FEMS Microbiology Letters 108 (1993) 133-138 © 1993 Federation of European Microbiological Societies 0378-1097/93/$06.00 Published by Elsevier

133

FEMSLE 05355

Correlation of cilofungin in vivo efficacy with its activity against Aspergillus fumigatus (1,3)-/3-D-glucan synthase *

Danielle Beaulieu, Julia Tang, Douglas J. Zeckner and Thomas R. Parr Jr.

Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, Indiana, USA

(Received 26 December 1992; accepted 14 January 1993)

Abstract: (1,3)-fl-o-Glucan synthase is a cell wall synthesis enzyme that is the target of cilofungin, an antifungal agent of the lipopeptide class. Cilofungin's glucan synthase inhibitory activity, MIC, and effective dose 50% in a systemic infection mouse model tend to correlate for Candida albicans. This correlation is not seen in AspergiUus fumigatus. MICs for cilofungin against A. fumigatus were consistently > 125 ~g /ml while the effective dose 50% in a systemic aspergillosis model was determined to be 20.6 mg/kg, To begin to understand this discrepancy, we examined the A. fumigatus glucan synthase. This cell wall enzyme was prepared and its activity was measured by [14C]-glucose incorporation from UDP-[U-14C]glucose into an acid insoluble polymer formed in the presence of a-amylase. Enzyme activity in crude membrane preparations was measured in the presence of several antifungal agents. Enzyme inhibition results showed that 1/zg/ml of papulacandin B, echinochandin B, aculeacin A and cilofungin all inhibited A. fumigatus glucan synthase activity (40-71%) while 1 /~g/ml of amphotericin B, fluconazole, ketoconazole and nikkomycin did not affect enzyme activity. A correlation was therefore established between the inhibitory effect of cilofungin on the A. fumigatus glucan synthase and the effective dose 50% obtained in a systemic aspergillosis mouse model.

Key words: Glucan synthase; Cilofungin; Aspergillus fumigatus; Minimum inhibitory concentrations; Animal model

Introduction

Aspergillosis is the second most common fun- gal infection requiring hospitalization of patients. Of all the species of Aspergillus recognized, A. fumigatus is the most common human pathogen

Correspondence to: (Present address) D. Beaulieu, Bristol- Myers Squibb, Pharmaceutical Research Institute, Microbiol- ogy Department (104), 5 Research Parkway, Wallingford CT, 06492, USA. * This work was presented in part at the 92nd American

Society for Microbiology General Meeting, Abst. F-78.

[1]. For many fungi, chitin and/3-(1,3)-glucans are believed to be structurally important cell wall polysaccharides [2,3]. Antifungal agents of the lipopeptide class, such as cilofungin, disrupt the synthesis of /3-(1,3) glucans, a crucial (up to 50- 60%) component of the fungal cell wall. The cellular target of cilofungin is the major cell wall synthesis enzyme (1,3)-/3-D glucan synthase (GS) (E.C.2.4.1.34; UDP-glucose: 1,3-/3-o-glucan 3-/3- D-glucosyltransferase). The in vivo (infection) and in vitro (MIC) effects of cilofungin and other lipopeptides on Candida albicans, C. tropicalis and Neurospora crassa are well known and docu-

134

mented [3-7]. The inhibitory effect of cilofungin on the glucose polymerization activity of the C. albicans GS is also well established [8-10]. How- ever, little is known about this lipopeptide's activ- ity against A. fumigatus, except that in the deter- mination of MIC values, cilofungin showed very little activity [5,11]. In contrast, in mouse protec- tion studies, animals challenged with A. fumiga- tus could be protected from death by cilofungin [11].

In this study we establish a correlation be- tween the in vivo activity of cilofungin against A. fumigatus in a murine systemic infection model and the inhibitory effect of cilofungin on the A. fumigatus GS.

