Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Oct 14.
Published in final edited form as: Org Lett. 2006 Feb 16;8(4):649–652. doi: 10.1021/ol052839v

Synthesis of 1-deoxysphingosine derivatives with conformationally restricted pyrrolidinediol head groups

Ann M Dougherty †, Frank E McDonald †,✉, Dennis C Liotta †,✉, Steven J Moody ‡, David C Pallas ‡, Carrie D Pack §, Alfred H Merrill §
PMCID: PMC4604444  NIHMSID: NIHMS63014  PMID: 16468733

Abstract

graphic file with name nihms63014u1.jpg

A family of cyclic 1-deoxysphingolipid derivatives of structure 4 has been designed and synthesized, which may serve as tumorigenesis suppressors for various cancers. Compound 4 is a second-generation analog developed from sphingosine (1), in which a hydroxyl substituent is moved from C1 to C5, and a methylene is added for conformational rigidity between the C2-nitrogen substituent and C4. The synthetic chemistry for pyrrolidine ring closure at C3-C4 features ring-closing metathesis followed by hydroboration-oxidation.

Sphinoglipids are natural products found in most cell membranes, and are structurally characterized by a long carbon chain “sphingoid” base that is derivatized with amide-linked fatty acids and various polar headgroups.1 Sphingoid bases have additional functions as cellular mediators and protein kinase C (PKC) inhibitors,2 affecting the growth, differentiation, migration, and apoptosis of cells. Extensive research efforts have resulted in several synthetic approaches to sphingoid bases3 and structural analogs, encouraged in part by recent discoveries regarding the anticancer activity of sphingolipids. In 2003 Menaldino et al. reported that 1-deoxysphingoid bases of general structure (3, Figure 1) were growth inhibitory and cytotoxic at concentrations up to ten-fold lower than for sphingosine (1), and up to fifty-fold more potent than the corresponding N-acylated ceramide 2 in HT29 and DU145 cell lines.4 As the primary alcohol of sphingosine is phosphorylated in vivo (resulting in undesired mitogenic/anti-apoptotic activity), the design of 1-deoxyanalogs 3 prevents phosphorylation, and by moving the hydroxyl group to the 5-position, lipophilicity of 3 is similar to sphingosine (1). In order to minimize N-acylation activity, we have further modified the 1-deoxysphingoid lead structure 3 to a cyclic pyrrolidine-diol 4, which also provides conformational restriction of the polar groups. This communication describes the synthesis and biological evaluation of several stereoisomers of 4, prepared in highly convergent fashion.

Figure 1. Structures of Sphingosine (1), Ceramide (2), and 1-Deoxyanalogs 3 and 4.

Figure 1

Open in a new tab

We envisioned that the cyclic pyrrolidine-diol analog 4 could be prepared from the functionalized dihydropyrrole 5, which in turn would arise from ring-closing metathesis5 of the diallylamine 6 (Figure 2).

Figure 2. Retrosynthesis for cyclic 1-deoxysphinganine.

Figure 2

Open in a new tab

We initially planned to prepare 2-amino-3-butene from the amino acid L-alanine, but racemization occurred under all conditions attempted via alpha-aminoaldehyde intermediates.6 Therefore, enantioselective synthesis of 2-amino-3-butene was accomplished by borohydride reduction of the chiral sulfinimide7 derived from methyl vinyl ketone (7), providing sulfinamine 8 as the major product of a 7: 1 mixture of diastereomers (Scheme 1). The minor diastereomer was separated from 8 by careful silica gel chromatography.8 Acidic cleavage9 of the chiral auxiliary and Cbz-protection of nitrogen provided compound 9.

Scheme 1. Enantioselective synthesis of N-Cbz-2-amino-3-butene (9).

Scheme 1

Open in a new tab

Several approaches were explored for preparation of the fragment bearing carbons 4 and 5. The best route involved asymmetric epoxidation of 11 to 12 (Scheme 2), followed by LDA elimination10 to give the allylic diol 13. Differentiation of the primary alcohol as the bromide and protection of secondary alcohol as the silyl ether afforded chiral non-racemic synthon 14 for carbon-nitrogen coupling. The enantiomer of 14 was likewise prepared beginning with epoxidation of 11 with d-DIPT.

Scheme 2. Stereoselective synthesis of allylic bromide (14).

Scheme 2

Open in a new tab

Reaction of carbamate 9 with sodium hydride11 and N-alkylation of each enantiomer of allylic bromide 14 provided the dienes 15 and 16 (Scheme 3), which each underwent ring-closing metathesis to dihydropyrroles 17 and 18 in excellent yield, using the Hoveyda metathesis catalyst.12

Scheme 3. Coupling of 9 and 14, and ring-closing metathesis to diastereomeric dihydropyrrolidines 17 - 18.

