Multidrug resistance, inherent or acquired, is a frequent characteristic of cancer cells and is difficult to predict and to manage. Multidrug resistance is caused by multiple mechanisms, including the dysfunctional metabolism of the lipid second messenger ceramide. The cytotoxic effect of various chemotherapeutics is decreased when the generation of ceramides is impaired, which results in the ineffectiveness of routine dosage and the need for higher, even more toxic drug levels. Needless to emphasize, this is a most undesirable situation; patients and oncologists would welcome its possible correction. Here we review ceramide metabolism with relationship to both blocking and potentiating the toxic response of cancer cells to chemotherapy. It is hoped that administering agents to target ceramide metabolism in combination with chemotherapy will improve response rates, especially in those patients with metastatic disease.

The link between ceramide metabolism and chemotherapy response was alluded to by earlier studies with anthracyclines. Shortly thereafter, the association between anthracycline resistance and accelerated ceramide metabolism through glucosylceramide synthase (GCS) was revealed. The signaling pathways activated by daunorubicin, including a SMase-initiated ceramide pathway, have been the subject of a recent review.¹

The enzyme ceramide synthase, by way of the de novo pathway, was shown to mediate daunorubicin-induced apoptosis in leukemia cells.² Exposing cells to drug-heightened ceramide synthase activity, promoting ceramide formation and increasing thepercentage of apoptotic cells. Introducing Fumonisin B₁ to inhibit ceramide synthase negated the cytotoxic impact ofdaunorubicin, a result which clearly affirmed ceramide's role in anthracycline action. In a round of similar studies using human leukemia cell models, Jaffrezou et al³ found that daunorubicin likewise increased ceramide levels, albeit via SM hydrolysis and not ceramide synthase; nevertheless, apoptosis was the end result. Doxorubicin also promotes ceramide elevation, this work being conducted in MCF-7 breast cancer cells.^4,5Moving into the realm of multidrug resistance (MDR) brought recognition to the enzyme catalyzing ceramide glycosylation, GCS, as an important facet of cellular response to chemotherapy.

Studies comparing chemotherapy-sensitive and chemotherapy-resistant cancer cells demonstrated a clear distinction in cerebroside content shown by an accumulation of a glycosylated form of ceramide, glucosylceramide, in the drug-resistant cells.⁶ This finding holds true for drug-resistant cancer cells derived from breast, ovary, melanoma,⁷ and prostate (author's unpublished observations), and recent studies have confirmed this in a multidrug-resistant cell line derived from HT29 human colon carcinoma.⁸

Investigations into the mechanism underlying the glucosylceramide surplus in chemotherapy-resistant cells revealed that exposure of wild-type cells to doxorubicin prompted an increase in ceramide and cell death attributable to apoptosis, whereas in chemotherapy-resistant cells treated in the same fashion, the ceramide generated was converted to glucosylceramide and the cytotoxic response was nil.⁴ This indicated that drug-resistant cells had an enhanced capacity for ceramide metabolism through the glycosylation pathway. An evaluation of ceramide toxicity showed wild-type cells susceptible to cell-permeable C₆-ceramide, whereas drug-resistant cells were tolerant.⁹ Analysis of the metabolic fate of ceramide in these cells showed wild-type MCF-7 contained only free C₆-ceramide whereas doxorubicin-resistant cells contained only high amounts of glucosyl-C₆-ceramide. With this work, the importance of GCS in regulating cellular response to ceramide and perhaps to ceramide-generating drugs such as the anthracyclines, became apparent.

To more clearly establish the influence of ceramide metabolism on drug resistance, the GCS gene was introduced into MCF-7 cells using a retroviral tetracycline-on expression system.¹⁰ The cell line that was developed, “MCF-7/GCS,” expressed an 11-fold higher level of GCS activity in the expression-on mode, compared to the parent cell line, and demonstrated strong resistance to doxorubicin and to ceramide analogs.¹¹ Ceramide resistance displayed by MCF-7/GCS cells paralleled the activity of the expressed GCS. In addition, resistance to TNF-α, which employs ceramide as a second messenger in the cell-killing response, was also a property of MCF-7/GCS cells.¹² It was shown that TNF-α had little influence on the induction of apoptosis or on growth arrest in MCF-7/GCS cells, and lipid metabolism studies revealed that TNF-α promoted a ceramide increase in MCF-7 cells and a glucosylceramide increase in MCF-7/GCS cells. Further, TNF-α-induced caspase activity was halted as a result of GCS transfection. Doxorubicin and TNF-α resistance in MCF-7/GCS cells was related to hyperglycosylation of ceramide and not to shifts in the levels of P-gp, Bcl-2, or TNF receptor 1 expression.^11,12

