The independent and combined effects of RAR-, RXR-, and VDR-selective ligands on the growth of squamous cell carcinoma in vitro

Headnote

Abstract

Headnote

Objective: The purpose of our study was to analyze the independent and combined effects of RAR-, RXR-, and VDR-selective ligands on the growth of squamous cell carcinoma to develop new, more effective, and less toxic chemopreventive therapy.

Method: The effects of 13-cis retinoic acid (13-cis RA), LG1069 (a highly selective RXR ligand), and vitamin D^sub 3^ (D^sub 3^) were analyzed in vitro in the SCC-25 human squamous cell carcinoma cell line. SCC-25 cells were grown in culture medium containing vehicle or hormone at the indicated concentrations. Growth of surviving cells was then assessed using a hemocytometer. The presence of functional vitamin D3 receptor (VDR) was confirmed using gene-transfer experiments with a promoterlac Z reporter plasmid.

Headnote

Results: D^sub 3^ and 13-cis RA have equipotent antiproliferative effects on SCC-25 cells. Furthermore, equimolar LG1069 completely blocks the growth-suppressive effects of D3 but has no effect on the action of 13-cis RA.

Conclusion: D^sub 3^ and its analogs, administered alone or in combination with 13-cis RA, may provide more effective and less toxic chemopreventive therapy for the prevention of second primary carcinomas of the head and neck.

Sommaire

Objectif Nous avons analyse l'effet individuel et en combinaison de ligands selectifs des RAR, RXR et VDR sur la croissance des epitheliomas malpighiens dans le but de developper de nouvelles approches de chimioprevention plus efficaces et moins toxiques.

Methode: Nous avons analyse l'effet in vitro de l'acide retinoique 13-cis (AR13-cis), du LG1069 (un ligand tres selectif du RXR) et de la vitamine D3 (D3) sur la lignee cellulaire SCC-25 (epithelioma malpighien). Les cellules de SCC-25 sont mis en culture dans un milieu de croissance contenant les facteurs mentionnes en concentration variable. La croissance cellulaire est mesuree par hemocytometre. La presence de recepteurs fonctionnels pour la Vit D3 (VDR) est confirmee par la technique du transfert de gene par un plasmide promoteur-rapporteur (VDRE-lac Z).

Resultats: D3 et AR13-cis ont des effets antiproliferatifs equipotents sur les cellules SCC-25. De plus le LG1069 en concentration equimolaire bloque l'effet suppresseur de D3 mais n'a aucun effet sur l'action de AR13-cis.

Conclusions: D3 et ses analogues administres seuls ou en combinaison avec I'AR13-cis pourraient offrir une protection plus efficace et moins toxique dans la chimioprevention de seconds primaires malpighiens dans la sphere ORL.

Key words: chemopreventive therapy, RAR-selective ligands, RXR-selective ligands, VDR-selective ligands, squamous cell carcinoma, 13-cis retinoic acid, vitamin D3

Regardless of primary site, patients who present with early-stage carcinoma of the head and neck have the best overall prognosis. However, the major cancerrelated cause of death in this patient population and among long-term survivors of more advanced disease is the development of second primary carcinomas (SPCs). The incidence of SPCs ranges from 4 to 7% annually and does not seem to decrease with time. Furthermore, 70 to 80% of these synchronous or metachronous SPCs occur in the aerodigestive tract (head and neck, lung, and esophagus).1,2

The concept of `field cancerization' introduced by Slaughter et al.3 in 1953 has been used to explain the frequency with which synchronous and metachronous SPCs are seen in patients with head and neck carcinoma. According to this concept, the entire epithelial lining of the upper aerodigestive tract is subject to the same carcinogenic insults that resulted in the development of a primary invasive carcinoma. Therefore, at the time of presentation, several foci of malignant and premalignant epithelial changes may coexist in different stages of evolution. This theory along with the observation that carcinogenesis is a reversible process in its early stages have resulted in the rapidly expanding study of chemoprevention.

