Astemizole and an Analogue Promote Fungicidal Activity of Fluconazole

Medical Mycology March 2010, 48, 255–262

Astemizole and an analogue promote fungicidal activity of fl uconazole against Cryptococcus neoformans var. grubii and Cryptococcus gattii

KIEM VU & ANGIE GELLI Department of Pharmacology, School of Medicine, University of California, Genome and Biomedical Sciences Facility, Davis, California, USA Downloaded from https://academic.oup.com/mmy/article/48/2/255/1013134 by guest on 27 September 2021

Cryptococcus neoformans is the leading cause of fungal meningitis, a life-threatening infection that occurs predominately in immuocompromised patients. Current drug therapies are limited to amphotericin B, fl ucytosine and the azoles since the echinocandins have no demonstrated activity against yeast like pathogens. Fluconazole, a drug belonging to the azole class and often the only available antifungal in the developing world, is fungistatic and therefore not effective in clearing cryptococcal infections in immunosuppressed individuals. Here we report that astemizole and a closely related analog (A2) promoted in vitro fungicidal activity of fl uconazole against Cryptococcus neoformans var. grubii and Cryptococcus gattii . Astemizole, a second-generation antihistamine drug used as an

H1 antagonist, has also been found to have antimalarial activity. Disk diffusion assays and MIC and MFC analysis confi rmed that the inhibitory concentrations of these drug combinations were fungicidal. When tested in vivo , astemizole or A2 in combination with fl uconazole signifi cantly improved the survival of Galleria mellonella (wax moth caterpillar) that had been previously challenged with C. neoformans but not when caterpillars were challenged with a fl uconazole-resistant strain. The fi ndings reported here suggest that fungicidal combinations between azoles and other existing drugs may represent an alternative strategy for improving treatments for fungal infections. Keywords fl uconazole, astemizole, Cryptococcus neoformans, fungicidal, drug

Introduction meningitis is a grave infection and even when treated with antifungals it causes high morbidity and mortality in The incidence of opportunistic invasive fungal infections immunocompromised patients and in seemingly normal in immunocompromised individuals continues to be a individuals [5 ]. Lifelong suppressive therapy is usually serious health issue [ 1 – 3 ]. Advancements in the treatment required in AIDS patients suffering from cryptococcosis of malignancies, autoimmune disease and organ transplan- because these infections are commonly not cleared in these tation have resulted in a cohort of patients who are highly individuals [ 6 ]. susceptible to life-threatening fungal infections. Candida Fluconazole is a commonly used drug from the azole species represent the fourth-most common cause of blood- class of antifungals [7 ,8 ]. It is often the drug of choice for stream infections [4 ] and Cryptococcus neoformans var. long-term maintenance therapy in AIDS patients because grubii and Cryptococcus gatti are the leading cause of fun- it can be administered orally, lacks signifi cant side effects gal infections of the nervous system [ 2 ,3 ,5 ]. Cryptoccocal and penetrates the central nervous system [6, 9 ]. Azoles target lanosterol 14α-demethylase – an essential enzyme in Received 12 January 2009; Received in Final version form 28 April 2009; the biosynthesis of ergosterol [ 7 ,8 ]. A signifi cant downside Accepted 1 June 2009 of fl uconazole and other azoles is that they are fungistatic Correspondence: A. Gelli, Genome and Biomedical Sciences Facility, Room 3503, Department of Pharmacology, School of Medicine, University rather than fungicidal and as a result cryptococcal infec- of California, 451 Health Sciences Drive, Davis, CA 95616, USA. tions will persist in the absence of a functional immune Tel: 1 530 754 6446; fax: 1 530 754 7710; E-mail: [email protected] system. The use of fl uconazole for maintenance therapy

