A mediator role of ceramide in the regulation of neuroblastoma Neuro2a cell differentiation.

Current studies indicate that ceramide is involved in the regulation of important cell functions, namely cell growth, differentiation, and apoptosis. In the present study, the possible role of ceramide in the differentiation of neuroblastoma Neuro2a cells was investigated. The following results were obtained. (a) Ceramide content of Neuro2a cells, induced to differentiate by retinoic acid (RA) treatment rapidly increased after addition of RA, was maintained at high levels in RA-differentiated cells and returned to the starting levels with removal of RA and reversal of differentiation; under the same conditions, the sphingosine content remained unchanged. (b) After a short pulse with [H]sphingomyelin or [H]sphingosine or L-[H]serine, the metabolic formation of ceramide was markedly higher and more rapid in RA-differentiated than undifferentiated cells. (c) Inhibitors of ceramide biosynthesis (Fumonisin B1, β-chloroalanine and L-cycloserine) diminished the extent of the differentiating effect of RA and concomitantly Cer content decreased. (d) The activity of neutral sphingomyelinase increased after addition of RA, maintained high levels in RA-differentiated cells, and returned to the initial levels with removal of RA. (e) Experimental conditions that cause an elevation of ceramide content (treatment with sphingosine or ceramide or C-ceramide or bacterial sphingomyelinase) inhibited cell proliferation and stimulated neurite outgrowth; dihydro-analogues of sphingosine, ceramide, and C-ceramide had no effect on differentiation. (f) treatment with Fumonisin B1 completely inhibited sphingosine-induced differentiation. These data suggest a specific bioregulatory function of ceramide in the control of Neuro2a cell growth and differentiation and pose the general hypothesis of a mediator role of ceramide in the differentiation of cells of neural origin.

Current studies indicate that ceramide is involved in the regulation of important cell functions, namely cell growth, differentiation, and apoptosis. In the present study, the possible role of ceramide in the differentiation of neuroblastoma Neuro2a cells was investigated. The following results were obtained. (a) Ceramide content of Neuro2a cells, induced to differentiate by retinoic acid (RA) treatment rapidly increased after addition of RA, was maintained at high levels in RAdifferentiated cells and returned to the starting levels with removal of RA and reversal of differentiation; under the same conditions, the sphingosine content remained unchanged.

(b) After a short pulse with [ 3 H]sphingomyelin or [ 3 H]sphingosine or L-[ 3 H]serine, the metabolic formation of ceramide was markedly higher and more rapid in RA-differentiated than undifferentiated cells. (c) Inhibitors of ceramide biosynthesis (Fumonisin B1, ␤-chloroalanine and L-cycloserine) diminished the extent of the differentiating effect of RA and concomitantly Cer content decreased. (d)
The activity of neutral sphingomyelinase increased after addition of RA, maintained high levels in RA-differentiated cells, and returned to the initial levels with removal of RA. (e) Experimental conditions that cause an elevation of ceramide content (treatment with sphingosine or ceramide or C 2 -ceramide or bacterial sphingomyelinase) inhibited cell proliferation and stimulated neurite outgrowth; dihydro-analogues of sphingosine, ceramide, and C 2 -ceramide had no effect on differentiation. (f) treatment with Fumonisin B1 completely inhibited sphingosine-induced differentiation. These data suggest a specific bioregulatory function of ceramide in the control of Neuro2a cell growth and differentiation and pose the general hypothesis of a mediator role of ceramide in the differentiation of cells of neural origin.
Increasing evidence indicates important roles for molecules of sphingoid nature in the modulation of cell response to different extracellular signals. These molecules include sphingosine, ceramide, and some derivatives of them, N-methylated forms of sphingosine, sphingosine-1-phosphate, and ceramide-1-phosphate (2,3). Ceramide (N-acyl-erythro-sphingosine) has been shown to possess bioeffector properties and to act as a key molecule in a new signal transduction pathway, the sphingo-myelin pathway or cycle (4 -7). In fact, in several cell lines, especially of the immune system, the activation of certain growth factor receptors by vitamin D3 and cytokines (tumor necrosis factor ␣, interleukin-1␤, and ␥-interferon) induces sphingomyelin hydrolysis by activation of sphingomyelinase, resulting in the elevation of the intracellular levels of ceramide. This, in turn, acts as mediator of the elicited physiological effects, presumably by controlling the activity of specific protein kinases and protein phosphatases. In particular, ceramide has emerged as a candidate for regulatory roles in biological processes that are intimately connected to each other, including cell proliferation, oncogenesis, differentiation, and apoptosis (reviewed in Refs. 4 -10).