To date, retinoids are the most studied class of chemopreventive agents. Vitamin A and its metabolites, such as 13-cis retinoic acid (13-cis RA), have significant effects on the growth and differentiation of squamous cell carcinoma.45 Retinoids interact with two classes of nuclear receptors, the RARs and the RXRs (both of which have alpha, beta, and gamma subtypes), which act directly to modulate the expression of specific target genes. The natural ligand for the RXR is 9-cis RA, whereas many naturally occurring retinoids, including all-trans RA, 9-cis RA, and 13-cis RA, interact with RARs. Retinoids regulate gene expression by binding to their cognate receptors, which then interact with specific DNA sequences known as response elements. These response elements are located in the promoter region of target genes, most of which remain unknown. Functional interactions between RARs and RXRs are further complicated by the fact that heterodimerization of the two receptors is a prerequisite step in responseelement binding and transcriptional regulation of target genes.6 Similarly, the related nuclear receptors for thyroid hormone and vitamin D3 also function as heterodimers with RXRs7 (Fig. 1).

Based on the observed effects of retinoids, several prospective randomized trials have been carried out to assess the efficacy of 13-cis RA in the prevention of SPCs in patients with carcinoma of the upper aerodigestive tract.8-1 These trials have shown promise with statistically significant decreases in the incidence of SPCs in treated versus control arms. However, toxicity has been a limiting factor. Most patients who received 13-cis RA experienced mild-to-moderate symptoms of hypervitaminosis A.8

It has long been recognized that 1,25-dihydroxyvitamin D^sub 3^ (D^sub 3^) is capable of reversing the hyperplastic epithelial changes associated with psoriasis.11 Vitamin D3 has recently been shown to induce differentiation and suppress the growth of squamous cell carcinoma in vitro.12-14 Given that the vitamin D^sub 3^ receptor (VDR) must heterodimerize with RXRs, the potential exists for synergistic tumour suppression when vitamin D^sub 3^ and retinoids are co-administered. Furthermore, the mechanisms through which these two classes of hormone cause differentiation and growth suppression of squamous carcinoma may target different events in the process of carcinogenesis. It may be possible, with combined therapy or new chemopreventive strategies, to decrease treatment-associated morbidity and subsequently allow prolonged therapy with increased efficacy.

The objective of this study is to evaluate the independent and combined effects of 13-cis RA, LG1069, a synthetic highly selective ligand for the RXR,ls l6 and D^sub 3^ on the growth of squamous cell carcinoma in vitro.

Materials and Method

la,25-Dihydroxyvitamin D^sub 3^ was kindly supplied by Dr. M.R. Uskokovic (Hoffmann-La Roche). 13-Cis retinoic acid was purchased from ICN. LG1069 was a generous gift from Dr. Tim Willson (Glaxo-Welcome Research and Development). All hormones were dissolved in dimethyl sulfoxide, and stock solutions were stored in the dark at -20 deg C.

Cell Cultures

The SCC-25 squamous cell carcinoma cell line was obtained from the American Type Cultures Collection (ATCC-Rockville, MD). The cells were cultured in Dulbecco's modified eagle medium (DMEM) (Gibco BRL) supplemented with 10% fetal bovine serum (FBS) (Gibco BRL), 1% 100x penicillin-streptomycin (Gibco BRL), and 0.4 mu g/mL hydrocortisone (ICN), at 37 deg C in a humidified atmosphere of 5% CO^sub 2^.

Cell-Growth Assays

SCC-25 cells were seeded in six-well plates at 15,000 cells per well in 2 mL of culture medium. After 24 hours, the culture medium was changed to charcoalstripped medium containing vehicle (DMSO) or hormone at the indicated concentrations. The media was changed every 2 days. On the designated day (2 to 10), cells were washed with 2 mL of phosphate-buffered saline and removed from the plate by incubation with 0.5 mL of 0.05% trypsin-EDTA. The cells were then counted using a hemocytometer. All treatment conditions were tested in quadruplicate, and statistical significance was determined using the independent variable two-tailed Student's t test.

Assay for Functional Vitamin D^sub 3^-Receptor Activity in SCC-25 Cells

SCC-25 cells were grown to 60% confluency in six-well plates in charcoal-stripped medium. Cells were then washed with 2 mL of Opti-MEM I reduced serum media (Gibco BRL). One mL of Opti-MEM I was then added to each well. Transfections were carried out using Lipofectin (Gibco BRL) according to the supplier's protocol. The promoter-reporter plasmid used was a VDRE-lac Z construct in the p61OAZ plasmid (Fig. 2). A tk-luciferase reporter plasmid, donated by Dr. Mark Featherstone (McGill University Cancer Center), was used as an internal control for transfection efficiency. Twelve hours following transfection of SCC-25 cells, the culture medium was changed to charcoal-stripped medium containing hormone at the indicated concentrations. Twenty-four hours later, the cells were lysed using lysis buffer (Promega), and beta-galactosidase assays were carried out as described by Tora et al.17 Transfections were all performed in quadruplicate. Transfection efficiency was standardized using the Luciferase Assay System with reporter lysis buffer (Promega) according to the supplier's protocol.

