A diffusible analogue of N3-(4-methoxyfumaroyl)-L-2,3-diaminopropanoic acid with antifungal activity

Systemic fungal infections, caused mainly by the opportunistic pathogen Candida albicans, are regarded as one of the most important problems to be solved in modern chemotherapy (Polak & Hartman, 1991 ; Ghannoum, 1997 ). None of the antimycotics used in clinical practice meets all the criteria expected for a good chemotherapeutic agent, so there is a strong need for novel compouds. Potentially interesting and valuable agents in chemotherapy are amino acid analogues. They act mainly as enzyme inhibitors. Unfortunately, their use is often limited by restrictions on their transport into cells. Amino acid permeases are the only means of transport for these compounds, but they are rather selective in their action and do not take up the more highly modified analogues. Consequently, only a limited number of such compounds exhibit activity against cells. Examples of such amino acid analogues include some antifungal agents: azoxybacillin, cispentacin, BAY 10-8888 and RI 331 (Aoki et al., 1994 ; Konishi et al., 1989 ; Capobianco et al., 1993 ; Ziegelbauer et al., 1998 ; Yamaki et al., 1988 ). Bypassing the carrier transport by appropriate modification of the amino acid analogues may result in free diffusion of novel derivatives. Those modifications have become a challenge in the design of an active agent.

In our research programme, aimed at the development of novel antifungal agents, we have focused on a glutamine analogue, N³-(4-methoxyfumaroyl)-L-2,3-diaminopropanoic acid (FMDP), a potent and selective inhibitor of glucosamine-6-phosphate synthase (GlcN-6-P synthase) EC 2 . 6 . 1 . 16 from C. albicans (Andruszkiewicz et al., 1986 , 1993 ; Milewski et al., 1985 ). This enzyme, which plays a key role in the biosynthesis of glucosamine-containing microbial cell wall macromolecules, chitin, mannoprotein and peptidoglycan, is considered to be a target for potential antimicrobial drugs (Chmara et al., 1986 ; Milewski et al., 1986 ; Borowski, 2000 ).

In our previous studies it has been found that FMDP, an amino acid analogue, exhibits only moderate antifungal activity and the presence of some amino acids, especially glutamine and glutamic acid, strongly decreases its antifungal efficacy (Cybulska et al., 1997 ). In order to overcome this problem, we have performed chemical modifications of the FMDP molecule aimed at the construction of latent lipophilic derivatives, which could be transported into cells by diffusion. Following uptake, the modifying group could be removed intracellularly. Very commonly, the development of pro-drugs, in particular in the penicillin group, involves the formation of esters or other easily hydrolysable derivatives of the active agent in order to increase penetration of compounds into the cell (Daehne et al., 1970 ). Recently, we synthesized and examined the FMDP derivative acetoxymethyl ester. The structure of this compound is shown in Fig. 1. This novel derivative inhibits the activity of isolated pure GlcN-6-P synthase and its enzyme inhibitory potency is only a few-fold lower than that of the unsubstituted FMDP (IC₅₀=3·2 µM for FMDP and 11·5 µM for the FMDP ester; Zgódka et al., 1999 ). In this paper, we present evidence that the FMDP acetoxymethyl ester penetrates into the fungal cells by free diffusion and, once inside, generates free FMDP upon enzymic cleavage. As a consequence, the compound inhibits growth of C. albicans cells. Our results show that the pro-drug approach can be successfully applied to the design of derivatives of GlcN-6-P synthase inhibitors which bypass the amino acid carrier system and exhibit good antifungal activity.

(6K): Fig. 1. Structures of FMDP and the FMDP acetoxymethyl ester.

Chemicals.
FMDP and the acetoxymethyl ester of FMDP were synthesized in our laboratory by methods described previously (Andruszkiewicz et al., 1986 ; Zgódka et al., 1999 ). 2,4,6-Trinitrobenzene sulfonic acid (TNBS) was from Sigma. Other chemicals were of the highest grade commercially available. Filter disks were manufactured by Whatman.

Organisms.
Candida albicans ATCC 26278 cells were stored on Sabouraud Dextrose Agar slants (1% Bacto peptone, 1% yeast extract, 2% glucose, 2% agar) at 4 °C and transferred monthly.

Growth conditions.
Minimal medium (YNBG) contained 1·7 g YNB (Yeast Nitrogen Base; Difco), 20 mg L-Trp, 20 mg L-Met, 10 mg L-His and 10 g glucose in 1000 ml water. Overnight cultures in Sabouraud Dextrose medium were diluted in fresh medium to a concentration of 10⁶ cells ml^-1 and incubated for about 2 h at 30 °C with shaking (200 r.p.m.) to obtain exponential-phase culture cells.

