAP immunoreactivity and confocal laser scanning microscopy. Exposure of neurons to βAP resulted in increased sensitivity to the [Ca 2+] i-elevating and neurodegenerative effects of excitatory amino acids. Moreover, [Ca 2+] i responses to membrane depolarization and calcium ionophore were greatly enhanced in βAP-treated neurons. Neurons in low cell density cultures were more vulnerable to βAP toxicity than were neurons in high cell density cultures. Basic fibroblast growth factor (bFGF), but not nerve growth factor (NGF), significantly reduced both the loss of calcium homeostasis and the neuronal damage otherwise caused by βAP. In AD, βAP may endanger neurons by destabilizing calcium homeostasis and bFGF may protect neurons by stabilizing intracellular calcium levels. Aggregation of βAP seems to be a major determinant of its [Ca 2+] i-destabilizing and neurotoxic potency.

Expression of the IL-6 gene induces differentiation of a human monoclonal EBV-transformed B cell line

To study the role of interleukin (IL)-6 as a growth and differentiation factor for Epstein-Barr virus (EBV)-transformed B lymphocytes, we transfected the cDNA coding for human IL-6 in a monoclonal IgG1-secreting EBV B cell line. Two independent clones were selected that constitutively secreted high amounts of IL-6. These clones showed enhanced levels of IL-6 and tumor necrosis factor alpha secretion when compared to non-IL-6 transfected controls. Moreover, they could efficiently be recovered from low cell density cultures in limiting dilutions when plated on a feeder layer of heterologous EBV B cells. IL-6-induced phenotypical changes comprised a significant rise in immunoglobulin secretion levels and enhanced membrane expression of CD25 (the beta chain of the IL-2 receptor) and of the B cell differentiation antigen CD40. IL-6-dependent down modulation of CD38 and of the adhesion structure VLA4 were also observed. Our data support the notion that IL-6 can serve as an growth and differentiation factor for EBV B cells.

Cell Density-Dependent DNA Fragmentation and Its Suppression by Heparin in Primary Culture of Adult Rat Hepatocytes

We show here that the internucleosomal DNA fragmentation, which is a biochemical hallmark of apoptosis, was induced in a cell density-dependent manner in primary culture of adult rat hepatocytes. This DNA fragmentation could be suppressed by a gene expression inhibitor, indicating the active nature of this process. Moreover, the viability changes in high and low cell density cultures showed a tendency corresponding with the incidence of the DNA fragmentation in them. These results suggest that in hepatocytes there may be a cell density-dependent apoptosis mechanism. In this report, we also show that heparin could suppress this DNA fragmentation with high specificity, and the cell death to some extent.

Thrombin Promotes Aortic Endothelial Cell Spreading and Microfilament Formation in Nonconfluent Monolayer Cultures

Thrombin is present at sites of vascular injury. The objective of this study was to determine if thrombin may regulate endothelial repair. To address this our study was designed to determine the effect of thrombin on cell shape and microfilament distribution of porcine aortic endothelial cells in low-density cultures. Since these cells grow as islands of cells, low-density cultures serve as a model for conditions in which reendothelialization occurs from several foci following prominent patchy endothelial denudation. Thrombin incubation for 1, 2, and 4 hr at 0.5, 4, and 8 U/ml caused a significant increase in both the surface area of endothelial cells and the amount of microfilament bundles occupying the cell surface area. The increase in microfilament bundles was more than would be expected from the increased surface area. However, the pattern of distribution of microfilaments and of vinculin adhesion plaques within the cells was not altered by thrombin. Attempts at understanding the molecular mechanisms involved were undertaken. The role of protein kinase C was studied. Phorbol 12-myristate 13-acetate (PMA) caused cell spreading as well but no increase in microfilaments within the first 2 hr of incubation. Some cells showed mild disruption of microfilaments. Following 24 hr incubation with PMA, which reduces protein kinase C, the cells showed retraction and arborization along with thinning and disruption of microfilaments. Removal of PMA and addition of thrombin reversed these changes to normal within 2 hr while washout of PMA on its own did not alter cell shape or microfilaments. Thrombin also partially reversed the effects of H-7, a protein kinase C inhibitor, which on its own also caused cell retraction and disruption of microfilaments. Neomycin sulfate did not alter the thrombin effect suggesting that the participation of the phosphotidylinositol system is not directly required. In conclusion, thrombin promotes endothelial activities which are associated with long-term endothelial repair; however, these effects do not appear to be directly due to the involvement of conventional isoforms of protein kinase C or the phosphotidylinositol system.