A process that is based on the regulation of proliferation/ differentiation and differentiation/apoptosis is neural development. Several cell systems (neurons, glial cells, neurotumoral cells) that undergo morphological and functional differentiation in culture are available to study this process in vitro. Some studies suggest that sphingolipids and sphingoid molecules may be involved in the regulation of neural development. In fact, exogenously added glycosphingolipids are capable to affect differentiation of neurons in primary culture and to induce differentiation of neuroblastoma cells in vitro (for a review, see Ref. 11). Moreover, in cultured hippocampal neurons, sphingolipid biosynthesis is necessary for axonal outgrowth (12), and inhibition of sphingolipid biosynthesis and degradation causes opposite effects on axonal branching (13). Furthermore, induced expression of G D3 and/or b-series gangliosides is followed by differentiation of Neuro2a cells (14). Finally, in T9 glioma cells, addition of a cell-permeable ceramide analog (C 2 -ceramide) causes growth inhibition and formation of processes, in analogy with nerve growth factor, which produced the same effects with a concomitant increase of the cellular level of ceramide (15).
On these premises, we decided to carry out a systematic investigation on the involvement of ceramide as a bioregulator in neural differentiation and the associated processes of proliferation and apoptosis. In the present study, we investigated the role played by ceramide in the differentiation of neuroblastoma Neuro2a cells. Initially, we determined the ceramide concentration and the metabolic routes leading to ceramide in Neuro2a cells induced to differentiate by treatment with retinoic acid, under strictly standardized conditions. Then, the ceramide levels of Neuro2a cells were increased by different treatments, and the effects on differentiation were observed. The data obtained strongly suggest that ceramide is involved in the regulation of Neuro2a cell differentiation.
Cell Cultures-The murine neuroblastoma cell line, clone NB2a (Neuro2a, CCL-131, American Cell Type Culture Collection), was used. Cells were cultured in Falcon dishes in DMEM supplemented with 10% FCS, 4 mM L-glutamine, 1 mM sodium pyruvate, 100 units/ml potassium penicillin G, and 100 g/ml streptomycin sulfate in 5% CO 2 , 95% airhumidified atmosphere. To induce neurite outgrowth, cells were plated at a 1.8 ϫ 10 4 /cm 2 cell density, and 48 h after plating the medium was replaced with 2% FCS-DMEM containing 20 M retinoic acid (22), the incubation being continued for different times up to 48 h. Control experiment showed that incubation up to 48 h with 2% FCS-DMEM, in the absence of retinoic acid, caused only a very modest outgrowth of processes.  (18). In some experiments, cells were submitted, after pulse, to a 2-4-h period of chase in 10% FCS-DMEM devoid of 3 H molecules. At the end of the pulse or pulse-chase period, cells were consecutively washed with 10% FCS-DMEM (twice) and phosphate-buffered saline (twice), harvested by scraping, and lyophilized (20). The influence of endocytosis and lysosomal activity on the metabolic processing was assessed by performing experiments at 4°C (condition that blocks endocytosis) or in the presence of 50 M chloroquine (drug which blocks the activity of lysosomal enzymes), as previously reported (19). Volatile radioactivity, 3 H 2 O, released in the culture medium, was determined by fractional distillation of the culture medium under carefully controlled conditions, collection of the distilled fractions, and measurement of the radioactivity by liquid scintillation counting (23).
Treatment of Cultured Cells with Exogenous Sphingosine, Natural Ceramide, C 2 -Ceramide, or Bacterial SM-ase-5 ϫ 10 4 -10 5 cells were plated in 35-mm Falcon dishes and grown in 10% FCS-DMEM for 24 h before the experiments. At the time of experiments, dishes were washed three times with 2% FCS-DMEM and incubated for different times (up to 24 h) in the same medium (1 ml/dish) containing D-erythro-Sph (0.1-10 M) or C 2 -Cer (1-10 M), or natural Cer (0.5-10 M) or bacterial SM-ase (100 milliunits/ml). C 2 -Cer (which is a cell-permeable analog of Cer) and Sph solutions were prepared by adding small volumes of stock solutions in absolute ethanol to 2% FCS-DMEM (the final concentration of ethanol never exceeded 0.1%). The solutions were allowed to stand at 37°C for 1 h before treatment of the cells. Natural Cer was dissolved in ethanol/dodecane, 98:2 (v/v) according to Ji et al. (25), and added to the medium to reach the final wanted concentration. Under these conditions, Cer is able to penetrate into cells. The final concentration of ethanol and dodecane never exceeded 0.5 and 0.01%, respectively. Neither ethanol nor dodecane at the used concentrations caused visible changes of cell morphology. Bacterial SM-ase was added to the culture medium just before use. In parallel experiments, 1 M DL-threosphinganine or dihydro-Cer or C 2 -dihydro-Cer, dissolved as D-erythro-Sph, Cer, or C 2 -Cer, was used.
