Antifungal activity of a Candida albicans GGTase I inhibitor-alanine conjugate. Inhibition of Rho1p prenylation in C. albicans

Article abstract of DOI:10.1016/S0960-894X(03)00320-2

An alanine conjugate of a Candida albicans geranylgeranyl transferase I inhibitor was synthesized to facilitate its uptake into the fungal cell. The antifungal activity of CaGGTase-Ala conjugate is demonstrated. It is also shown that the CaGGTase-Ala conjugate affects prenylation of endogenous Rho1p, but has no effect on prenylation of endogenous Ras1p.

Full text of DOI:10.1016/S0960-894X(03)00320-2

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1935–1937
Antifungal Activity of a Candida albicans GGTase I
Inhibitor-Alanine Conjugate. Inhibition of Rho1p Prenylation
in C. albicans
Krishna K. Murthi,a,* Susan E. Smith,b Arthur F. Kluge,a
Gustave Bergnes,a Patrick Bureaua and Vivian Berlinb
aGPC Biotech, Inc., Department of Chemistry, 610 Lincoln Street, Waltham, MA 02451, USA
bGPC Biotech, Inc., Department of Yeast Genetics, 610 Lincoln Street, Waltham, MA 02451, USA
Received 15 November 2002; accepted 10 February 2003
Abstract—An alanine conjugate of a Candida albicans geranylgeranyl transferase I inhibitor was synthesized to facilitate its uptake
into the fungal cell. The antifungal activity of CaGGTase-Ala conjugate is demonstrated. It is also shown that the CaGGTase-Ala
conjugate affects prenylation of endogenous Rho1p, but has no effect on prenylation of endogenous Ras1p.
# 2003 Elsevier Science Ltd. All rights reserved.
b-1,3 d-glucan is an essential component of the fungal
cell wall1 and the enzyme b-1,3 d-glucan synthase,
which regulates the synthesis of b-1,3 d-glucan has been
targeted for antifungal drug discovery.2 Candida albi-
cans Rho1p is required for cell viability, plays a role in
cell integrity and is a key regulator of b-1,3 d-glucan
synthase.3,4 Rho1p is a small GTP binding protein that
plays an essential role in cellular integrity and mor-
phology of fungal cells.5 Geranylgeranyl transferase I
(GGTase I), a zinc metalloenzyme, catalyses the transfer
of a geranylgeranyl group to Rho1p which is required
for its functional localisation to the membrane.6 The
essentiality of GGTase I in C. albicans was validated
genetically by knock-down experiments.7 Our antifungal
drug discovery strategy targeted C. albicans GGTase I
with the premise that a GGTase I inhibitor would block
the b-1,3 d-glucan synthase pathway, disrupt cell wall
formation and result in lysis and cell death.
Compound 1 showed no antifungal activity in cellular
assays. We reasoned that the lack of activity of 1 in whole
cells was due to its inability to enter the fungal cell.
Fungal cells utilize active transport systems such as per-
meases to transport nutrients such as short peptides and
amino acids into the cell. This transport system has been
used to transport meta-fluorophenylalanine9 into fungal
cells and achieve inhibition of growth. By analogy to this
work we decided to modify the CaGGTase I inhibitor 1 to
be a potential substrate for the permease transport system.
The substrate specificity of the prenyl transferases,
GGTase I and farnesyl transferase (FTase) has been
reported.8 We designed C. albicans GGTase I inhibitors
based on the terminal tetrapeptide sequence CA1A2L,
which is the key recognition element of GGTase I.
Replacement of the A1A2 dipeptide with a 4-phenyl
piperidine spacer lead to the very potent C. albicans
GGTase I inhibitor (IC50: <5 nM) 1.
Chemistry
The alanine conjugate 3 was designed to be a substrate for
active transport and was synthesized as shown in Figure
1.10 Boc-4-phenylpiperidine-4-carboxylic acid was cou-
pled to the dipeptide Leu-Ala-OtBu (EDC, HOBt, DIEA,
CH2Cl2) to give intermediate 2. Tert-butoxycarbamate
deprotection (TFA, CH2Cl2), S-Tr-N-Boc-cysteine cou-
pling (EDC, HOBt, DIEA, CH2Cl2), followed by overall
deprotection (TFA, CH2Cl2, Et3SiH) gave 3.
*Corresponding author. Fax: +1-781-890-9005; e-mail: krishna.
murthi@gpc-biotech.com
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00320-2

