New PDMP analogues inhibit process outgrowth in an insect cell line

Article abstract of DOI:10.1016/j.bmcl.2004.01.013

D-threo-1-Phenyl-2-aminodecanoyl-3-morpholinopropanol (D-threo-PDMP) has previously been shown to inhibit the biosynthesis of glycosphingolipids (GSLs) in mammals and mammalian cell lines by the inhibition of glucosylceramide synthase. New D-threo-PDMP analogues were synthesized from D-serine, and found to suppress neurite extension in an embryonic insect cell line from the moth Manduca sexta, and in explanted neural tissue from insect pupae. Inhibition occurred at lower concentrations than D-threo-PDMP. The observed suppression of neurite formation was found to be reversible after the removal of the compounds. Due to their small size and short life cycle, M. sexta is shown to be an ideal model organism for studies of GSL effects in cellular development, and for drug development studies.

Full text of DOI:10.1016/j.bmcl.2004.01.013

Bioorganic & Medicinal Chemistry Letters 14 (2004) 1487–1490
New PDMP analogues inhibit process outgrowth in an
insect cell line§
Jacob P. Slavish,a Donna K. Friel,a Lynne A. Olandb and Robin Polta,*
aDepartment of Chemistry, University of Arizona, Tucson, AZ 85721, USA
bDepartment of Neuroscience, University of Arizona, Tucson, AZ 85721, USA
Received 18 May 2003; revised 22 December 2003; accepted 10 January 2004
Abstract—d-threo-1-Phenyl-2-aminodecanoyl-3-morpholinopropanol (d-threo-PDMP) has previously been shown to inhibit the
biosynthesis of glycosphingolipids (GSLs) in mammals and mammalian cell lines by the inhibition of glucosylceramide synthase.
New d-threo-PDMP analogues were synthesized from d-serine, and found to suppress neurite extension in an embryonic insect cell
line from the moth Manduca sexta, and in explanted neural tissue from insect pupae. Inhibition occurred at lower concentrations
than d-threo-PDMP. The observed suppression of neurite formation was found to be reversible after the removal of the
compounds. Due to their small size and short life cycle, M. sexta is shown to be an ideal model organism for studies of GSL effects
in cellular development, and for drug development studies.
# 2004 Elsevier Ltd. All rights reserved.
1. Introduction
Given the biological utility and pharmacological poten-
tial of d-threo-PDMP, researchers have attempted to
synthesize more potent inhibitors.9,10 Shayman et al.11
recently reported that electron-rich aromatic d-threo-
PDMP analogues gave increased inhibition of gluco-
sylceramide synthase. d-threo-40-hydroxy-1-phenyl-2-
palmitoylamino-3-pyrrolidino-1-propanol and d-threo-
(30,40-ethylenedioxy)-1-phenyl-2-palmitoylamino-3-pyr-
rolidino-1-propanol inhibited the glucosylceramide syn-
thase from Madin–Darby canine kidney (MDCK) cell
homogenates at an IC50 of 90 nM.
The glycosyltransferase inhibitor d-threo-PDMP has
displayed an ability to inhibit the enzyme glycosylcer-
amide synthase,1 which is responsible for the initial
glycosylation of ceramide.2 By inhibiting this glycosyla-
tion, d-threo-PDMP depletes the display of glyco-
sphingolipids (GSLs) on cellular surfaces,3 but also
increases intracellular ceramide levels.4 Many reports
have suggested that ceramide plays an active role in
cellular apoptosis (programmed cell death).5,6
Since d-threo-PDMP has revealed the ability to reduce
GSL levels and amplify ceramide levels, d-threo-PDMP
has been employed to halt cellular expansion and
development. Treatment of embryonic rat hippocampal
neurons with 50 mM d-threo-PDMP in culture sig-
nificantly reduced the length of the total axon plexus
and the number of axonal branch points after three days
of administration.7 Hynds et al.8 have reported that d-
threo-PDMP at 20 mM concentration inhibited neurite
outgrowth from SH-SY5Y cells, a neuroblastoma cell
line.
In this paper we report the synthesis and biological
effects of d-threo-PDMP analogues (7a–c) that bear
ether substituents on the aromatic ring; specifically, 4-
methoxy (7a), 4-tert-butoxy (7b), and 3,4-methylene-
dioxy (7c). The analogues were tested against embryonic
MRRL-CH1 cells from M. sexta to determine their
effectiveness in neurite outgrowth suppression and
cytotoxicity. Additionally, the effects on neurite exten-
sion from neuronal explants (antennal sensory neurons)
were observed. All ether-bearing analogues suppressed
cellular growth at lower concentrations than the lead
compound, d-threo-PDMP, and the suppressive effects
were reversible at lower concentrations (Scheme 1).
§Supplementary data associated with this article can be found, in the
online version at, doi:10.1016/j.bmcl.2004.01.013
* Corresponding author. Tel.: +1-520-621-6336; fax: +1-520-621-
8407; e-mail: polt@u.arizona.edu
Because of its short life cycle, a relatively simple GSL
series, and well characterized neuroanatomy and devel-
opmental profile, the insect M. sexta is an ideal model
0960-894X/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.01.013