Effect of Inhibitors of Sphingosine and Ceramide Synthesis on Neuro2a Differentiation-The involvement of serine-palmitoyl transferase in RA-induced differentiation in vivo was assessed by adding 2.5 mM ␤-chloroalanine or L-cycloserine, known inhibitors of this enzyme (26,27), 2 h prior and during RA treatment. The involvement of Cer synthase in the same process was ascertained running experiments in the presence of Fumonisin B1 (12,28), the known inhibitor of this enzyme.
Morphological Differentiation and Measurement Of Neurite-like Processes Outgrowth-The degree of morphological differentiation was assessed by phase-contrast microscopy. In particular, 200 -300 cells in 4 -5 random fields in each dish were counted, and cells bearing neuritelike processes longer than the major cell body diameter (after treatments not exceeding 8 h), or bearing a neurite with a length at least double that of cell diameter (after treatment longer than 8 h), were scored as differentiated. Data were expressed as neurite-bearing cells (short or long neurites) as percentage of total cells counted. Cell aggregates were not scored, and cells with more than one neurite were only counted once. Cell viability was assessed by the trypan blue exclusion method.
[ 3 H]Thymidine Incorporation-5 ϫ 10 4 -10 5 cells were grown in 35-mm Falcon dishes in 10% FCS-DMEM. 24 h after plating cells were cultured in 2% FCS-DMEM containing different sphingoids or bacterial SM-ase (as described above) for 24 h. The medium was removed and replaced with 1 ml of DMEM containing 0.5 Ci of [ 3 H]thymidine (29). After 2 h at 37°C, cells were harvested with phosphate-buffered saline and treated with 10% trichloroacetic acid. The insoluble residue, filtered on microfiber glass filters GF/C (Whatman International Ltd., Maidstone, UK), was submitted to radioactivity counting.
Other Methods-Radioactivity was determined by liquid scintillation counting, fluorography, and radiochromatoscanning (Digital Autoradiograph, Berthold, Germany) (19,20). The content of Cer and Sph was determined by the method of Preiss et al. (30) and Ohta et al. (31), respectively. Total protein was assayed (32) using bovine serum albumin as the standard. Mg 2ϩ -stimulated neutral sphingomyelinase (N-SM-ase) activity was assayed on the cell homogenate as previously reported (33). Ganglioside content was determined (34) as lipid bound sialic acid (using N-acetylneuraminic acid as the standard) on the aqueous phase (see above) desalted by Sephadex G-25 column chromatography. Neutral glycolipids and sphingomyelin were purified from the organic phase (after alkaline methanolysis) by Unisil silicic acid column chromatography (35). Gangliosides and neutral glycolipids, separated by HPTLC (see above), were quantified by densitometric scanning (Camag TLC densitometer) (23) after visualization with a p-dimethylaminobenzaldehyde reagent (36) or a diphenylamine reagent (37), respectively. SM was determined after perchloric acid digestion (38,39). Statistical significance of differences was determined by the Student t test.
Ceramide and Sphingosine Levels during RA-induced Differentiation of Neuro2a Cells-After RA treatment in low (2%) FCS medium, Neuro2a cells underwent differentiation, which started to be appreciable after 60 -120 min and resulted in a very elaborated network of processes after 24 -48 h, in agreement with reported findings (22). As shown in Fig. 1, the content of endogenous Cer (1.02 Ϯ 0.11 nmol/mg protein, in undifferentiated cells) increased rapidly with time in Neuro2a cells after RA addition until about 8 h (1.55 Ϯ 0.46 nmol/mg protein). This level remained almost constant in differentiated cells (1.69 Ϯ 0.33 and 1.74 Ϯ 0.36 nmol/mg protein at 24 and 48 h, respectively). The increase of Cer content could be acknowledged at the first investigated time after RA treatment (30 min), indicating that the cell response to RA, in terms of Cer production, was very prompt. The removal of RA from the medium caused reversal of differentiation, as expected (22), with a parallel decrease of the Cer content to the starting level. It is noteworthy that the Sph content (100 Ϯ 12 pmol/mg protein) appeared to be unaffected by RA treatment. Concomitant treatment of Neuro2a cells with RA and 25 M Fumonisin B1, the inhibitor of Cer synthase (12, 28), reduced but not blocked the process of differentiation (Fig. 7, lanes RA and RA ϩ Fumonisin B1) and caused a concomitant decrease of Cer content (1.34 Ϯ 0.12 nmol/mg protein). It is noteworthy that upon RA plus Fumonisin B1 treatment, the Cer content remained substantially higher than that of undifferentiated cells.