1936
K. K. Murthi et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1935–1937
Figure 1. Synthesis of alanine conjugate.
Discussion
As shown in Table 1, compounds 1, 1a and 3 are all
potent inhibitors of CaGGTase I but not CaFTase, as
revealed by using recombinant components in appro-
priate in vitro prenylation assays similar to those pre-
viously described.11 Neither 1 nor 1a afforded any
antifungal activity against SC5314 cells, a well-studied
C. albicans clinical isolate (MIC >128 mg/mL). In con-
trast, the alanine conjugate 3 showed potent antifungal
activity with a minimum inhibitory concentration
(MIC) of 1 mg/mL and a minimum fungicidal concen-
tration (MFC) of 2 mg/mL (Table 1).
Figure 2. Compound 3 affects the localisation of C. albicans Rho1p but
not Ras1p. Lanes 1–3 DMSO treated samples, lanes 4–6 compound 3
treated samples. Lanes 1and 4, whole cell extracts; lanes 2 and 5, cyto-
solic fractions; lanes 3 and 6, membrane fractions. Protein marker
molecular weights (kD) are indicated. Ã And # indicate proteins cross-
reacting with the anti-Rho1p and anti-Ras antibodies, respectively.
of Ras1p, an FTase substrate, using a monoclonal anti-
body (Molecular Probes, panRAS) that recognises a
sequence conserved in C. albicans Ras1p (data not
shown). The localisation of the ꢀ44 kD C. albicans
Ras1p15 to the membrane fraction was unaffected by
either DMSO alone or compound 3 (Fig. 2). These
results indicated that compound 3 specifically inhibited
GGTase I in vivo and did not generally affect prenyl-
ation in C. albicans cells. It is interesting that compound
3 effects cell killing after 5 h of treatment. This obser-
vation could be explained if the geranylgeranylation of
only newly synthesised Rho1p (i.e., that synthesised by
the cell to compensate for the turn-over of Rho1p and/
or depletion as a result of cell division) rather than pre-
existing geranylgeranylated Rho1p, was inhibited. Fur-
ther work is necessary to determine if this is the case,
but in support of this idea, we have observed that a
large proportion of tagged Rho1p expressed from a
heterologous, regulatable promoter was localised to the
cytosol in the presence, but not the absence, of com-
pound 3.
Presumably compound 3 is transported into the fungal
cell by an amino acid or peptide permease. To test this
hypothesis, the MIC was determined in the presence of
amino acids in the media which could compete for the
permease. The addition of amino acids at a concen-
tration of 1mM abolished the activity of compound 3 as
indicated by an MIC >64 mg/mL suggesting that trans-
port by a peptide transport system was responsible for
its activity.
Next, given that compound 3 has antifungal activity
and appeared to enter into cells, we asked whether this
compound affected the activity of GGTase I in C. albi-
cans cells. To do this, we studied the cellular localisation
of the GGTase I substrate, Rho1p. Treatment of
C. albicans cells with compound 3 should, by inhibition
of the CaGGTase I, prevent the membrane targetting of
Rho1p mislocalising it to the cytosol. Initially, a kill
curve experiment13 was carried out to determine both
the appropriate concentration and timeframe over
which compound 3 acted. We found that at a concen-
tration of 3 mg/mL, cell killing occurred 5 h after expo-
sure to compound 3. These conditions were then used in
a prenylation assay;14 SC5314 cells were treated with
compound 3, extracts were generated and subjected to
high speed centrifugation. The localisation of Rho1p
was studied by Western blot analysis using a specific
Rho1p peptide antibody.6 In mock, DMSO treated
C. albicans cells, Rho1p was exclusively localised to the
membrane fraction (Fig. 2).
Taking these results together the apparent inhibition of
GGTase I in vivo correlates with cidal activity of this
compound.
References and Notes
1. Georgopapadakou, N. H.; Tkacz, J. S. Trends Microbiol.
1995, 3, 98.
2. (i) Stone, E. A.; Fung, H. B.; Kirschenbaum, H. L. Clinical
Therapeutics 2002, 24, 351. (ii) Denning, D. W. J. Anti-
microbial Chemotherapy 2002, 49, 889.
In contrast, cells treated with compound 3 showed a
significant proportion of Rho1p in the cytosolic fraction
(Fig. 2). As a control, we also analysed the localisation
3. Kondoh, O.; Tachibana, Y.; Ohya, Y.; Arisawa, M.;
Watanabe, T. J. Bact. 1997, 179, 7734.
4. Smith, S. E.; Csank, C.; Reyes, G.; Ghannoum, M. A.;
Berlin, V. FEMS Yeast Res. 2002, 2, 103.
Table 1. Antifungal activity12 of compounds 1, 1a and 3
5. Cabib, E.; Drgonova, J.; Drgon, T. Annu. Rev. Biochem.
1998, 67, 307.
Compd CaGGTase I
IC50, (mM)
CaFTase
IC50, (mM)
MIC mg/mL MFC mg/mL
6. Omer, C. A.; Gibbs, J. B. Mol Microbiol. 1995, 11, 219.
7. C. albicans GGTase I is composed of 2 non-identical sub-
units, a and b. We cloned the genes encoding these subunits
and to assess their requirement for C. albicans viability we
employed 2 alternative but complementary approaches. The
hemizapper or pop-in/pop-out method (Scherer, S.; Davis,
1
<0.005
0.005
0.013
0.2
NT
>128
>128
>128
>128
1a
3
1
2