1488
J. P. Slavish et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1487–1490
Scheme 1. Synthesis of ether-substituted d-threo-PDMP analogues.14 (a) 2equiv PPTS/THF/H 2O; (b) 1 equiv 1,10-diimidazole carbonyl/THF; (c) (i)
4 equiv HF/H2O/CH3CN; (ii) 1 equiv TsCl/pyridine; (d) 5 equiv morpholine/THF/Á (e) (i) 2M KOH (aq)/Á; (ii) 1 equiv NO2-C6H4-O2C-C9H19/0.1
equiv HOBt/pyridine.
system for the study of GSL roles in neural develop-
ment. Moreover, since this insect (tobacco hornworm,
or Sphinx moth) has been extensively studied, drug
effects can be observed at the cellular, tissue, morpho-
logical, and even behavioral studies on the adult moth
become feasible. Pupation (metamorphosis) may repre-
sent an acceptable model for metastasis in mammalian
tumors. The role of glycosphingolipids in insects
(arthrosides) may be studied more conveniently in
this organism than in mammals, where the glyco-
sphingolipids are more numerous and structurally
complex.
in >90% yield in all cases. Saponification and N-acyla-
tion completed the synthesis of the d-threo-PDMP
analogues in 61–68% yield over two steps.15
3. Biological assays on D-threo-PDMP analogues
In our analogue tests, MRRL-CH1 cells16 were exposed
to d-threo-PDMP analogues at 0.5, 1, and 5 mM for 48 h.
The results of the 24 h exposure to the d-threo-PDMP
analogues are illustrated in Figure 1. After 48 h of
exposure to the analogues, the drug media was removed
and fresh media was provided. Cells were allowed to
grow for an additional 6 days to see if the inhibition was
reversible. The results are illustrated in Figure 2. The
reversibility of each analogue is summarized in Table 2.
2. Synthesis of D-threo-PDMP analogues
Starting with the imino ester of d-serine12 (1), the
aromatic ring (yields given in Table 1) was introduced in
a stereoselective manner via a tandem reductive-alkyl-
ation reaction to provide the desired threo-isomer.13 The
threo adducts were isolated by column chromatography
and elaborated as described previously.14
Cells were originally exposed to the d-threo-PDMP
analogues at concentrations varying from 5 to 20 mM
for 48 h. In all cases, the analogues at these concentra-
tions were deemed cytotoxic. This observation was
intriguing because early testing had shown that d-threo-
PDMP exhibited only weak inhibition at 20 mM
(unpublished work). The dosage concentrations were
then reduced to 1.0 mM and 0.5 mM.
The O’Donnell Schiff base of imines 2a–c was removed
with pyridinium p-toluenesulfonate in 70% yield in all
cases. Treatment of b-amino alcohol 3a–c with 1,10-di-
imidazole carbonyl gave the carbamate 4a–c in 55–64%
yield. The silyl group was removed by aqueous HF to
afford crude hydroxyl carbamate. The crude hydroxyl
carbamate was treated with tosyl chloride in pyridine to
furnish the tosylate 5a–c in 48–83% yield over two
steps. The tosyl group was displaced with morpholine in
refluxing THF to afford the morpholino carbamate 6a–c
After the cells were exposed to 0.5 mM of the d-threo-
PDMP ether analogues, the effects were rapid. After 24
h, although normal cellular morphology and membrane
integrity were maintained, cells appeared to be unable
to form neurite extensions, and showed reduced adhe-
sion to their substrate. Prior reports have stated that
d-threo-PDMP effects on neurite extensions were not
observed until 3 days after exposure.7 Cellular adhesion
decreased, resulting in an increase in cellular aggrega-
tion compared to the untreated cells.8 The loss of
adhesion may be a consequence of the decrease in
endogeneous ganglioside levels.17
Table 1. Yields of d-threo isomer from the tandem reductive-alkyl-
ation
Table 2. Inhibitory effects of the analogues on MRRL-CH1 cell line
outgrowth
Analogue
5.0 mM
1.0 mM
0.5 mM
Aromatic Ring
d-threo-PDMP
d-threo-ADMP
d-threo-PipDMP
d-threo-BDMP
No Effect
Irreversible
Irreversible
Irreversible
No Effect
Irreversible
Irreversible
Irreversible
No Effect
Reversible
Reversible
Reversible
a
b
c
Yield
48%
44%
39%