Metabolic The radioactivity present in the total lipid extract, as well as tritiated water released in the culture medium (volatile radioactivity), produced by complete Sph degradation, was a little higher in differentiated than undifferentiated cells and markedly increased with pulse time in both cell types. Conversely, [ 3 H]Sph diminished with pulse time and represented only a minor portion of total incorporated radioactivity (radioactivity in the total lipid extract plus volatile radioactivity) at all the investigated times (about 5% at 30 min and 2% at 120 min). 3 H 2 O was a minor product of [ 3 H]Sph metabolism, since after 30 and 120 min of pulse it accounted only for about 8% and 12%, respectively, of the total incorporated radioactivity in both undifferentiated and RA-differentiated cells. [ 3 H]Cer represented by far the major radiolabeled metabolite of exogenous [ 3 H]Sph, accounting for more than half of total incorporated radioactivity in both undifferentiated and RA-differentiated cells (Fig. 3). It was produced more rapidly and at a higher extent in differentiated cells, whereas its utilization for the biosynthesis of complex sphingolipids (Glc-Cer, SM, and gangliosides) was higher in undifferentiated cells.
In pulse-chase experiments with L-[ 3 H]serine (Table II), control and RA-differentiated Neuro2a cells incorporated similar amounts of radioactivity into total sphingolipids (mainly Cer, SM, gangliosides, and neutral glycosphingolipids). At all investigated times [ 3 H]Cer represented the major 3 H-sphingolipid (from 52 to 87% of total sphingolipids) and was produced in significantly higher amounts by differentiated than undifferentiated cells. On the basis of these results, the effect of inhibitors of serine-palmitoyl transferase (a key enzyme in sphingolipid biosynthesis) on RA-induced differentiation was investigated. As shown in Table III, treatment of Neuro2a cells with 2.5 mM ␤-chloroalanine or L-cycloserine resulted in a timedependent, substantial, although not complete, inhibition of   RA-induced differentiation.

FIG. 2. Incorporation of radioactivity into the total lipid extract, Sph, and water after feeding undifferentiated (control) and differentiated (RA-treated) Neuro2a cells with 40 nM
The feeding experiments with [Sph-3 H]SM provided similar results (Fig. 4). After 2 h of feeding with 4 M [Sph-3 H]SM, followed or not by 4 h chase, [ 3 H]Cer represented the major radiolabeled metabolite in both undifferentiated and RA-differentiated cells, but its metabolic formation was significantly higher in the differentiated ones (2.4-fold and 1.8-fold at 0 and 4 h chase, respectively) (Fig. 4). Conversely, other 3 H-metabolites produced during [Sph-3 H]SM metabolism (mainly Glc-Cer, gangliosides, Sph, and water) were markedly lower in differentiated than control cells (Fig. 4). Also in these experiments 3 H 2 O (which could be properly measured after 4 h chase) constituted a very minor metabolite (about 6.5 and 2.5% of total incorporated radioactivity in undifferentiated and RA-differentiated cells, respectively). When [Sph-3 H]SM was administered under conditions that block endocytosis or lysosomal degradation, [ 3 H]Cer formation was only partially reduced in both control and RA-differentiated cells (Table IV). This indicates that only a portion of [Sph-3 H]SM is internalized into cells and processed in the lysosomes, the remainder being produced at an extralysosomal level (possibly the plasma membrane). The amount of [ 3 H]Cer produced in the presence of chloroquine or at 4°C was much higher (2.7-and 1.9-fold, respectively) in RA differentiated than undifferentiated cells. On these premises, and since neuroblastoma cells are known to contain an Mg 2ϩstimulated N-SM-ase (41,42), we investigated the possible role of this enzyme in Cer formation during RA-induced Neuro2a cell differentiation. As shown in Fig. 5, the activity of Mg 2ϩstimulated N-SM-ase, increased during RA-induced differentiation, was maximal in the fully differentiated cells and decreased upon removal of RA, paralleling reversal of cell differentiation. Fumonisin B1, at the concentrations used with cells in culture, did not affect the in vitro assay of N-SM-ase.