Microglial conditioned medium promotes survival and development of cultured mesencephalic neurons from embryonic rat brain.

We previously reported that microglial conditioned medium (Mic-CM) has a neurotrophic effect on cultured rat neocortical neurons [Nakajima et al. (1989): Biomed Res 10:411-423]. In order to investigate the interaction between microglia and neurons in more detail, we determined the effects of Mic-CM on the primary cultured mesencephalic neurons from 16-day embryonic rats. The addition of Mic-CM to the culture medium significantly enhanced the survivability of neurons and promoted neurite extension in a low cell-density culture condition. In a high cell-density culture condition, Mic-CM markedly increased dopamine uptake, which was quantified by assessing the specific [3H]dopamine uptake, and also increased the dopamine content of cultured cells. Furthermore, the number of mesencephalic dopaminergic neurons, which was determined by quantitative analysis of tyrosine hydroxylase (TH)-immunoreactive cells, increased significantly in the presence of Mic-CM. These results suggest that Mic-CM enhances survival or maturation of TH-positive neurons present in cultures of the embryonic mesencephalon and that these neurotrophic effects may be due to a diffusible factor(s) from microglia.

The cytotoxic action of lymphokine activated killer cells upon the human glioma cell line U251 is stimulated by bispecific monoclonal antibody (MoAb) constructs.

The ability of IL-2 stimulated mononuclear cells to kill the human glioblastoma cell line U251 has been investigated. Highest cytotoxic activity was generated in low cell density cultures incubated for 15 days with 250-1000 U/ml IL-2. Sub-optimal killing was noted, with cells only exposed to IL-2 for three days. Under the latter conditions, bispecific monoclonal antibodies (MoAbs) of either anti-CD3 or anti-CD16 and an anti-NCAM MoAb stimulated LAK cell activity markedly. Anti-CD16 conjugates were found more effective than anti-CD3 and (Fab')2 constructs more efficacious than those made with whole Ig molecules. Maximal stimulation of LAK cell activity was noted with bispecific MoAbs. Little effect was observed with either single or mixtures of monomeric MoAbs. Furthermore, no effect of bispecific MoAbs was observed when target cells lacked expression of NCAM. These results could be of clinical importance as it is not always feasible to screen LAK cells for optimal activity before administration to patients. Whilst bispecific MoAbs have no effect on optimally stimulated LAK cells, they are not inhibitory and can stimulate killing under sub-optimal IL-2 stimulation.

Internalization efficiency of the transferrin receptor

Quantitative ultrastructural and biochemical methods have allowed us to obtain a coherent set of data on the internalization efficiency of the transferrin receptor (TfR). In confluent cell cultures we find that (1) the initial internalization rate of transferrin is approximately 10% per minute, and (2) around 10% of cell-surface TfRs are present in coated pits. From these data a lifetime of coated pits of ca. 1 min is derived. Furthermore, we show that coated pits constitute 1.1–1.4% of the plasma membrane area in confluent cell cultures. Thus, the TfR is concentrated six- to ninefold in coated pits compared to resident plasma membrane proteins. Moreover, we show that the concentration of TfRs in coated pits is cell density dependent, since only around 5% of the receptors are present in coated pits in low-density cultures. Correspondingly, the internalization of TfRs in high-density cell cultures is roughly twice as efficient as that in low-density cell cultures. The reduced TfR internalization efficiency at low cell density is accounted for by a concomitant decrease to 0.55% in the relative surface area occupied by coated pits.

Heterogeneous expression of car☐ypeptidase E and proenkephalin mRNAs by cultured astrocytes