The role of ceramide metabolism in regulating cellular the response to chemotherapy has also been demonstrated employing a reverse tactic, by introducing GCS-antisense into doxorubicin-resistant cells. These studies showed that decreasing cellular glycosylation potential through GCS antisense transfection heightened chemotherapy sensitivity in doxorubicin-resistant cells, effectively reversing resistance.¹³ The antisense cell line displayed by RT-PCR, Western blot, and in vitro assays, decreased GCS mRNA, GCS protein, and GCS enzymatic activity. These cells were 28 times more sensitive to doxorubicin compared to the parent MCF-7-AdrR cell line. Under doxorubicin stress, GCS-antisense transfected cells displayed time- and dose-dependent increases in endogenous ceramide and dramatically higher levels of caspase activity, compared to control cells. GCS antisense transfection also heightened sensitivity to Vinca alkaloids and taxanes but had little impact on sensitivity to either 5-FU or cisplatin.¹⁴ Inasmuch as the data from gene transfection studies positions GCS as a force to overcome multidrug resistance in chemotherapy (Fig. 1), manipulation of ceramide metabolism by drug intervention can also provide a viable therapeutic avenue. Ceramide levels are readily enhanced by the introduction of ceramide generating agents alone or in conjunction with inhibitors of ceramide metabolism ( Fig. 2).

Figure 1

Shifting cellular chemotherapy tolerance through GCS gene transfection.

Figure 2

Manipulation of Cellular Ceramide Levels Through Drug Intervention. The drugs shown exert a stimulatory influence (+) on SPT (serine palmitoyltransferase) and Cer Syn (ceramide synthase), or have an inhibitory impact (−) on GCS (glucosylceramide (more...)

A number of classical P-glycoprotein substrates influence ceramide metabolism by hindering its conversion to glucosylceramide. Tamoxifen, used in the Dartmouth regimen for treatment of melanoma,¹⁵ evaluated for treatment of pancreatic carcinoma and malignant gliomas,^16,17 and employed in a biochemotherapy regimen for patients with metastatic melanoma,¹⁸ inhibits glucosylceramide synthesis in vinblastine-resistant carcinoma.¹⁹ In doxorubicin-resistant cells, clinically relevant concentrations of tamoxifen, verapamil, and cyclosporin A, markedly decrease glucosylceramide levels with IC₅₀ values of 1.0, 0.8, and 2.3 mM, respectively.²⁰ Toremifene, an anti-estrogen with chemistry and pharmacology similar to tamoxifen, is equally effective.⁹ These studies suggest that the P-glycoprotein-targeted agents act not only by blocking cellular drug efflux but also through enhancing ceramide levels. RU486 (mifepristone) inhibits growth and induces apoptosis in MCF-7 cells,²¹ and retards ceramide glycosylation in MCF-7-doxorubicin-resistant cells.⁹ Ketoconazole overcomes resistance to doxorubicin and vinblastine in cancer cells,²² displays activity in advanced prostate cancer patients,²³ and it inhibits glucosylceramide synthesis (author's unpublished observations). Studies with tamoxifen analogs corroborate the idea that MDR reversal and inhibition of cellular glucosylceramide synthesis are allied. For example, exposure of doxorubicin-resistant cells to triphenylethylene, the tamoxifen nucleus minus the dimethylethanolamine moiety, neither inhibits ceramide glycosylation nor sensitizes cells to doxorubicin, but inclusion of tamoxifen with doxorubicin decreases glucosylceramide production, enhances ceramide generation, and sensitizes cells.^20,24 Similarly, cis-tamoxifen is devoid of chemosensitizing activity, indicating a stereochemical requirement.

Studies using chemical inhibitors of GCS further highlight a relationship between ceramide metabolism and chemotherapy efficacy. PPMP (1-phenyl-2-palmitoylamino-3-morpholino-1-propanol), a commercially available inhibitor of ceramide glycosylation,²⁵ sensitizes breast cancer cells to doxorubicin,²⁰ and neuroblastoma to Taxol and vincristine.²⁶ These enzyme inhibitors are structural analogs of the natural GCS substrate. A chemical cousin, PPPP (1-phenyl-2 palmitoylamino-3-pyrrolidino-1-propanol) completely blocks glucosylceramide synthesis in vincristine-resistant leukemia and in combination, enhances vincristine-induced cytotoxicity,²⁷ and pretreatment of melanoma cells with PPPP markedly reduces tumor formation and metastatic potential in mice.²⁸ In metastatic human colon cancer, raising the levels of ceramide by direct administration of ceramide analogs and ceramidase inhibitors induces apoptotic cell death, preventing tumor growth.²⁹ Preferential killing of drug-resistant epidermoid carcinoma cells over drug-sensitive counterparts upon exposure to PDMP and PPPP has also been shown.³⁰ A reduction in glucosylceramide levels accompanied PDMP-induced apoptosis, prompting the authors to suggest that manipulation of glucosylceramide levels is an avenue for preferential destruction of drug-resistant cancer cells. Spinedi et al³¹ showed that PDMP suppresses glucosylceramide synthesis and potentiates the apoptotic effect of C₆-ceramide in neuroepithelioma, suggesting that glucosylceramide synthesis is a mechanism to escape ceramide-governed apoptosis. Similarly, by targeting ceramide metabolism, Sietsma et al²⁶ showed that PDMP increases neuroblastoma sensitivity to chemotherapy, and that the increase involves a sustained elevation of ceramide.