© 2010 ISHAM DOI: 10.3109/13693780903081968 256 Vu & Gelli in immunocompromised patients has been associated with cascontent/registry/index.html ) and purchased from Ana- the development of fl uconazole-resistance in cryptococcal logix Inc (Burlington, WI). All drugs were dissolved in infections [9 ]. Given their low toxicity and good infi ltration dimethylsulfoxide to achieve the following concentrations; into the nervous system, azoles would be ideal drugs if they fl uconazole 4 mg/ml, astemizole 4 mg/ml, A1 10 mg/ml, were fungicidal instead. Interestingly, it has been shown A2 10 mg/ml and A3 10 mg/ml. that azoles become fungicidal against Candida spp. when combined with inhibitors of calcineurin [ 10 ,11 ]. Disk diffusion halo assays In this study we found that astemizole and a closely related analog (A2) promoted in vitro fungicidal activity of Strains were inoculated into liquid medium and grown fl uconazole against Cryptococcus neoformans var. grubii overnight at 30°C. Approximately 2 107 cells were inocu- and Cryptococcus gattii. Astemizole is a second-generation lated into 8 ml of top agar pre-warmed at 42°C and subse- Downloaded from https://academic.oup.com/mmy/article/48/2/255/1013134 by guest on 27 September 2021 antihistamine drug that is used as an H1 antagonist and it also quently spread onto YPD plates. All drugs and the solvent possesses antimalarial activity [ 12, 13 ]. The MFCs observed control were applied to 6-mm BBL disks (Becton Dickinson). were similar to the MIC analysis and the results of the These disks were placed on the solidifi ed top agar surface MFCs confi rmed that the inhibitory concentrations of these of the YPD plates as indicated and the strains were grown drug combinations were fungicidal against C. neoformans at 30°C for 48 h. All disk diffusion assays were performed var. grubii and C. gattii. Disk diffusion assays were con- as explained above. Microscope observation at 2.5 sistent with the MIC analysis. The combination of drugs magnifi cation was used to examine whether there was an also signifi cantly improved the survival of the wax moth enhancement of halo clearing that resulted from the com- caterpillar, Galleria mellonella, but not when caterpillars bination of drugs versus drug alone [ 10 ]. where challenged with a fl uconazole-resistant strain of C. neoformans var. grubii . This system was demonstrated to Antifungal drug activity testing by CLSI/NCCLS criteria function as a useful model to examine the in vivo effi cacy of antifungal agents in larvae of G. mellonella previously Checkerboard titrations were performed in order to assess challenged with C. neoformans var. grubii [ 14, 15 ]. Collec- drug interactions according to the Clinical and Laboratory tively, these observations support the notion that fungicidal Standards Institute (CLSI) (previously known as NCCLS – combinations between azoles and other existing drugs National Committee for Clinical Laboratory Standards may lead to improvements in treatment regimes for fungal [ 16 ]. In vitro testing was carried out in RPMI 1640 medium infections. containing L-glutamine, without sodium bicarbonate and buffered to pH 7.0 with MOPS in 96-well plates (96- well cell culture cluster, fl at-bottom, Costar). Inocula of Materials and methods C. neoformans (100 μl) were prepared in accordance with the CLSI standard (M27-A2) [ 16 ], added to the 96-well Strains, media and chemicals plates and incubated for 48 h at 35°C without shaking. The strains used in this study were as follows; Cryptococ cus Readings were taken by measuring the optical density neoformans var. grubii (H99, ATCC 208821, Rockville (OD600 ) with a microplate reader. The MIC of drugs alone MD) , Cryptococcus gattii (NIH444), and strains T1-5796 or in combination was defi ned as the lowest drug con- and 89-6105797 (both of which represent fl uconazole- centration in a well at which 100% reduction in optical resistant C. neoformans var. grubii isolates). The latter density was observed compared to the no-drug control three strains were generously provided by J. Heitman well. Fractional inhibitory concentrations (FICs) and FIC (Duke University). All test isolates were grown on YPD indices ( FIC) were determined as previously described (Yeast extract, Peptone, Dextrose (2% vol/vol). The top [4 ]. The FIC index is defi ned as the sum of the FICs for agar used in the disk diffusion assays was 0.7% Bactor each of the drugs and the FICs are defi ned as the MIC Agar (Difco) in water. Fluconazole and astemizole were of each drug when used in combination divided by the purchased from Sigma-Aldrich (St. Louis, MO). The three MIC of the drug when used alone [ 4 ]. Drug interactions astemizole analogues, i.e., (A1, 4-Piperidinamine, N-[1- were based on FIC indices and classifi ed as synergistic [(4-fluorophenyl)methyl]-1H-benzimidazol-2-yl]-1-[(4- ( FIC 0.5), indifferent ( FIC 1 through 4), or methoxyphenyl)acetyl];, A2, 1H-Benzimidazole-2-amine, antagonistic ( FIC 4) [4 ]. Minimum fungicidal concen- 1-[2-(4-methoxyphenyl)ethyl]-N-[1-[2-(4-methoxyphenyl) trations (MFCs) were determined by plating 100 μl from ethyl]-4-piperidinyl] and A3, 1H-Benzimidazole-2-amine, each well with growth inhibition onto YPD plates that were N-[(4-methoxyphenyl)methyl]-1-[2-(1-piperidinyl)ethyl] then incubated at 30°C for 72 h. The lowest concentration were identifi ed through the Chemical Abstracts Service that yielded three or fewer colonies was recorded as the (CAS) registry website (http://www.cas.org/expertise/ MFC [ 16 ].