Effect of Induced Increase of Ceramide Content on Neuro2a Cell Differentiation and Growth-Undifferentiated Neuro2a cells were treated with Sph, natural Cer, C 2 -Cer, or bacterial SM-ase, conditions that are known to increase the cellular Cer level, and the differentiation process was morphologically assessed. The first set of experiments relied on the metabolic studies described above, indicating that exogenous Sph is actively taken up by Neuro2a cells and rapidly and predominantly N-acylated to ceramide. As shown in Fig. 6a, the addition of exogenous Sph (0.1-10 M) resulted in a dose-dependent increase of the cellular level of Cer. Parallelly, neurite outgrowth was markedly stimulated in a dose-and time-depend-     (Fig. 6, b and c, and Fig. 8, B and C). When Neuro2a cells were treated with 25 M Fumonisin B1, the differentiating effect of exogenous Sph was not detectable (Fig.  7, lanes Sp and Sp ϩ Fumonisin B1). It is noteworthy (and as reported above) that treatment with Fumonisin B1 inhibited the differentiating effect of RA as well, but only partially. Treatment of Neuro2a with 1 M natural Cer, 1 M C 2 -Cer, or bacterial SM-ase (100 milliunits/ml) for 2-4 h also resulted in the stimulation of neurite outgrowth (up to 2.5-fold with bacterial SM-ase and natural Cer and 3.5-fold with C 2 -Cer) (Figs. 8, D-F, and 9). In contrast to the marked inhibitory effect exerted on Sph and RA-induced differentiation, Fumonisin B1 did not affect the differentiation promoted by bacterial SM-ase after 4 and 24 h (106 and 93% with respect to SM-ase treated cells, respectively). As shown in Fig. 10, the administration of exogenous Sph, or natural Cer or C 2 -Cer or bacterial SM-ase, at the conditions stimulating neuritogenesis, caused a marked diminution of [ 3 H]thymidine incorporation into DNA, indicating inhibition of cell proliferation.

Specificity of Ceramide-induced Differentiation of Neuro2a
Cells-To evaluate the specificity of ceramide-induced differentiation, we compared the effects of equimolar concentrations of different sphingoids on this process. As shown in Table V, under the experimental conditions used, where Sph, natural Cer, and C 2 -Cer were active, either a racemic mixture of threodihydro-Sph or C 2 -dihydro-Cer or dihydro-Cer did not exert any effects on Neuro2a cell differentiation. DISCUSSION The first piece of evidence provided by this study is that enhanced levels of Cer are characteristic of RA-differentiated Neuro2a cells. In fact, Cer (but not Sph) content increases during RA-induced differentiation, is maintained at high levels in differentiated cells, and returns to the basal values upon reversal of differentiation. At least two metabolic pathways, de novo Cer biosynthesis and SM degradation, seem to contribute to increasing the Cer content in differentiated cells. Both pathways appear to be more efficient in differentiated cells. Particularly, it is surprising the rapidity and efficiency by which exogenous Sph is acylated to Cer in RA-differentiated cells, with a concomitant lesser degree of Cer metabolic progression to more complex sphingolipids (gangliosides, SM), thus resulting in Cer accumulation. A similar situation has been reported to occur in GH 4 C 1 cells, where treatment with RA, at concentrations able to inhibit cell proliferation, causes a significant  and prolonged increase of cellular Cer content as a result of increased Sph N-acylation (43). Since Sph content is maintained constant during RA differentiation, an increased replenishment of Sph pool either by neosynthesis or sphingolipid degradation is requested in differentiated cells. The data here presented on L-[ 3 H]serine metabolism and on the effects of two inhibitors of serine palmitoyltransferase demonstrate that an increased neosynthesis of Cer does occur in RA-differentiated cells. This evidence suggests that the activity of serine palmi-toyltransferase, a rate-limiting enzyme in de novo Cer biosynthesis (44), is enhanced in RA-differentiated cells.