Scheme 3

Open in a new tab

A variety of conditions were explored for the introduction of the C3 alcohol via anti-Markovnikov hydration, with the best results obtained with hydroboration13 of the alcohols 19 and 20 with either borane-dimethyl sulfide or thexylborane followed by alkaline hydrogen peroxide oxidation and hydrogenolytic removal of the N-Cbz protective group.14 As expected, diastereomer 19 produced a single pyrrolidine-diol diastereomer 21, arising from stereoinduction from the C5-hydroxyl (Still-Barrish model)15 reinforcing the steric effect from the methyl substituent attached at C2. In contrast, diastereomer 20 gave a separable mixture of diastereomers 22 and 23, consistent with opposing effects of stereoinduction from C2 and C5 chiral allylic postions.16

After hydrogenolysis of the N-Cbz protective group, pyrrolidine-diol stereoisomers 24 – 26 and their enantiomers (prepared from ent-8 following the same synthetic route) were evaluated for cytotoxicity in DU-145 (human prostate carcinoma) and HT-29 (human colon carcinoma) cell lines (Table 1).17,18 We were pleased to observe that pyrrolidine-diol compounds 25 and ent-25 exhibited cytotoxicity against DU-145 cells in the same range as acyclic (2S,3S,5S)-3, and the compounds 25, ent-25, 26, and ent-26 were uniformly more cytotoxic against HT-29 cells in comparison with D-erythro-sphingosine (1).19 The range of biological activities among the various stereoisomers of cyclic structures 24 - 26 is relatively small, and thus further structure-activity studies on other cyclic aminodiols 24 - 26 as well as first-generation aminodiols 3 are warranted.

Table 1. Biological evaluation of pyrrolidine-diol analogs 24 - 26 and ent-24 - 26.

graphic file with name nihms63014f7.jpg
Open in a new tab

Supplementary Material

1si20060111_11

Scheme 4. Stereoselectivity of hydroboration of 19 vs. 20, and completion of target compound syntheses.

Scheme 4

Open in a new tab

Acknowledgments

This research was supported by the National Institutes of Health, through the National Cooperative Drug Discovery Grant program U19 CA 87525, and R01 CA 57327 to D. C. P.

Footnotes

Supporting Information Available. Experimental procedures and characterization data for compounds 8 - 26. This material is available free of charge via the Internet at http://pubs.acs.org.

Contributor Information

Frank E. McDonald, Email: fmcdona@emory.edu.

Dennis C. Liotta, Email: dliotta@emory.edu.