Results

Characterization of the SCC-25 Cell Line

The SCC-25 cell line was chosen based on its elevated expression of RAR-beta and gamma.18 The expression of RAR-beta and gamma has been shown to decrease as premalignant epithelial lesions progress to carcinoma in situ and, eventually, invasive squamous cell carcinoma. Loss of RAR-beta expression has been speculated to be a prerequisite for the progression of many epithelial malignancies.18,19 Chemopreventive therapy targets pre-invasive dysplastic epithelial lesions that, in theory, should have preserved RAR-beta expression. The SCC-25 cell line has levels of RAR-beta expression comparable to those of dysplastic epithelium, making it an ideal model for the evaluation of the differentiating effects of retinoids and other `mucosal stabilizers,' such as D^sub 3^.

The SCC-25 cell line was grown as described in Materials and Methods. The mean doubling time of the cell line, in the absence of either 13-cis RA or D^sub 3^, was 42 2 (SD) hours. To demonstrate the presence of functional VDR in SCC-25 cells, gene transfer experiments were carried by transient transfection of a D3dependent promoter-lac Z reporter vector (see Materials and Methods for details). The lac Z gene encodes a bacterial beta-galactosidase enzyme. VDR function can be assessed by measuring D^sub 3^-dependent beta-galactosidase activity in extracts of transfected cells. There was a 40fold induction of beta-galactosidase activity when cells were exposed to 10 nM of D^sub 3^, as compared to the negative control (10 nM LG1069) (Fig. 3). This D^sub 3^-dependent upregulation of beta-galactosidase activity confirms the presence of functional VDRs in SCC-25 cells.

Effects of 13-Cis Retinoic Acid and D^sub 3^ on Growth of SCC-25

Consistent with previously published results, 13-cis RA caused a dose-dependent suppression of SCC-25 growth (Fig. 4). Following a 10-day exposure to 10 nM 13-cis RA, growth, as compared to control, was decreased to 17.9+/- 8.8%. There was no statistically significant difference in the cumulative suppression caused by the three concentrations of retinoid used.

Figure

Figure 2 Figure 3

Similarly, D^sub 3^ caused a dose-dependent suppression of SCC-25 growth (Fig. 5). However, the difference in growth suppression caused by 1 nM, 10 nM, and 100 nM of D^sub 3^ was statistically significant (p < .01). Subsequent to a 10-day exposure to 10 nM D^sub 3^, growth, as compared to control, was decreased to 18.1 +/- 4.5% There was no statistically significant difference between the independent effects of 13-cis RA and D^sub 3^.

The effect of 13-cis RA and D3 when co-administered at a concentration of 10 nM is shown in Figure 6. The cumulative suppression caused by the two drugs in combination following 10 days of treatment was 9.2 +/3.15% as compared to 17.9 +/-8.8% (p = .15) and 18.1 +/-4.5% (p = .03) when 13-cis RA and D^sub 3^ were administered independently. This suggests that there is no synergistic or additive growth suppression in response to the two drugs in combination.

Differential Modulation by LG1069 of Growth-Suppressive Effects of 13-Cis RA and D^sub 3^

Although ligand binding by RXRs is not necessary for the formation of RAR-RXR and VDR-RXR dimers, RXR-selective ligands, such as LG1069, may modulate gene transcription. Varying concentrations of LG1069 were added to the culture medium to assess the effects of RXR ligand binding on the growth-suppressive effects of 13-cis RA and D^sub 3^. When 10 nM of LG1069 was administered in combination with 10 nM of 13-cis RA, there was no statistically significant difference (p = .65) in cumulative growth suppression (Fig. 7). However, 10 nM of LG1069 completely blocked the growth-suppressive effects of an equimolar concentration of D^sub 3^ (p < .0001). Following 10 days of treatment with the two drugs in combination, growth, as compared to control, was 95.9 +/-3.6% (Fig. 8). Furthermore, LG1069 (0.01 nM to 10 nM) showed a dose-dependent effect on growth suppression by 10 nM of D^sub 3^. At day 10, growth in the presence of 10 nM of D^sub 3^ and 0.01 nM of LG1069 was 39.9 +/- 7.2%, as compared to 18.1 +/-4.5%, when 10 nM of D3 was administered alone (p = .01).