The results obtained with the [Sph-3 H]SM feeding experiments showed that also SM degradation contributes to enhance the Cer level in differentiated Neuro2a cells. Cer formation in differentiated cells remains markedly elevated also when endocytosis or lysosomal degradation are inhibited, especially in RA-differentiated cells. This indicates that an extralysosomal, possibly plasma membrane-bound SM-ase is mainly responsible for the increased SM degradation in RA-differentiated cells. Consistent with this interpretation is the evidence, here provided, that the activity of the Mg 2ϩ -dependent N-SM-ase, an enzyme especially concentrated in neural tissues (45) and in cells of neuronal origin (41,42), increases during RA-induced differentiation of Neuro2a cells. It is also worth noting that previous studies have shown that N-SM-ase increases in rat brain parallelly with neuronal maturation (46). A further support to the notion that Cer increase concomitant to RA-induced differentiation is due to stimulation of both de novo biosynthesis and SM degradation comes from the observation that treatment with Fumonisin B1, which inhibits ceramide synthase (one of the enzymes of the biosynthetic route) (12, 28) but does not affect the activity of N-SM-ase, markedly reduced, but not suppressed, both the formation of neurite-like processes and the increase of Cer level induced by RA treatment. The increase of Cer content owing to RA-induced differentiation was about 70% and was accompanied by an overall decrease of the sphingolipid content (particularly SM but also neutral glycosphingolipids) of differentiated as compared to undifferentiated cells, as well as a decrease of the metabolic involvement of Cer in the biosynthesis of complex sphingolipids. It will be interesting to ascertain whether a particular pool of Cer, separated from the one used for biosynthetic purposes, is involved in differentiation and thus submitted to enhancement. In this case, the increase of "active" Cer might be still higher.
The second, important, piece of evidence provided by this study is that conditions leading to enhance the Cer content of undifferentiated Neuro2a cells, in the absence of inducers like RA, succeeded in stimulating neuritogenesis, concomitantly with inhibition of cell proliferation. In fact, supplying of exogenous Sph, natural Cer, or C 2 -Cer or treatment with bacterial SM-ase is followed by induction of neurite formation and inhibition of thymidine incorporation into DNA. We observed that Fumonisin B1, which blocks Cer biosynthesis from Sph or dihydro-Sph (12,28), completely inhibited Neuro2a cell differentiation upon treatment with exogenous Sph but did not affect the differentiating effect promoted by bacterial SM-ase treatment. Notably, the stimulation of Neuro2a cell neuritogenesis following treatments that enhance the Cer content appeared to be rather specific since dihydroderivatives of Sph, Cer, and C 2 -Cer did not exert such effect. This body of observations strangely suggests that the increase of Cer originated from neosynthesis and/or SM degradation is instrumental to Neuro2a cell differentiation. A support to this view comes from the finding that nerve growth factor causes growth inhibition and formation of processes in T9 glioma cells, with concomitant increase of cellular Cer (15). Furthermore, in hippocampal neurons, Fumonisin B1 treatment inhibits axonal growth, together with Cer biosynthesis, and Cer derivatives are able to reverse this effect and to cause a significant increase in axonal length (12). Moreover, exogenous sphingosylphosphocholine, which promotes neuritogenesis in different neuroblastoma cells including Neuro2a, is rapidly processed by cells, Cer being the main metabolic product (47). In agreement with this hypothesis is also the finding that in leukemia cells SM hydrolysis, following activation of N-SM-ase, triggers cell differentiation (48). The reported evidence that exogenous Cer was unable to induce neurite outgrowth (49), and Sph-inhibited neuritogenesis (50) in neuroblastoma cells, seemingly contrasting our results, can be explained on the basis of the different culture and general experimental conditions used by those investigators. Moreover, no evidence was provided by these authors for any increase of Cer levels under the adopted experimental conditions.
It is worth stressing that in Neuro2a cells the increased level of Cer in RA-induced differentiation occurs very early, persists along cell differentiation, but returns to the basal values upon reversal of differentiation. Therefore, Cer more than a trigger of differentiation seems to be a necessary instrument for differentiation. Hence, its role would be that of a bioregulator, its constant presence being needed for the expression of a particular functional state of the cell. If so, the enzymes directly involved in Cer formation and utilization should be considered as suitable targets of influences governing the transition of neural cells from the stage of proliferation to that of differentiation.
In conclusion, this work provides solid evidence for a bioregulatory implication of ceramide in the differentiation of Neuro2a cells and poses the general question of a mediator role of ceramide in the control and maintenance of differentiation in cells of neural origin.