Materials and Methods

Strains Aspergillus fumigatus strain WM-1 (obtained

from William Mertz, Johns Hopkins Hospital, Baltimore, MD) was used for the determination of effective doses 50% (EDs0), MICs, and for the preparation of the (1-3)-/3-o-glucan synthase (GS). A. fumigatus spores were stored at -80°C in a freezing solution composed of 5% lactose and 10% glycerol.

Antifungal MIC determinations MICs against A. fumigatus were determined

by a standard micro broth dilution method [12]. Cilofungin, fluconazole, and amphotericin B were used in concentrations varying between 125 /xg/ml and 0 .075 / ,g /ml , in two-fold serial dilu- tions. Antibiotic 3 liquid medium was used and final spore concentration per well (200 /,1 final volume) was 5 x 105 CFU.

ED5o determination EDs0 for cilofungin, fluconazole and ampho-

tericin B were determined in an A. fumigatus systemic murine infection model based on that described by Denning and Stevens [11]. Animals used were male CD-1 mice, 18-20 g, ten mice per group. Twenty-four hours prior to infection, the mice were immunosuppressed with 400R X- irradiation. Each mouse was infected i.v. with

l x 107 spores of A. fumigatus strain WM-1, via the lateral tail vein. Treatment regimens were as follows: cilofungin twice a day for 10 days i.p., fluconazole twice a day for 5 days p.o., and am- photericin B 0.4 h, 2, 4 and 6 days post-infection i.p. Dosing intervals were selected based on the pharmacokinetics of the compounds in mice. Doses tested were as follows: cilofungin and flu- conazole 100, 50, 25 and 12.5 mg/kg, ampho- tericin B 3, 1.5, 0.75, 0.38, 0.20, 0.10, 0.05 mg/kg. Survival of the mice was recorded over 15 days. Mice were given free access to food and water for the duration of the experiment. EDs0 values were estimated by the method of Reed and Muench [13].

Preparation of the A. fumigatus GS A. fumigatus mycelial cells were grown in

Sabouraud Dextrose broth, inoculated with ap- prox. 1.44 x 104 spores/ml, and incubated at 35°C for 24 h. Cells were recovered by filtration on Corning disposable sterile nylon filters, 0.22/~M. Filtered cells were washed with cold water, fil- tered dry and stored frozen at - 80°C. Two to 2.5 g of cells were resuspended in 15 ml total volume of cold extraction buffer (50 mM Hepes pH 7.7, 1 M sucrose, 50 mM NaC1, 1 mM DTT, 5 mM EDTA, 10 mM NaF, and 0.1 mM GTP), then broken in a Bead Beater (Biospec Products, Bartlesville, OK), using the 50-ml chamber and approximately 17 ml of dry 500-/,M pre-cooled glass beads. The lysing chamber was contained within a larger one filled with a salt-ice mixture for cooling purposes. Four to five cycles were required to disrupt the cells effectively; each cy- cle consisted of 15 s beating followed by a 3 min cooling period. The lysed cell suspension was recovered from the beads with a Pasteur pipette and the beads were washed twice with cold ex- traction buffer. The final volume of the lysate was 15-20 ml. Cleared lysates were obtained by low speed centrifugation (2000 x g, 4°C, 15 min). The lysates were then submitted to a high speed cen- trifugation (100 000 x g, 4°C, 60 min) in order to pellet the membranes. The supernatants had min- imal enzyme activity and were discarded. The pellets were homogenized in 20 ml of cold extrac- tion buffer using a cooled Wheaton-type homoge-

nizer. The membranes were then pelleted a second time by ultracentrifugation (as above) and the pellets were homogenized in 5-10 ml of cold extraction buffer. Protein concentrations were determined with the BioRad Protein Assay Con- centrate according to the manufacturer 's instruc- tions and adjusted to 1 m g / m l with extraction buffer. Membranes were then frozen in a dry ice/methanol bath and stored at -80°C. This material was defined as the membrane prepara- tions.