Discussion

The significant incidence, morbidity, and mortality associated with SPC of the head and neck clearly emphasizes the need for effective chemopreventive strategies. Given that most patients have received antecedent locoregional therapy, the treatment options available to this patient population may be limited, further complicating their care. Despite a reduction in the incidence of SPC associated with high-dose 13-cis RA therapy, toxicity and poor patient compliance have limited its clinical usefulness.8-lo,zo Furthermore, cessation of therapy has been associated with the recurrence of dysplastic squamous epithelium.21,22 Currently, clinical trials using low-dose 13-cis RA are underway.23 Whether these doses can significantly decrease the incidence of SPC while minimizing toxicity remains to be seen. However, toxicity data and the requirement for long-term maintenance therapy with 13-cis RA suggest the need for new chemopreventive strategies.

Vitamin D^sub 3^ has been shown to induce differentiation and suppress the growth of many malignancies including malignant melanoma, myeloid leukemia, and prostatic, colonic, breast, and squamous cell carcinomas.24 This study suggests that it is as effective as equimolar 13-cis RA in the suppression of squamous cell carcinoma in vitro. Furthermore, the absence of an additive or synergistic relationship between 13-cis RA and D^sub 3^ in SCC-25 cells is demonstrated.

RXR ligands are known to affect the growthinhibitory effects of 13-cis RA. Studies have demonstrated increased growth suppression when 13-cis and 9-cis RA are combined.25 The effect of RXR-selective ligands on D^sub 3^-dependent growth inhibition is less well characterized. This study demonstrates that LG1069 has little effect on the actions of 13-cis RA in SCC-25 cells. However, at equimolar concentrations, it completely blocks the effects of D^sub 3^. The nature of this antagonism remains to be determined. One possibility is that the presence of high RXR-ligand concentrations stimulates homodimerization of the RXR. This RXRRXR dimer formation subsequently reduces the pool of free RXR monomers available for heterodimerization with the VDR, resulting in decreased D^sub 3^-dependent transcription.7 A second possibility is that 9-cis RA and RXR-selective ligands up-regulate the expression of 24hydroxylase. The 24-hydroxylase enzyme is responsible for converting 1,25-dihydroxyvitamin D^sub 3^ to 1,24,25-trihydroxyvitamin D^sub 3^, its inactive form. When studied in a colon carcinoma model, co-administration of 9-cis RA and D^sub 3^ to HT-29 cells blocked the antiproliferative effects of D^sub 3^. However, under the same treatment conditions, D^sub 3^-dependent growth suppression in Caco-2 cells was unchanged. When compared to the Caco-2 cell line, 9-cis RA induced expression of the 24-hydroxylase gene only in HT-29 cells.26 These findings highlight the fact that malignancies in vivo are heterogeneous collections of cells that will display varying and complex responses to retinoids and D^sub 3^. Therefore, to suggest that SCC-25 cells are representative of squamous cell carcinoma in vivo is simplistic. The modulating effects of LG1069 on both RAR- and VDR-dependent growth suppression will be assessed in several other cell lines.

Figure 4

Figure 5

Figure 6

Figure 7 Figure 8

The mechanisms through which 13-cis RA and D^sub 3^ modulate cell growth and differentiation remain, for the most part, unknown. It has been shown that both hormones induce apoptosis and cell-cycle arrest.527,zs However, the exact genetic pathways through which these changes occur are unclear. It is very likely that D^sub 3^ and 13-cis RA target the process of carcinogenesis at different sites. Indirect evidence to support this hypothesis is found in the effects of either drug on activator protein-1 (AP-1) activity. Activator protein-1 is a transcription factor composed of elements of the Jun and Fos families of DNA-binding proteins, and interacts with the promoters of many genes that regulate cellular differentiation and proliferation. It is frequently hyperexpressed in malignant cells. Retinoids decrease the expression of AP-1.29 However, D^sub 3^ has been shown to suppress the growth of malignant cell lines without decreasing AP-1 activity.30 Therefore, a chemopreventive strategy using both D^sub 3^ and 13-cis RA in combination, targeting several pathways involved in the development of invasive squamous cell carcinoma, may be more efficacious than using either drug independently. Although not supported by the combined effects of D^sub 3^ and 13-cis RA on SCC-25 cells, a combined approach may allow the dose of either drug to be decreased and, therefore, limit toxicity. Furthermore, such a combined chemopreventive approach may prove much more effective in cases of premalignant and malignant epithelial lesions that demonstrate resistance to either hormone when administered independently.