Galleria mellonella killing assays San Diego, CA). Estimation of differences in survival (log rank and Wilcoxon tests) was analyzed by the Kaplan- The G. mellonella killing assays were carried out as Meier method using GraphPad Prism software. A P value previously described [2,3 ]. Briefl y, G. mellonella wax of 0.05 was considered signifi cant. Each experiment was moth caterpillars (or larvae) (Vanderhorst, Inc., St. Marys, repeated three times and the data shown here are from a Ohio) in the fi nal instar stage were housed in the dark and representative experiment. used within 7 days from the day of shipment. Fourteen caterpillars of the desired weight (∼300 mg 30 mg) were used for all assays. The inocula were prepared by Results growing a strain ( C. neoformans var. grubii ) overnight in We explored the theory that the combination of astemi- 10 ml of YPD. The culture was centrifuged and washed zole and azoles could potentiate the growth inhibition

with PBS. The optical density (OD ) of the PBS sus- Downloaded from https://academic.oup.com/mmy/article/48/2/255/1013134 by guest on 27 September 2021 600 caused by azoles alone. Astemizole belongs to a group pension was obtained and serial dilutions were made of compounds with antihistamine activity that possess a until a fi nal concentration of 150 cells/μl was obtained. N- (4-piperidinyl)-1 H -benzimidazol-2-amine moiety (cen- Prior to injection, the Hamilton syringe was cleaned by tral part of structure) (Fig. 1A). We identifi ed three closely purging the needle sequentially with 10% bleach, 70% related analogues of astemizole that retained part of this ethanol, and PBS. Special attention was paid during moiety but differed in other portions of the molecule each injection to make sure that no air bubbles were (Fig. 1A). Analogue 1 (A1), which is most similar to astem- introduced into the larvae. Injections of 8 μl aliquots of izole (score 92), contains a 4-methoxyphenyl-acetyl group C. neoformans inoculum (1.2 103 cells) were initiated instead of a 4-methoxyphenyl-ethyl group (a score close in the last left proleg (hemocoel) with a Hamilton syringe to 100 indicated high similarity to astemizole). Analogue (10 μl). Prior to injections the area was cleaned with an 2 (A2) with a score 89 contains two 4-methoxyphenyl- alcohol swap and ampicillin (20 mg of ampicillin/kg of ethyl groups instead of one and analogue 3 (A3), the least body weight) was co-administered to prevent infections similar to astemizole having a score of only 73, lacks the from naturally occurring bacteria. Following injections 4-fl uorophenyl-methyl group and the benzylamide group caterpillars were incubated in petri dishes containing ( Fig. 1 A). sawdust and the number of dead caterpillars was scored Disk diffusion assays were conducted with fl uconazole, daily. Drugs were administered 48 h following the injec- a drug belonging to the azole class of antifungals. A wild tion of the inoculum in the same manner, but a different type strain (H99) of Cryptococcus neoformans var. grubii proleg was used for each injection beginning with the displayed a turbid halo surrounding the disk with the last left proleg, followed by the last right proleg, and the highest concentration of fl uconazole (16 μg), which is penultimate left proleg. The drugs used in the caterpil- indicative of fl uconazole’s fungistatic activity (Fig. 1B). lar assays were fl uconazole, astemizole and A2. Each Astemizole and the A1, A2 and A3 derivatives (32 μg/ animal received a total of three injections: (i) inocula, ml) alone did not result in any detectable growth inhibi- (ii) fl uconazole, and (iii) either astemizole, A2 or 15% tion ( Fig. 2 B, left panel). In striking contrast, the combi- DMSO (dissolved in PBS). This ensured that all caterpil- nation of astemizole (32 μg) and fl uconazole (8 μg and lars received the same number of injections despite the 16 μg) resulted in a signifi cantly larger and clear halo. number of drugs that were tested. For example, the cat- Similarly the combination of fl uconazole and A2 resulted erpillars that were tested with one drug also received one in a large, clear halo whereas the combination of fl ucon- 15% DMSO injection. The control groups of caterpillars azole and A1 or A3 resulted in halos fi lled with residual that were injected with PBS alone or inoculum alone were cells similar to the results found with fl uconazole alone also each given two other injections of 15% DMSO. Fol- ( Fig. 1 B). lowing the injection of drugs, the caterpillars were incu- A closer examination of the halos surrounding A2 bated at 37°C. Most animals that die as a result of trauma and fl uconazole or astemizole and fl uconazole revealed a from injection do so within 24 h. As a result, any animals signifi cant clearing of fungal cells compared to halos that died within the fi rst 24 h of injection were excluded with fl uconazole alone, (Fig. 1C). A clear halo devoid of from data analysis. The number of animals that die within residual cells and microcolonies indicated fungicidal activ- 24 h of injection is about 3% (3 animals out of 100 ity and the results suggested that astemizole and A2 pro- injected). moted this fungicidal activity (Fig. 1C). In contrast, the combination of A1 and fl uconazole or A3 and fl uconazole Statistical analysis resulted in halos fi lled with residual cells similar to those All data were subjected to statistical analysis using resulting from fl uconazole alone suggesting that astemi- GraphPad Prism 4.0 statistical software (GraphPad zole and A2 were the only active compounds to cause