G S activity assay Polymerization activity of the GS present in

membrane preparations was based on the assay described by Tang and Parr [10], with some modi- fications. Briefly, a reaction mixture composed of 2 m g / m l a-amylase (Sigma Chemicals, St. Louis, MO), 3 m g / m l BSA (Sigma), 0.4 mM U D P G (Sigma), and 3.14 /~M UDP-[U-14C]glucose (1 /zCi/ml) (Amersham, Arlington Heights, IL), in PH 7.7 extraction buffer was mixed with 100/zg of protein in a 1 :4 ratio (100 /xl of membrane preparation +33 /zl of reaction mixture). Reac- tions were done at room temperature (membrane preparations and reaction mixtures were pre- warmed to room temperature) for 15 min and stopped by adding 1 ml of cold 5% trichloroacetic

135

acid (TCA), which precipitated the glucan poly- mer. The precipitated polymers were recovered by vacuum filtration on a Millipore filtering unit using 0.45/xM filters (Type HA filters, Millipore Corporation, Bedford, MA). Filters were rinsed with 5% cold TCA, dried and radioactivity pre- sent on the filters was counted.

Results and Discussion

M I C vs. ED5o MIC s and EDs0 of cilofungin, amphotericin B,

and fluconazole were determined for A. fumiga- tus strain WM-1 and values obtained are re- ported in Table 1. Figure 1 shows the survival of mice treated with cilofungin for the duration of the 15 day experiment. This figure shows a dose response to cilofungin reflected in the lower number of surviving animals at lower doses of cilofungin. Controls for the EDs0 determination were a group of animals treated with ampho- tericin B, one treated with fluconazole and one non-treated group (See Table 1 for values). Mean day of death for untreated controls was 4.3 days. Cilofungin showed efficacy in protecting mice in- fected with A. fumigatus , with an EDs0 of 20.6 mg/kg. The in vivo activity of cilofungin against

Table 1

Comparison of experimental and published EDs0 values and MIC values for A. fumigatus

Antifungal MIC EDso Source strain ~g/ml mg/kg

Cilofungin A. fumigatus WM-1 > 125 20 * A. fumigatus 10AF/8610 > 50 ND

Fluconazole A. fumigatus WM-1 > 125 > 100 * A. fumigatus HO604 ND 47-71 * A. fumigatus Hll-20 > 25.6 § > 80 ~:

Amphotericin B A. fumigatus WM-I 0.312 0.3 * A. fumigatus HO604 ND 0.1-0.9 * A. fumigatus Hll-20 0.4 § 4 :~

This study Denning and Stevens [11]

This study Troke et al. [20] Schmitt et al. [17]

This study Troke et al. [20] Schmitt et al. [17]

* Animal model used was a systemic mouse aspergillosis. * Animal model used was a mouse lung infection with strains Hll-20. § MIC values are the mean of 16 A. fumigatus strains. ND: Not determined.

136

A. fumigatus was not re f lec ted in the high M I C values.

I t is well known that the d e t e r m i n a t i o n of meaningfu l MICs of an t i fungal c o m p o u n d s is dif- ficult [14-16]. In this study, MICs for ampho- ter ic in 13 and f luconazole a re s imilar to those r e p o r t e d by Schmit t et al. [17] (See Tab le 1). In genera l , M I C values co r r e spond to EDs0 values for ampho te r i c in B and f luconazole . In the case of ci lofungin, high M I C values for A. fumigatus ( > 1 2 5 / x g / m l ) would p red ic t tha t this c o m p o u n d would be inact ive in vivo. The ED~0 in a mur ine systemic infect ion mode l was found to be m o d e r - a te at 20.6 m g / k g . Ci lofungin is known to be effect ive agains t C. albicans [4,12], and A. fumi- gatus [11], as the ED50 would indicate . Knowing tha t c i lofungin is effect ive in vivo agains t C. albi- cans, tha t the C. albicans GS is inh ib i ted in vi tro by ci lofungin, we hypo thes ized that the A. fumi- gatus might have a GS which could be inh ib i ted by ci lofungin.