Growth experiments
(a) Antifungal susceptibility tests.
To determine the 50% inhibitory concentrations (IC₅₀) and MICs of FMDP and its ester, C. albicans cells grown overnight in Sabouraud Dextrose medium at 30 °C were inoculated at a cell density of 10⁴ ml^-1 in YNBG medium containing the test compounds at concentrations of 0400 µM. Cultures were incubated for 48 h at 30 °C, and cell growth was quantified by measuring the OD₆₆₀ to determine the IC₅₀ values. The MIC was defined as the lowest drug concentration preventing visible growth.

(b) Determination of kinetics of growth inhibition.
YNBG liquid medium was inoculated with 10⁶ yeast cells ml^-1 from the overnight culture in Sabouraud broth. After a 3 h preincubation at 30 °C with shaking (200 r.p.m.), the compounds were added to give a final concentration in the range of 0160 µM. Incubation was continued for 8 h under the same conditions. Growth was determined spectrophotometrically at 660 nm at hourly intervals.

Uptake studies.
For uptake studies, exponential-phase C. albicans cells grown in Sabouraud Dextrose medium were washed with 50 mM potassium phosphate buffer, pH 7·0, containing 1% (w/v) glucose and suspended in the same buffer to a cell density of 10⁷ cells ml^-1 (or 10⁶ cells ml^-1 to determine the effects of cell density on FMDP ester uptake). Cell suspensions were preincubated for 15 min at 30 °C and uptake was started by adding FMDP or its ester at final concentrations from 0·05 to 4 mM. At that moment, and at 5 min intervals thereafter, 2 ml samples of the cell suspensions were withdrawn, immediately collected on filter disks (GF/A Whatman filters, pore size 0·22 µm), and the filtrates were used to determine the concentration of the antifungal agents. Then, 1 ml samples of the filtrates were taken and combined with 1·25 ml of a solution containing 4% Na₂B₄O₇ . 10H₂O and 0·8 mg TNBS ml^-1. Reactions were carried out at 37 °C for 30 min. The A₄₂₀ was measured and concentrations of FMDP or its ester were read from standard curves. Data were plotted as nmol amino acid (or amino acid ester) taken up by 1 mg (dry wt) cells versus time. The initial uptake velocities were determined from the slopes of the linear part of the curves in the 010 min region.

In some experiments, cells were preincubated with the inhibitor or ionophore for the indicated times at 30 °C, before the determination of uptake by the antifungal agents.

Analysis of FMDP ester metabolism.
Cell-free extracts from C. albicans were prepared by the method described previously (Milewski et al., 1991 ). FMDP ester was added to the extract, to give a final concentration of 100 µM. The mixture was incubated at 30 °C, and 1 ml samples were collected at 1 min intervals and deproteinized by addition of 1 ml ethanol. Precipitate was removed by filtration and the supernatant was analysed by TLC. TLC analyses were performed on Kieselgel 60 F 254 (Merck) and cellulose plates (Merck) in the following solvent systems: A: n-propanol:NH₃:CHCl₃ (12:8:1, by vol.); B: n-propanol:H₂O (7:3, v/v). The filtrates obtained after cell harvesting in uptake studies were also subjected to TLC analysis in the same systems.

Growth inhibitory effect of FMDP and the FMDP acetoxymethyl ester
The acetoxymethyl ester of FMDP exhibited a stronger antifungal effect than FMDP. A 50% reduction of fungal growth was noticed upon the action of 3 µM FMDP ester and for 10 µM FMDP. In terms of the MIC values, the difference between both agents was even higher (3·1 µg ml^-1/10 µM and 50 µg ml^-1/200 µM, respectively). The growth inhibitory effect of both compounds was inoculum size dependent. The MIC for the FMDP ester was enhanced to the value of 35 µM and that for FMDP to 400 µM, when the growth medium was inoculated with 10⁶ cells ml^-1. The better antifungal potency of the FMDP ester was also confirmed in kinetic studies. As shown in Fig. 2, this compound in concentrations above 20 µM caused, after some delay, the complete growth inhibition of C. albicans cells, while FMDP only reduced the growth rates. Microscopic observations of C. albicans cells treated with FMDP or its ester revealed substantial morphological changes of fungal cells, such as swelling, clumping and inhibition of septum formation. These effects are characteristic for compounds inhibiting chitin biosynthesis and were previously observed in our earlier studies on the mechanism of action of FMDP-peptides (Milewski et al., 1991 ). Growth inhibitory and morphological effects were totally reversed when N-acetyl-D-glucosamine, 5 mM, was added to the FMDP- or FMDP-ester-treated cultures (data not shown). These observations, taken together with the previously established high GlcN-6-P synthase inhibitory potency of both compounds (Zgódka et al., 1999 ), confirm that this enzyme is the only target crucial for growth of C. albicans cells which is inhibited upon the action of FMDP and its ester.