References

  • 1.Voet D, Voet JG. Biochemistry. 2nd. Wiley; New York: 1995. [Google Scholar]
  • 2.(a) Hannun YA, Loomis CR, Merrill AH, Jr, Bell RM. J Biol Chem. 1986;261:12604. [PubMed] [Google Scholar]; (b) Merrill AH, Jr, Sereni AM, Stevens VL, Hannun YA, Bell RM, Kinkade JM., Jr J Biol Chem. 1986;261:12610. [PubMed] [Google Scholar]
  • 3.(a) Garner P, Park JM, Malecki E. J Org Chem. 1988;53:11061. [Google Scholar]; (b) Herold P. Helv Chim Acta. 1988;71:354. [Google Scholar]; (c) Nimkar S, Menaldino D, Merrill AH, Liotta D. Tetrahedron Lett. 1997;38:7687. [Google Scholar]
  • 4.Menaldino DS, Bushnev A, Sun AM, Liotta DC, Symolon H, Desai K, Dillehay DL, Peng Q, Wang E, Allegood J, Trotman-Pruett S, Sullards MC, Merrill AH. Pharmacol Res. 2003;47:373. doi: 10.1016/s1043-6618(03)00054-9. For second-generation synthesis of aminodiol compounds 3: Wiseman JM, McDonald FE, Liotta DC. Org Lett. 2005;7:3155. doi: 10.1021/ol050829o.
  • 5.Kirkland TA, Grubbs RH. J Org Chem. 1997;62:7310. doi: 10.1021/jo970877p. [DOI] [PubMed] [Google Scholar]
  • 6.(a) Albeck A, Persky R. J Org Chem. 1994;59:653. [Google Scholar]; (b) McKillop A, Taylor RJK, Watson RJ, Lewis N. Synthesis. 1994:31. [Google Scholar]; (c) Takai K, Hotta Y, Oshima K, Nozaki H. Tetrahedron Lett. 1978;27:2417. [Google Scholar]; (d) Nishizawa R, Saino T, Takita T, Suda H, Aoyagi T, Umezawa H. J Med Chem. 1977;20:510. doi: 10.1021/jm00214a010. [DOI] [PubMed] [Google Scholar]
  • 7.(a) Borg G, Cogan DA, Ellman JA. Tetrahedron Lett. 1999;40:6709. [Google Scholar]; (b) Cogan DA, Liu G, Ellman J. Tetrahedron. 1999;55:8883. [Google Scholar]; (c) Ellman JA, Owens TD, Tang TP. Acc Chem Res. 2002;35:984. doi: 10.1021/ar020066u. [DOI] [PubMed] [Google Scholar]; (d) Weix DL, Ellman JA. Org Lett. 2003;5:1317. doi: 10.1021/ol034254b. [DOI] [PubMed] [Google Scholar]; (e) Zhou P, Chen B, Davis FA. Tetrahedron. 2004;60:8003. [Google Scholar]
  • 8.We also explored addition of vinylmagnesium bromide to the sulfinimide derived from acetaldehyde, but this reaction proceeded with poor diastereoselectivity, reaching a maximum of 3: 1 dr favoring compound 8 when 8 equiv. of CH2=CHMgBr was used.
  • 9.Liu G, Cogan DA, Ellman JA. J Am Chem Soc. 1997;119:9913. [Google Scholar]
  • 10.(a) Kang SH, Jun HS. Chem Commun. 1998:1929. [Google Scholar]; (b) Hao JL, Aiguade J, Forsyth CJ. Tetrahedron Lett. 2001;42:821. [Google Scholar]; (c) Gao Y, Klunder JM, Hanson RM, Masamune H, Ko SY, Sharpless KB. J Am Chem Soc. 1987;109:5765. [Google Scholar]
  • 11.For example, see: Briot A, Bujard M, Gouverneur V, Mioskowski C. Eur J Org Chem. 2002:139.
  • 12.Garber SB, Kingsbury JS, Gray BL, Hoveyda AH. J Am Chem Soc. 2000;122:8168.Kinderman SS, Doodeman R, van Beijma JW, Russcher JC, Tjen KCMF, Kooistra TM, Mohaselzadeh H, van Maarseveen JH, Hiemstra H, Schoemaker HE, Rutjes FPJT. Adv Synth Catal. 2002;344:736. (c) The Grubbs second-generation catalyst Cl2(PCy3)(IMes)Ru=CHPh also provided metathesis products 17 and 18, but a longer reaction time was required and it proved difficult to remove all traces of the Ru catalyst from the products.
  • 13.(a) Brown HC, Vara Prasad JVN, Gupta AK. J Org Chem. 1986;51:4296. [Google Scholar]; (b) Lutjens H, Knochel P. Tetrahedron: Asym. 1994;5:1161. [Google Scholar]
  • 14.Other hydroboration reagents such as 9-BBN, or hydrosilylation directed by the C5-alcohol gave no reaction and recovery of substrate. The silyl ether compounds 17 and 18 were either unreactive or gave unsatisfactory yields under all hydroboration conditions attempted.
  • 15.Still WC, Barrish JC. J Am Chem Soc. 1983;105:2487. [Google Scholar]
  • 16.After our synthetic work was completed, we learned of another study involving hydroboration of dihydropyrrole substituted with oxygenated substituents: Cren S, Wilson C, Thomas NR. Org Lett. 2005;7:3521. doi: 10.1021/ol051232b.
  • 17.DU-145 cell cytotoxicity assays were also conducted under serum-starved conditions (protocol b, table 1), in order to minimize artifacts of protein binding, but results are also reported without serum deprivation (protocol a), since serum removal can affect sphingolipid metabolism: Colombaioni L, Frago LM, Varela-Nieto I, Pesi R, Garcia-Gil M. Neurochem Int. 2002;40:327. doi: 10.1016/s0197-0186(01)00090-0.
  • 18.Compounds 24 - 26 exhibited poor solubility in ethanol solvent vehicles, therefore experiments were conducted in a 1: 1 ethanol: ethyl acetate mixture as well as in ethanol. Control experiments indicated that addition of ethyl acetate was neither toxic to nor inhibited the growth of DU-145 cells.
  • 19.N-Cbz compounds 21 - 23 and ent-21 - 23 were also evaluated, as well as N-acetyl derivatives of 24 - 26 and ent-24 - 26, and were generally less potent cytotoxic than the cyclic amines 24 - 26 and enantiomers.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1si20060111_11
NIHMS63014-supplement-1si20060111_11.pdf (5.8MB, pdf)

RESOURCES