Cultured astrocytes have been found to express neuropeptides, neuropeptide processing enzymes and their mRNAs. Although astrocytes were shown to display regional variation in their expression of these mRNAs, it was unclear whether all astrocytes cultured from the same brain region express similar mRNA levels or if this expression is heterogeneous. We examined the individual heterogeneity of astrocytes cultured from several brain regions by in situ hybridization. Astrocytes derived from the frontal cortex, hypothalamus and cerebellum of neonatal rat brains were cultured for 3 weeks and then analyzed by in situ hybridization using 35S-labeled cRNA probes to car☐ypeptidase E (CPE), proenkephalin (PE), and cyclophilin (1B15) mRNAs and an oligomeric probe complimentary to the first 45 bases of rat 18S rRNA. Frequency histograms generated by counting the grains produced over emulsion-coated cells demonstrated populations of astrocytes expressing high levels of mRNA for CPE and PE and a population expressing low to background levels of these mRNAs. In contrast, all cultured astrocytes expressed high levels of 1B15 mRNA and 18S rRNA. The percentage of cultured astrocytes expressing high levels of CPE mRNA was 42% for frontal cortex astrocytes and 23% for cerebellar astrocytes. While the percentages of cultured astrocytes expressing high levels of PE mRNA varied slightly between brain regions (24–28%), the level of PE mRNA expression per cell showed greater variation between regions. The effect of culture density on the expression of PE mRNA was also examined. Approximately 55% of the cells in low density cultures expressed PE mRNA, while only 5–10% of the cells in high density cultures expressed this mRNA. These results indicate that cultured astrocytes display individual heterogeneity with regard to neuropeptide biosynthesis and that the expression of neuropeptides by these cells is regulated.

Glucocorticoid Induction of the Adipocyte Clone 5 Gene Requires High Cell Density

We have previously reported the identification of a glucocorticoid and cell-density inducible gene (clone 5) that appears to play a regulatory role in adipocyte differentiation. In the current studies we have more carefully investigated the interplay between cell growth or differentiation and the glucocorticoid responsiveness of clone 5 RNA. We find that inducibility by steroid hormone is independent of the differentiated state but surprisingly, occurs only when cells are at high cell density. In subconfluent TA1 cells clone 5 RNA is refractory to induction whereas well characterized glucocorticoid-inducible promoters such as the those from mouse mammary tumor virus and alpha-1-acid glycoprotein are induced under all conditions. These results point to the necessity of ancillary factors, absent in low density cell culture, that are required for glucocorticoid responsiveness of the clone 5 gene.

Interleukin-6 production by murine B cells and B cell lines

We have analyzed interleukin (IL)-6 gene transcription and IL-6 secretion by murine B cells in vitro. Mitogenic doses of lipopolysaccharide (LPS) or LPS in combination with F(ab′) 2 goat anti-mouse IgM antibodies (GAMμ), but not GAMμ alone, induced B cells to synthesize and release IL-6. In time course experiments, the accumulation of IL-6 mRNA was first detectable at 24–36 hr of culture and the levels were maintained through 60 hr; these kinetics correlated well with increases in supernatant IL-6 levels and were coincident with vigorous cell cycle activity. We also analyzed constitutive and LPS-induced IL-6 gene expression by the murine B cell lines: 70Z/3, 38C-13, WEHI-231, X16C, WEHI-279, and BCL 1. Only the WEHI-279 and BCL 1 lines produced detectable IL-6 constitutively, and the BCL 1 cells could be further induced by treatment with LPS. Of the remaining cell lines, only WEHI-231 and X16C could be stimulated with LPS to produce IL-6. To evaluate whether IL-6 could influence proliferation and Ig secretion by the cell lines, low cell density cultures were established in the presence of various doses of human rIL-6 and were assessed over time for levels of [ 3H]thymidine uptake and supernatant Ig. Under these conditions, IL-6 had no effect on either cell function.

Direct measurements of nanomolar nitrate uptake by the marine diatom Phaeodactylum tricornutum (Bohlin). Implications for studies of oligotrophic ecosystems

A new automated procedure for nanomolar nitrate analysis was applied to the study of uptake of low nitrate concentrations (< 100 ngat l−1) by phytoplankton. The precision of this analytical method (± 3 ngat l−1) made it possible to monitor the absorption of very low quantities of nitrate over short term periods by a low cell-density culture of the marine diatom Phaeodactylum tricornutum, where the levels of particulate nitrogen and chlorophyll were equivalent to those found in oligotrophic areas (0.5 µgat N l−1 and 0.4 µg l−1 respectively). By continuous monitoring of nitrate disappearance from the culture medium, we are able to describe accurately the transient uptake responses of the diatom after a spike addition of trace quantities of nitrate and thus to provide new information on the still largely unknown small-scale phenomenon of pulsed nitrate supply in the upper layer of stratified oceans and rapid uptake of these nitrate patches by phytoplankton. The results show that a N-limited culture of Phaeodactylum tricornutum is immediately capable of taking up trace quantities of nitrate (< 100 ngat l−1) at high rates (0.10–0.14 h−1) . These initial rates are one order of magnitude higher than the theoretical rates calculated from the Michaelis-Menten equation and are close to the level of V max (0.15 h−1) obtained when cells are exposed to saturating nitrate concentrations. This rapid initial uptake would be a considerable advantage in oligotrophic areas where nanomolar nitrate supply is thought to be episodic. The present results suggest that phytoplankton evolve adaptations to utilize the available nitrate at the spatial and temporal scales at which it occurs. On the other hand, we can consider this physiological adaptation as evidence of nitrate pulses in the field which would invalidate the steady-state approach to the oligotrophic ecosystems.