(15K): Fig. 2. Growth inhibitory effect of FMDP and FMDP acetoxymethyl ester. Cells from an overnight culture in YNBG were diluted with YNBG medium to a cell density of 106 cells ml-1. The FMDP or its ester was added to give a final concentration of 0160 µM. All suspensions were incubated at 30 °C with shaking for 8 h. The rate of growth was monitored by determining the OD660 at hourly intervals. (a) FMDP; (b) FMDP acetoxymethyl ester. , 0; •, 5 µM; , 10 µM; , 20 µM; , 40 µM; , 80 µM; x, 160 µM.

Kinetics of uptake of FMDP and the FMDP acetoxymethyl ester
The uptake of low-molecular-mass compounds by microbial cells is usually measured using radioisotopic derivatives. Since such derivatives were not available, we applied a colorimetric method, based on the formation of yellow products upon the reaction of TNBS with the amino group of organic compounds. Since either FMDP or its ester was the only amino-containing compound added to the incubation medium, the decrease in concentration of TNBS-positive substance in filtrates remaining after removal of fungal cells allowed determination of the initial uptake rates of these compounds. The uptake rates were found to be dependent on the concentrations of FMDP or the FMDP ester in the medium. At low concentrations, the initial rates of FMDP uptake were higher than those of the ester [1·49 nmol min^-1 (mg dry wt)^-1 vs 0·56 nmol min^-1 (mg dry wt)^-1, respectively, at 100 µM]. At concentrations higher than 1 mM, the FMDP ester was taken up much faster than its unsubstituted counterpart. The dependence of the initial velocity of FMDP acetoxymethyl ester uptake on the initial concentration of this compound in the medium showed a linear fit; coefficient of determination, r²=0·9976 (Fig. 3). This dependence was maintained up to an FMDP ester concentration of 4 mM. The initial velocity of FMDP uptake exhibited saturation kinetics with a V_max=5 nmol min^-1 (mg dry wt)^-1. This behaviour is characteristic for carrier-mediated transport, while the kinetics of FMDP ester uptake follows the pattern characteristic for passive diffusion.

(12K): Fig. 3. Kinetics of uptake of FMDP and the FMDP acetoxymethyl ester. Exponential-phase C. albicans cells grown in modified Sabouraud medium were harvested, washed and suspended in potassium phosphate buffer, pH 7·0, containing 1% glucose (107 cells ml-1). Cell suspensions were preincubated for 15 min at 30 °C, FMDP or its ester was added to give final concentrations of 0·054 mM, and the incubation was continued. Uptake of compounds was measured at 30 °C over a period of 30 min. Samples of the cell suspensions were withdrawn at time intervals (5 min), cells were removed by filtration and the concentration of tested compounds remaining in filtrates was determined by the TNBS method and then plotted against time. The initial uptake rates were determined from the slopes of the linear part of the uptake curves in the 010 min regions. The initial uptake velocities were then plotted against initial compound concentration to give the presented graphs. , FMDP; •, FMDP ester. Bars represent SD.

Fungal cultures used in uptake kinetics studies contained 10⁷ cells ml^-1. At this cell density, the amount of FMDP acetoxymethyl ester which was accumulated by C. albicans cells increased from a value of 1·6 nmol min^-1 (mg dry wt)^-1 at an initial concentration of 0·2 mM to 10·45 nmol min^-1 (mg dry wt)^-1 at an initial concentration of 2 mM. The decrease in cell density to 10⁶ ml^-1 resulted in a proportional decrease in the amount of FMDP ester accumulated by cells to 0·121 nmol min^-1 (mg dry wt)^-1 at an initial concentration of 0·2 mM and 1·31 nmol min^-1 (mg dry wt)^-1 at initial concentration of 2 mM. These results also indicate that the acetoxymethyl ester of FMDP is transported across biological membranes by diffusion.