Act ive GS was successful ly p r e p a r e d f rom A. fumigatus mycel ia l ceils. C r u d e m e m b r a n e p r e p a - ra t ions were able to synthesize glucose po lymers in vitro. The effect of known GS inhibi tors on these m e m b r a n e p r e p a r a t i o n s was then assayed.

Inh ib i t ion of GS activity by d i f ferent an t i fungal agents was assayed at 10, 1 and 0.5 / x g / m l d rug concent ra t ions . Tab le 2 shows the resul ts of these

oo I i i H ~ • o 9o ! ~ " I I . e 4

80 I ~ J ~ I . . ~ 1 . I . I I • e .

"~ 60 ~ 5O ~ 40

3O 20 10

0 0 I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

D a y s Fig. 1. Percent survival of immunosupressed mice treated with varying doses of cilofungin. (B) 100 mg/kg, (e) 50 mg/kg, (D) 25 mg/kg, (A) 12.5 mg/kg, (o) control, during a sys- temic Aspergillus fumigatus infection. Survival data were used in the Reed and Muench method [13] for the calculation of

the effective dose 50% for cilofungin.

Table 2

Percentage of inhibition of the AspergiUus fumigatus (1,3)-/3- o-glucan synthase by different antifungal compounds at 10, 1 and 0.5/zg/ml

Antifungal % Inhibition at compound 10/~g/ml 1/zg/ml 0.5 #g/ml

Lipopeptides cilofungin 78 69 60 echinocandin B 74 49 27 aculeacin A 76 71 60

Other classes papulacandin B 80 40 21 fluconazole 3 0 0 ketoconazole 6 0 0 amphotericin B 18 0 0 nikkomycin 7 0 0

All drugs were dissolved in 100% dimethyl sulfoxide (DMSO) at 100× the target concentration and then diluted to 10× the target concentration with water. Reactions were done as described above except that 90/xl of glucan synthase prepara- tion + 10/xl of the 10 × concentrated drug were used instead of 100 #1 of glucan synthase preparation. Control reactions contained corresponding DMSO concentrations instead of the drug. The drug and the glucan synthase preparation were allowed to interact for 5 rain before the reactions were started. Processing of the reactions was the same as described in Materials and Methods. Percentage of inhibition was deter- mined by comparison of the counts in the drug containing reaction with the solvent controls.

exper iments . An t i funga l s be longing to the l ipo- p e p t i d e class were able to inhibi t GS activity to varying degrees . Pa pu l a c a nd in B, and ant ib io t ic known to inhibi t f l -g lucan synthesis [18] inhib i ted po lymer iza t ion activity in this assay. O t h e r classes of an t i fungal drugs were not effect ive in inhibi t - ing the po lymer iza t ion activity of the GS (see Table 2).

I t has been d e m o n s t r a t e d in Saccharomyces cerevisiae that mutan t s res is tan t to acu leac in A, a l i popep t ide , a re not d i f ferent f rom the p a r e n t s t ra in with respec t to cell wall po lysacchar ide compos i t ion and in vi tro GS activity, but tha t the re is a major d i f fe rence in cell surface hy- d rophobic i ty [19]. T h e s e au thors sugges ted that this d i f fe rence poin ts to the impor t a nc e of cell surface hydrophobic i ty in the accessibi l i ty of the ta rge t to the ant ib io t ic molecules . One reason why l i p o p e p t i d e M I C values for A. fumigatus are always very high (like they would be for a resis-

tant strain) might be because growth condi t ions used in these exper iments are such that the cell surface hydrophobici ty is changed and access to the GS target is restricted. This would lead to an apparen t lack of sensitivity in vitro to the l ipopeptide. W h e n A. fumigatus is grown in vivo, for example in an infected mouse, the growth condi t ions may be al tered so that cell surface hydrophobicity is such that the l ipopept ide now has access to the GS. These condi t ions would lead to modera te EDs0 values, indicative of greater susceptibil i ty to the l ipopept ide in vivo.