Balanced nutrient fortification enables high‐density hybridoma cell culture in batch culture

Cells of an in vitro culture system are not the same as for an in vivo system, metabolically and physiologically; ineffective utilization of nutrients occurs by cells in vitro. Therefore, a simpler approach is needed to examine closely and overcome differences between in vivo and in vitro cells.Recognizing the ineffectiveness of nutrient utilization in vitro, we have constructed, a balanced, fortified high-density medium based on RPMI 1640 medium previously optimized for relatively low-density cell culture. The high-density medium was used to cultivate a hybridoma line in a batch spinner flask culture. In this fortified medium, a hybridoma cell line 2c3.1 was cultivated to near 1 x 10(7) cells/mL in batch suspension culture. During the culture, glucose, glutamine, and 10 essential amino acids of concentrations five times richer than normal in the medium were almost thoroughly consumed. Combined analysis of these consumption profiles reveals that the balanced, fortified nutrient supply contributes much to cellular activity to overcome the limitations of in vitro cellular growth. Intermediate metabolites, such as ammonium ion and lactic acid, were produced over concentrations reported until now to be inhibitory. This observation suggests that the major conclusive factor against cellular growth over the critical cell density is not so-called inhibitory metabolites. As a result of the high-density culture, 5-8 times higher production of a monoclonal antibody for hepatitis B surface antigen (anti-HBs) was obtained.Active cellular consumption of all the essential nutrients and the corresponding production of MAb strongly support the potential of our approach to overcome the growth limitation of cells in vitro and to obtain high-density hybridoma cell culture.

Direct effect of cyclosporin A on CD8+ T lymphocytes in a primary mixed lymphocyte reaction

Cyclosporin A (CsA) blocks the development of cytotoxic T lymphocytes (CTL) in a primary mixed lymphocyte reaction (MLR) when added during the first few days of culture. We found that addition of recombinant IL-2 (rIL-2) or an IL-2-containing supernatant (2 degrees MLR SN) either alone or in combination was unable to reverse the suppression. Purified CD8+ responder cells were inhibited as effectively by CsA as unfractionated cells, suggesting that CsA has a direct, lymphokine-independent effect on CD8+ cells. The frequency of CTL precursors (CTLp) responding in a primary MLR was measured by limiting dilution in the presence and absence of CsA. Using either unfractionated cells or cell-sorter-purified CD8+ responder cells, no suppression was seen at very low cell numbers but increased markedly as the cell number per culture increased. These results imply that either a CD8+ regulatory/suppressor cell is being diluted out at low cell numbers, or that low cell-density culture conditions might override the inhibitory effects of CsA.

Reasons for the apparent difference in the effects of produced and added ethanol on culture viability during rapid fermentation by Saccharomyces cerevisiae

By feeding ethanol at various high rates to low cell density cultures of Saccharomyces cerevisiae it was shown that the sharp fall in viability when ethanol is produced during rapid fermentations is in part a direct consequence of the high rate of change of extracellular ethanol concentration. Nevertheless, the fall in viability in high cell density rapid fermentations which produced 98 g L(-1) ethanol in 3 h considerably exceeded that of control low cell density cultures to which ethanol was added at the same rate. This difference was shown to be not due to intracellular ethanol accumulation or to differences in glucose concentration between the cultures. The concentrations of a range of potentially toxic fatty acids, higher alcohols, and esters were measured during rapid fermentations, but when added at these concentrations to control cultures in the presence of ethanol they had no significant toxic effect. However, when rapid fermentations were conducted in rich medium containing 80 g L(-1) yeast extract, the apparent difference in toxicity of produced and added ethanol virtually disappeared. Magnesium was shown to be the component of yeast extract primarily responsible for this effect. The high rate of fall of viability when ethanol is rapidly produced is suggested to be partly due to the inability of the cells to adapt quickly enough to the rising ethanol concentration and partly to an increased demand for magnesium at higher ethanol concentrations which cannot be met in Mg-unsupplemented high cell density fermentations.