The TLC analysis of the spent medium filtrates collected during uptake determination did not reveal any ninhydrin-positive substances, except the FMDP ester, for 25 min. After that time, the appearance of another ninhydrin-positive but UV-negative substance was observed. Its R_F value amounted to 0·03 and was lower than that of the diaminopropanoic acid (R_F=0·37) and FMDP (R_F=0·93), but we were not able to identify this compound. We suppose that this unidentified substance could be a product of FMDP intracellular metabolism, extruded by C. albicans cells.

Effects of metabolic inhibitors on FMDP and FMDP acetoxymethyl ester uptake
The influence of several metabolic inhibitors and ionophores on the uptake of these compounds is shown in Table 1. Metabolic inhibitors such as sodium azide and sodium arsenate reduced the initial velocity of FMDP uptake to 2 and 12% of the control value, respectively, when the cells were preincubated with the appropriate inhibitor for 10 min. These inhibitors and the proton ionophore carbonyl cyanide m-chlorophenylhydrazone did not affect the rate of FMDP acetoxymethyl ester uptake into C. albicans. This indicates that uptake of this compound is neither energy-dependent nor proton-linked. N-Ethylmaleimide, an agent known to covalently modify cysteine side chains, did not reduce FMDP ester uptake.

Table 1. Effects of metabolic inhibitors and ionophores on uptake of FMDP and its ester

Summarizing the results of the uptake studies, we believe that the linear kinetics, dependence of accumulation rate on cell density and lack of effects of metabolic inhibitors, proton ionophores and thiol group blockers on the rate of the FMDP acetoxymethyl ester transport into fungal cells clearly indicate that this compound enters the candidal cells by free diffusion. On the other hand, the saturation kinetics and complete inhibition of uptake by metabolic inhibitors indicate that FMDP is taken up by an active transport mechanism, most likely by an amino acid permease(s). Since these carrier proteins are unlikely to accept amino acid esters (Horak, 1986 ; Prasad, 1987 ), their apparent reluctance towards the acetoxymethyl derivative of FMDP cannot be surprising.

Metabolism of the FMDP ester
FMDP acetoxymethyl ester was added to a crude cell-free extract from C. albicans, and the mixture was incubated at 30 °C. TLC analysis of deproteinized samples taken from this mixture at 1 min intervals revealed the immediate formation of FMDP. Traces of the FMDP ester were detected for 5 min, but totally absent from samples taken later. When proteins present in the cell-free extract were denatured by heating for 3 min at 100 °C prior to the addition of the FMDP ester, decomposition of this compound was not observed. The ester was also stable in 50 mM potassium phosphate buffer (pH 7·0) as well as in the chromatographic solvent system used in TLC analysis. We can therefore conclude that the FMDP acetoxymethyl ester is rapidly hydrolysed to FMDP by unidentified enzymes present in the cytoplasm of C. albicans cells. It is very likely that this process also takes place inside intact fungal cells treated with the agent. Intracellular fast hydrolysis of the FMDP ester must lower the intracellular concentration of this agent to nearly negligible amounts, thus permitting its continuous influx. Moreover, the generated free FMDP is expected to be eliminated from the cytoplasmic environment due to the binding to its target, GlcN-6-P synthase. If the hydrolysis did not occur, the FMDP ester could be accumulated by free diffusion only until its intracellular concentration equilibrates with the extracellular one. According to Ziegelbauer et al. (1998) , the intracellular volume of 10⁷ C. albicans cells is roughly 5 µl. Therefore, such an amount of fungal cells, present in 1 ml of suspension used in our uptake studies, would have been able to accumulate only 0·5% of the antifungal agent present in the medium. Our results clearly show that the accumulation rate is much higher. One may therefore assume that the intracellular hydrolysis of the FMDP ester keeps the concentration gradient of this compound as a driving force for continuous free diffusion.

Our earlier studies on physico-chemical properties of the acetoxymethyl ester of FMDP showed that the apparent lipophilicity of this compound is higher than that of FMDP (Zgódka et al., 1999 ). Results of the studies presented in this paper suggest that the formation of acetoxymethyl esters can be a way of obtaining amino acid analogues lipophilic enough to diffuse through the fungal cell membrane. The application of the lipophilic pro-drug approach for the design of derivatives of GlcN-6-P synthase inhibitors was shown to be a possible way of obtaining novel compounds exhibiting better antifungal properties than the parent compound.

The authors acknowledge the financial support of these studies by the State Committee for Scientific Research (Warsaw), and in part by the Chemical Faculty of the Technical University of Gdask.