Work with SDZ PSC 833 ([3'-keto-bmt1]-[val-2]-cyclosporine), signaled an important step forward in demonstrating both the cytotoxic principles of ceramide and the impact of manipulating the ceramide pathway for therapeutic purposes. SDZ PSC 833 is a second generation MDR modulator that interferes with P-glycoprotein-mediated drug efflux.³² Our group first reported that SDZ PSC 833 alone strongly activates cellular ceramide formation, and that the increase in ceramide was mirrored by a progressive decline in cell survival.³³ Cells having an enhanced capacity for ceramide glycosylation were more resistant to SDZ PSC 833, as opposed to drug-sensitive cells.^6,24 Lipid metabolism studies revealed that SDZ PSC 833 resistance was allied with rapid conversion of ceramide to glucosylceramide.²⁴ Studies with P-glycoprotein negative cells and with mdr-1-transfected cells showed that SDZ PSC 833 elicits ceramide generation independently of P-glycoprotein.³⁴ In other studies it was revealed that drug sensitivity in leukemia could be enhanced by SDZ PSC 833 through perturbations in SM-ceramide pathways and not by modified drug efflux parameters.³⁵

Tilly and colleagues, using an inhibitor of ceramide-induced cell death, showed that cell destruction by chemotherapy was preventable if ceramide metabolism was manipulated.³⁶ Numerous studies now clearly demonstrate that co-administration of agents that target ceramide metabolism enhances the cytotoxic impact of chemotherapy. For example, vinblastine toxicity is intensified by SDZ PSC 833 in P-glycoprotein-poor cells, and both drugs at low dose promote ceramide formation.³⁷ Doxorubicin and tamoxifen synergize to increase ceramide formation, complementing cytotoxicity.^5,20 Another example of enhanced drug efficacy through ceramide targeting is shown by the influence of RU486 on doxorubicin.^5,9 A three-component regimen consisting of doxorubicin and two agents that enhance ceramide formation, tamoxifen and SDZ PSC 833, can bring cell viability to zero.²⁴ Suramin, a polysulfonated naphthylurea, disrupts glycolipid metabolism and elicits apoptosis through ceramide pathways in a number of cancer cell lines.³⁸ In Bcl2-transfected prostate cancer cells that are resistant to doxorubicin, a doxorubicin/suramin combination was shown effective in circumventing resistance.³⁹ Suramin is currently in clinical trials for treatment of breast and prostate cancer.^40–42 Aberrant ceramide generation is also involved in prostate cancer cell resistance to chemotherapy,⁴³ and it was recently shown that ceramidase is overexpressed in prostate cancer.⁴⁴ These findings justify exploring a ceramide-targeted approach for the treatment of prostate cancer.

More evidence that the ceramide pathway can be manipulated for therapeutic purposes stems from work with the novel synthetic retinoid, N-(4-hydroxyphenyl) retinamide (4-HPR). 4-HPR has been shown to significantly increase the levels of ceramide and induce mixed apoptosis/necrosis in highly drug-resistant human neuroblastoma cell lines.⁴⁵ The cytotoxicity of 4-HPR can be enhanced by adding modulators of ceramide metabolism,⁴⁶ such as L-threo-dihydrosphingosine (safingol), a sphingosine analog and protein kinase-C inhibitor,⁴⁷ and PPMP and tamoxifen. Safingol has been shown to be synergistic with 4-HPR and produces a 100- to 10,000-fold increase in cytotoxicity, compared to 4-HPR alone, in neuroblastoma, lung, melanoma, prostate, colon, and pancreatic cancer cell lines.⁴⁶The safingol/4-HPR combination is minimally toxic to fibroblasts and bone marrow myeloid progenitor cells.

The involvement of ceramide metabolism in cancer cell response to chemotherapy is far-reaching. Exogenous addition of ceramide enhances Taxol-induced apoptosis in leukemic T cells.⁴⁸ Ceramide also enhances Taxol-induced apoptosis in head and neck squamous cell carcinoma, suggesting that a Taxol/ceramide combination therapy may be a promising alternative to conventional treatment of head and neck cancers.⁴⁹ Research in breast cancer shows that Taxol elicits de novo ceramide generation and apoptosis; however, addition of ceramide synthesis inhibitors, such as L-cycloserine, checks Taxol-induced apoptosis.⁵⁰ These data imply that ceramide is obligate for Taxol-elicited cytotoxicity. The theme on the interplay of ceramide with chemotherapy response continues to expand. Recent reviews by Norman Radin caption this tract.^51,52

Many studies now indicate that, in addition to SMase, de novo enzymes of ceramide synthesis are also viable targets for chemotherapeutic agents. Etoposide regulates serine palmitoyltransferase, the rate-limiting enzyme in de novo ceramide synthesis.⁵³ In human neuroblastoma, 4-HPR elevates ceramide via coordinate activation of serine palmitoyltransferase and ceramide synthase.⁵⁴

Controlling ceramide metabolism is an attractive approach to cancer treatment. The majority of drugs that would be used are already available in the clinic, expediting the application of principle to practice. Ceramide metabolism can also be impacted by gene therapy. This has been illustrated through the application of antisense GCS, which is effective in the complete reversal of doxorubicin resistance in a breast cancer cell model.¹⁴