J. P. Slavish et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1487–1490
1489
Figure 1. Effects of 0.5 mM d-threo-PDMP analogues on neurite outgrowth of MRRL-cell line.
Figure 2. Growth is normal six days after the removal of the d-threo-PDMP analogues.
After removal of the d-threo-PDMP ether analogues, a
majority of the cells were able to develop healthy
extensions, similar to the neurite extensions observed in
the untreated cells. At 1 mM, suppression of outgrowth
was also observed, and recovery was reduced. Some
recovery was observed where cells were exposed to
d-threo-PipDMP and d-threo-ADMP, but not d-threo-
BDMP. d-threo-BDMP appeared to be cytotoxic even
at 1 mM.
d-threo-PDMP has been shown to deplete cells of
endogeneous glycolipids, which can suppress cellular
outgrowth and reduce cellular adhesion. We have
synthesized three novel d-threo-PDMP ether analogues
from d-serine that reduce cellular extensions at lower
concentrations than d-threo-PDMP. The suppression of
outgrowth is reversible after the removal of the analogue.
Shayman et al.11 found that their novel d-threo-PDMP
ether analogues, which have a pyrrolidine head group in

1490
J. P. Slavish et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1487–1490
place of the morpholine, depleted endogeneous glyco-
lipids without affecting cellular growth. This observation
convolutes the understanding of the true effect(s) of
d-threo-PDMP and its analogues and the role of cell-
surface carbohydrates. Isolation of GSLs from the
Manduca sexta is also in progress.18 In the future, we
hope to gain an understanding of the perturbations of
GSL expression caused by the d-threo-PDMP ether
analogues.
75, 66, 60, 54, 50, 36, 32, 30, 29, 24, 22, 21, 15. (1R,2R)-
(+)-1- para-t-Butoxyphenyl-2-decanoylamino-3-(N-mor-
pholino)-1-propanol [D-threo-BDMP] (7b). [a]2D5=+4.5ꢀ (c
1.0, CHCl3). HRMS (FAB+H) calcd: 463.3536, found:
463.3549. 1H NMR (300 MHz, CDCl3) d 7.3 (d, J=8.8
Hz, 2H), 6.9 (d, J=8.8 Hz, 2H), 5.9 (d, J=7.3 Hz, 1H),
4.9 (d, J=2.9 Hz, 1H), 4.2 (m, 1H), 3.7 (t, J=4.6 Hz, 4H),
2.7 (m, 6H), 2.1 (t, J=7.8 Hz, 2H), 1.5 (m, 2H), 1.3 (s,
3H), 1.25 (s, 12H), 0.9 (t, J=5.8 Hz, 3H). 13C NMR
(75.0 MHz, CDCl3): d 174, 155, 136, 127, 124, 79, 76, 67,
60, 54, 51, 37, 32, 29, 26, 23, 14. (1R,2R)-(+)-1-(3,4-
Methylenedioxy)-phenyl-2-decanoylamino-3-(N-morpholino)-
1-propanol [D-threo-PipDMP] (7c). [a]2D5=+3.8ꢀ (c 1.0,
CHCl3). HRMS (FAB+H) calcd: 435.2859, found:
435.2844. 1H NMR (300 MHz, CDCl3): d 7.4–7.1 (m,
4H), 6.05 (d, J=7.3 Hz, 1H), 6.0 (s, 2H), 5.22 (d, J=2.9
Hz, 1H), 4.18 (m, 1H), 3.71 (t, J=4.6 Hz, 4H), 2.70 (dd,
J=13.1, 6.2 Hz, 1H), 2.65 (m, 5H), 2.33 (s, 3H), 2.05 (t,
J=7.8 Hz, 2H), 1.50 (m, 2H), 1.25 (s, 12H), 0.90 (t, J=5.8
Hz, 3H). 13C NMR (75.0 MHz, CDCl3) 173, 139, 134,
130, 128, 125.9, 125.7, 72, 66, 62, 60, 55, 50, 37, 32, 29.4,
29.3, 29.2, 29.1, 26, 23, 19, 14.
Acknowledgements
We thank the U.S. Army (DAMD 17-99-1-9539) for
financial support. Support for L.A.O. was partially
supplied by NIH (P01-NS28495). D.K.F. was partially
supported by a McNair Fellowship.
References and notes
16. RRL-CH1 cell line procedure: The cell line, MRRL-CH1,
which was derived originally from embryonic Manduca
sexta by Dwight Lynn (USDA), is maintained in a 10-mL
volume in 50-mL Falcon flasks in a 27ꢀ incubator in room
air. The cells grow attached to the bottom of the flask,
and are passaged approximately every 3–4 weeks by using
the outflow of a medium-filled pipette to dislodge cells