K. K. Murthi et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1935–1937
1937
R.W. Proc. Natl. Acad. Sci. U.S.A., 1979, 76, 4951) permitted
a statistical analysis to determine whether a gene is essential.
Using this approach we were unable to recover homozygous
null mutants for either the a- or the b- subunit genes indicating
that GGTase I is essential (P>0.995). In addition, we placed
the a- and b- subunit genes under the control of a regulatable
promoter (see ref 4 for an example). Significant reduction of
either of the GGTase I subunits resulted in a loss of cell via-
bility, cell lysis and mislocalisation of geranylgeranyl transfer-
ase substrates (manuscripts in preparation).
8. Zhang, F. L.; Casey, P. J. Annu. Rev. Biochem. 1996, 65, 241.
9. Kinsbury, W. D.; Bochm, J. C.; Mehta, R. J.; Grappel, S. F.
J. Med. Chem. 1983, 26, 1725.
10. Bergnes, G.; Berlin, V.; Come, J.; Kluge, A.; Murthi, K.;
Pal, K.; U.S. Patent 6423519, 2002.
11. (i) Stirtan, W. G.; Poulter, C. D. Biochem. 1997, 36, 4552.
(ii) Gomez, R.; Goodman, L. E.; Tripathy, S. K.; O’Rourke, E.;
Manne, V.; Tamanoi, F. Biochem. J. 1993, 289, 25. (iii) Caplin,
B. E.; Marshall, M. S. Methods Enzymol. 1995, 250, 51 .
12. MIC and MFC determinations followed standard proce-
dures using SC5314 C. albicans cells (Fonzi, W. A.; Irwin, M.
Y. Genetics, 1993, 134, 717) except that YNBGlc medium was
used: YNB w/o amino acids (Difco) supplemented with 1 mg/mL
histidine, 2 mg/mL methionine, 2 mg/mL tryptophan, 200 mg/mL
glutamine and 2% glucose (all obtained from Sigma). To deter-
mine the effect of high amino acid concentration on the MIC
this medium was supplemented with an amino acid mixture at
a final concentration of 1mM (Sigma).
13. Kill curve methodology. A culture of SC5314 in YNBGlc
medium was set up at 35 ꢁC and shaken overnight at 220 rpm.
The SC5314 cells were then washed, resuspended in fresh
YNBGlc and counted. 2 flasks containing 100 mL YNBGlc
were inoculated with 1Â109 (1Â107 cells/mL) SC5314 cells.
Either compound 3 or vehicle (DMSO) was added to the cul-
tures which were then incubated at 220 rpm, 35 ꢁC. At various
time points, a range of appropriate serial dilutions (10À5–10À9
)
of the cultures were made and 100 mL cells plated onto
Sabourand plates. The plates were incubated at 35 ꢁC for 48 h
prior to counting.
14. Prenylation assay. A culture of SC5314 in YNBGlc med-
ium was set up at 35 ꢁC and shaken overnight at 220 rpm. The
following morning the cell number was determined and cells
were resuspended at 1Â107 cells/mL in YNBGlc and treated
with either compound 3 at 3 mg/mL or an equivalent volume of
DMSO vehicle. Cells were harvested after 5 h exposure to
compound, a time based on the killing kinetics of compound 3
(not shown). After the incubation, the cells were washed with
1M sorbitol, transferred to 2 mL screwcap tubes, pelleted and
frozen at À80 ꢁC. Whole cell extracts were generated and 50–
100 mL was subjected to high speed centrifugation (4 ꢁC,
48,000 rpm for 1h using a TI120.1 rotor in a Beckmann
tabletop centrifuge) to resolve the membrane and cytosolic
fractions. SDS.PAGE and western blotting analysis were then
carried out.
15. Feng, Q.; Summers, E.; Guo, B.; Fink, G. J. Bacteriol.
1999, 65, 241.