Despite the significant antiproliferative effects of D^sub 3^, its use in vivo is limited by the serious side effects of hypercalcemia, hypercalciuria, and bone resorption.lz Considerable efforts have been made to develop potent analogs of D^sub 3^ devoid of these toxicities. Two such compounds, EB1089 and Calcipotriol (MC903), have been developed. EB1089 has shown great potential as a chemopreventive agent. In vivo, it has a halflife similar to that of D^sub 3^ but is at least 100-fold less potent than D^sub 3^ in calcium mobilization. Furthermore, EB1089 has been shown to be at least 10-fold more potent than either D^sub 3^ or Calcipotriol on growth inhibition of squamous cell carcinoma in vitro.12 Similar results have been demonstrated on breast carcinoma cell lines.31 EB1089 and other related compounds will significantly reduce the toxicity associated with D^sub 3^ treatment and increase the potential for new chemopreventive strategies using D^sub 3^ analogs.

This study represents the first to compare the effects of D^sub 3^ and 13-cis RA on the growth of squamous cell carcinoma in vitro. The results suggest that equimolar D^sub 3^ and 13-cis RA are equally as effective at suppressing the growth of squamous cell carcinoma. Furthermore, there does not seem to be an additive or synergistic relationship between the two. The indirect evidence that D^sub 3^ and 13cis RA exert their antiproliferative effects through different molecular pathways also suggests that combined therapy in vivo may be more effective than treatment using either one independently. The potential use of D^sub 3^ analogs alone or in combination with 13-cis RA represents a new form of potentially more effective and less toxic chemoprevention for SPC of the head and neck.

References

References

References

1. Vokes EE, Weichselbaum RR, Lippman SM, et al. Head and neck cancer. N Engl J Med 1993; 328:184-194.

2. Hong WK, Lippman SM, Wolf GT. Recent advances in head and neck cancer-larynx preservation and cancer chemoprevention. Cancer Res 1994; 53:5113-5120.

References

3. Slaughter DP, Southwick HW, Smejkal W. "Field cancerization" in oral stratified squamous epithelium: clinical implications of multicentric origin. Cancer 1953; 6:963-968.

4. Lippman SM, Benner SE, Hong WK. Cancer chemoprevention. J Clin Oncol 1994; 12:851-873.

5. Lippman SM, Kessler JF, Meyskins FL. Retinoids as preventive and therapeutic anticancer agents. Cancer Treat Rep 1987; 71:493-515.

6. Shalinsky DR, Bischoff ED, Lamph WW, et al. A noval retinoic acid receptor-selective retinoid, ALRT1550, has potent antitumor activity against human oral squamous cell carcinoma xenografts in nude mice. Cancer Res 1997; 57: 162-168.

References

7. Haussler MR, Haussler CA, Jurutka PW, et al. The vitamin D hormone and its nuclear receptor: molecular actions and disease states. J Endocrinol 1997;154:557-573.

8. Hong WK, Lippman SM, Itri LM, et al. Prevention of second primary tumors with isotretinoin in squamous cell carcinoma of the head and neck. N Engl J Med 1990; 323:795-801.

9. Benner SE, Pajak TF, Lippman SM, et al. Prevention of second primary tumors with isotretinoin in squamous cell carcinoma of the head and neck: long-term follow-up. J Natl Cancer Inst 1994: 86:140-141.

References

10. Pastorino U, Infante M, Maioli M, et al. Adjuvant treatment of stage I lung cancer with high dose vitamin A. J Clin Oncol 1993; 11:1216-1222.

11. Holick MF. Will 1,25 dihydroxyvitamin D3, MC903, and their analogs herald a new pharmacologic era for the treatment of psoriasis. Arch Dermatol 1989; 125:1692-1697. 12. Yu J, Papavasiliou V, Rhim J, et al. Vitamin D analogs: new therapeutic agents for the treatment of squamous cancer and its associated hypercalcemia. Anticancer Drugs 1995; 6: 101-108.