that have grown to confluence, then seeding a new flask
containing 10 mL fresh DL medium with approximately
300 mL of the cell suspension.
1. Radin, N. S.; Shayman, J. A.; Inokuchi, J. Adv. Lipid Res.
1993, 26, 183.
2. Hakomori, S. Annu. Rev. Biochem. 1981, 50, 733.
3. Kolter, T.; Sandhoff, K. Angew. Chem., Int. Ed. 1999, 38,
1532.
4. Bieberich, E.; Freischutz, B.; Suzuki, M.; Yu, R. K. J.
Neurobiol. 1999, 72, 1040.
5. Ghafourifar, P.; Klein, S. D.; Schucht, O.; Schenk, U.;
Pruschy, M.; Rocha, S.; Richter, C. J. Biol. Chem. 1999,
274, 6080.
6. Quillet-Mary, A.; Jaffrezou, J.-P.; Mansat, V.; Bordier,
C.; Naval, J.; Laurent, G. J. Biol. Chem. 1997, 272, 21388.
7. Schwarz, A.; Rapaport, E.; Hirschberg, K.; Futerman,
A. H. J. Biol. Chem. 1995, 270, 10990.
8. Hynds, D. L.; Takehana, A.; Inokuchi, J.; Snow, D. M.
Neurosci. 2002, 114, 731.
9. Abe, A.; Radin, N. S.; Shayman, J. A.; Wotring, L. L.;
Zipkin, R. E.; Sivakumar, R.; Ruggieri, J. M.; Carson,
K. G.; Ganem, B. J. Lipid Res. 1995, 36, 611.
10. Carson, K. G.; Ganem, B.; Radin, N. S.; Abe, A.; Shay-
man, J. A. Tetrahedron Lett. 1994, 35, 2659.
11. Lee, L.; Abe, A.; Shayman, J. A. J. Biol. Chem. 1999, 274,
14662.
12. O’Donnell, M. J.; Polt, R. J. Org. Chem. 1982, 47, 2663.
13. Polt, R.; Peterson, M. A.; DeYoung, L. J. Org. Chem.
1992, 57, 5469.
For the current experiments, 1.2mL DL medium and 120
mL of cell suspension from the 50-mL flask were added to
each well of a sterile 12-well plate. Analogues were added
2–4 days after initiation of the well cultures. 48 h after the
drugs were added, 0.9 mL of medium were removed from
each well and replaced with 0.9 mL fresh medium, and the
step was repeated to decrease the concentration of the
remaining analogue by 16-fold. Digital photographs were
taken just before adding the analogue, at 24 and 48 h after
analogue administration, and at 24, 48, and 144 h after
analogue removal.
Because the PDMP analogues were lipophilic, each ana-
logue was dissolved in DMSO in an amount appropriate
to yield a 100 mM stock solution, and then diluted in 1
mL DL medium to yield a 100 mM solution. Warming,
vortexing, and sonication were necessary to force the
analogues into solution, and the analogue-containing
solutions were used immediately after preparation. Fur-
ther dilutions from the 100 mM solution were made by
removing from each well the volume of medium-contain-
ing analogue to be added. All doses were tested in dupli-
cate. Vehicle control wells were given an amount of
DMSO equal to the amount of DMSO that the cells
exposed to the highest analogue dose tested received.
17. Vyas, A. A.; Patel, H. V.; Fromholt, S. E.; Heffer-Lauc,
M.; Vyas, K. A.; Dang, J.; Schachner, M.; Schnaar, R. L.
Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 8412.
14. Mitchell, S. A.; Oates, B. D.; Razavi, H.; Polt, R. J. Org.
Chem. 1998, 63, 8837.
15. (1R,2R)-(+)-1- para-Methoxyphenyl-2-decanoylamino-3-
morpholinopropan-ol [D-threo-ADMP] (7a). [a]2D5=+7.0ꢀ
(c 1.0, CHCl3). HRMS (FAB+H) calcd: 421.3066, found:
421.3076. 1H NMR (300 MHz, CDCl3) d 7.3 (d, J=8.8
Hz, 2H), 6.9 (d, J=8.8 Hz, 2H), 5.8 (d, J=7.3 Hz, 1H),
4.9 (d, J=2.9 Hz, 1H), 4.2 (m, 1H), 3.8 (s, 3H), 3.7 (t,
J=4.6 Hz, 4H), 2.7 (m, 6H), 2.05 (t, J=7.8 Hz, 2H), 1.50
(m, 2H), 1.25 (bs, 12H), 0.90 (t, J=5.8 Hz, 3H). 13C
NMR (75.0 MHz, CDCl3) d 174, 155, 136, 126, 124, 79,
18. Abeytunga, D. T. U.; Glick, J. J.; Gibson, N. J.; Oland,
L.; Somogyi, A.; Wysocki, V. H.; Polt, R., submitted.