In conclusion, we established a correla t ion be- tween in vivo activity of ci lofungin against A. fumigatus, and GS inhibi t ion by this l ipopeptide, despite the fact that MIC values against this fungus are always high. In vitro inhibi t ion of this GS could provide a useful method to predict in vivo activity of l ipopept ide compounds against A.

fumigatus.

Acknowledgements

We would like to thank the following people for excellent technical support: L. Thomas, C. Boylan, P. Raab and T. Butler of the Lilly Core Lab, and Jim Farmer . We would also like to thank Rober t Gordee for giving us strains and Thal ia Nicas and Dave Pres ton for their helpful suggestions on this manuscr ipt .

References

1 Rogers, A.L. and Kennedy, M.J. (1991) In: Manual of Clinical Microbiology (Balows, A., Haussler Jr., N.J., Herr-

137

man, K.L., Isenberg, H.D. and Shadomy, H.J., Eds.), pp. 659-673. 5th Edn., American Society for Microbiology, Washington, DC.

2 Bartnicki-Garcia, S. (1968) Annu. Rev. MicrobioL 22, 87- 108.

3 Borgia, P.T. and Dodge, C.L. (1992) J. Bacteriol. 174, 377-383.

4 Gordee, R.S., Zeckner, D.J., Ellis, L.F., Thakkar, A.L. and Howard, L.C. (1984) J. Antibiot. 37, 1054-1065.

5 Gordee, R.S., Debono, M. and Parr Jr., T.R. (1992) In: New Approaches for Antifungal Drugs. (Fernandes, P.B., Ed.), pp. 46-63. Birkhauser, Boston, MA.

6 Hobbs, M., Perfect, J. and Durack, D. (1988) Eur. J. Clin. Microbiol. 7, 80-81.

7 Pfaller, M., Gordee, R., Gerarden, T., Yu, M. and Wen- zel, R. (1989) Eur. J. Clin. Microbiol. Infect. Dis. 8, 564-567.

8 Sawistowska-Schr6der, E.T., Kerridge, D. and Perry, H. (1984) FEBS Lett. 173, 134-138.

9 Taft, C.S., Stark, T. and Selitrennikoff, C.P. (1988) Anti- microb. Agents Chemother. 32, 1901-1903.

10 Tang, J. and Parr Jr., T.R. (1991) Antimicrob. Agents Chemother. 35, 99-103.

11 Denning, D.W. and Stevens, D.A. (1991) Antimicrob. Agents Chemother. 35, 1329-1333.

12 Gordee, R.S., Zeckner, D.J., Howard, L.C., Alborn Jr., W.E. and Debono, M. (1988) Ann. NY Acad. Sci. 544, 294-309.

13 Reed, L.J. and Muench, H. (1938) Am. J. Hyg. 27, 493-497. 14 Hanson, L.H. and Stevens, D. (1989) Antimicrob. Agents

Chemother. 33, 1391-1392. 15 Galgiani, J.N. (1987) Antimicrob. Agents Chemother. 31,

1867-1870. 16 Galgiani, J.N., Reiser, J., Brass, C., Espinel-Ingroff, A.,

Gordon, M.A. and Kerkering, K.M. (1987) Antimicrob. Agents Chemother. 31, 1343-1347.

17 Schmitt, H.J., Edwards, F., Andrade, J., Niki, Y. and Armstrong, D. (1992) Exp. Chemother. 38, 118-126.

18 Baguley, B.C., R6mmele, G., Gruner, J. and Wherli, W. (1979) Eur. J. Biochem. 97, 345-351.

19 Font de Mora, J., Gil, R., Sentandreu, R. and Herrero, E. (1991) Antimicrob. Agents Chemother. 35, 2596-2601.

20 Troke, P.F., Andrews, R.J., Marriott, M.S. and Richard- son, K. (1987) J. Antimicrob. Chemother. 19, 663-670.