References

13. McElwain MC, Modzelewski RA, Yu WD, et al. Vitamin D: an antiproliferative agent with potential for therapy of squamous cell carcinoma. Am J Otol 1997; 18:293-298. 14. Kornfehl J, Formanek M, Temmel A, et al. Antiproliferative effects of the biologically active metabolite of vitamin D3 (1,25 [OH]2 D3) on head and neck squamous cell carcinoma cell lines. Eur Arch Otorhinolaryngol 1996; 253:341-344. 15. Zavacki AM, Lehmann JM, Seol W, et al. Activation of the orphan receptor RIP14 by retinoids. Proc Natl Acad Sci U S A 1997; 94:7909-7914.

References

16. Miller VA, Benedetti FM, Verret AL, et al. Initial clinical trial of a selective retinoid X receptor ligand, LGD1069. J Clin Oncol 1997; 15:790-795.

17. Tora L, White JH, Bron C, et al. Human estrogen receptor has two independent non-acidic transactivation functions. Cell 1989; 59:477-487.

18. Hu L, Crowe DL, Rheinwald JG, et al. Abnormal expression of the retinoic acid receptors and keratin 19 by human oral and epidermal squamous cell carcinoma cell lines. Cancer Res 1991; 51:3972-3981.

19. Xu XC, Ro JY, Lee JS, et al. Differential expression of nuclear retinoid receptors in normal, premalignant, and malignant head and neck tissues. Cancer Res 1994; 54: 3580-3587.

References

20. Shah JP, Strong E, DeCosse JJ. Effect of retinoids on oral leukoplakia. Am J Surg 1983; 146:466-470. 21. Hong WK, Endicott J, Itri LM, et al. 13-Cis-retinoic acid in the treatment of oral leukoplakia. N Engl J Med 1986; 315: 1501-1505.

22. Giannini F, Maestro R, Vukosavijevic T, et al. All-trans, 13cis, and 9-cis retinoic acids induce a fully reversible growth inhibition in HNSCC cell lines: implications for in vivo retinoic acid use. Int J Cancer 1997; 70:194-200. 23. Radiation Therapy Oncology Group. RTOG 91-15: Chemoprevention trial to prevent second primary tumors with lowdose 13-cis retinoic acid in head and neck cancer. 1993.

References

24. Walters MR. Newly identified actions in the vitamin D endocrine system. Endocr Rev 1992; 13:719-764. 25. Apfel CM, Kamber M, Klaus M, et al. Enhancement of HL-60 differentiation by a new class of retinoids with selective activity on retinoid X receptor. J Biol Chem 1995; 270:30765-30772. 26. Kane KF, Langman M, Williams GR. Antiproliferative responses of two human colon cancer cell lines to vitamin D3 are differentially modified by 9-cis retinoic acid. Cancer Res 1996; 56:623-632.

References

27. Narvez CJ, Welsh J. Differential effects of 1,25-dihydroxyvitamin D3 and tetradecanoylphorbol acetate on cell cycle and apoptosis of MCF-7 cells and a vitamin D3-resistant variant. Endocrinology 1997; 138:4690-4698.

28. Blutt SE, Allegretto EA, Pike JW, et al. 1,25-Dihydroxyvitamin D3 and 9-cis retinoic acid act synergistically to inhibit the growth of LNCaP prostate cells and cause accumulation of cells in G1. Endocrinology 1997; 138:1491-1497. 29. Huang C, Ma WY, Dawson MI, et al. Blocking activator protein-1 activity, but not activating retinoic response element, is required for the antitumor promotion effect of retinoic acid. Proc Natl Acad Sci U S A 1997; 94:5826-5830. 30. Nayeri S, Danielsson C, Kahlen JP, et al. The anti-proliferative effect of vitamin D3 analogues is not mediated by inhibition of the AP-1 pathway, but may be related to promoter selectivity. Oncogene 1995; 11:1853-1858. 31. Colston KW, Mackay AG, Chander S, et al. Novel vitamin D analogues suppress tumor growth in vitro. In: Norman AW, Bovillon R, Thomasset M, eds. Vitamin D: gene regulation, structure-function analysis and clinical application. Proc 8th Workshop on Vitamin D. Berlin: de Gruyter, 1991:465-466.