Fig. 1 Astemizole and analogues of astemizole differ in their ability to promote fungicidal activity of fl uconazole against Cryptoccus neoformans . (A) The chemical structures of astemizole and its three most similar analogues, A1, A2 and A3. (B) Disk diffusion assays reveal that a combination of fl uconazole and astemizole and fl uconazole and A2 demonstrate the strongest inhibition of growth. Clear halos are visible surrounding disks containing the combination of drugs even at concentrations of fl uconazole (8 μg) that no longer promote growth Fig. 2 Astemizole and A2 promote fungicidal activity of fl uconazole inhibition when added alone. The inhibition of growth becomes more against Cryptococcus neoformans var. grubii and Cryptococcus gattii . prominent for all drug combinations with increasing concentrations of Disk diffusion assays demonstrate an enhanced inhibition of growth of: fl uconazole (16 μg). (C) Microscopic observation at 2.5 magnifi cation (B) C. neoformans var. grubii, and (C) C. gattii strains in the presence of of disk diffusion assays of C. neoformans exposed to drug combinations. fl uconazole in combination with astemizole or with A2. This effect was In the presence of fl uconazole (16 μg), the addition of astemizole or A2 more prominent in the C. gatti strain. A1 and A3 in combination with demonstrates a signifi cant clearing of halos devoid of residual cells or fl uconazole did not demonstrate signifi cant inhibition of growth when microcolonies compared to fl uconazole alone. A1 and A3 show no compared to fl uconazole alone. Disk diffusion assays were performed as μ μ signifi cant clearing of halos. A wild type strain (H99) was grown in YPD explained previously with 16 g of fl uconazole and 32 g of astemizole, medium overnight subsequently 210 7 cells were resuspended in top agar A1, A2 and A3. and poured onto YPD solid medium. Disks containing the indicated concentrations of fl uconazole and astemizole alone and in combination were added on to the solidifi ed medium and a disk containing DMSO were fungicidal against C. neoformans var. grubii and μ (8 l) was also added. Cells were incubated for 48 h at 30°C. C. gattii ( Table 1 ). For C. gattii the MIC of fl uconazole decreased almost 10-fold when the drug was combined with A2 (Table 1). The calculated FIC index of fl uconazole fungicidal activity of fl uconazole (Fig. 1C). Similar results and A2 was 0.41 indicating a synergistic relationship and were observed with six other strains of C. neoformans similarly between fl uconazole and astemizole ( FIC (data not shown). 0.35). This was consistent with the disk diffusion assays We next examined the effect of astemizole and A2 that revealed large, clear halos of C. gattii surrounding on fl uconazole activity through standard CLSI in vitro disks with fl uconazole and A2 or fl uconazole and astemi- susceptibility testing methods [ 16 ]. The MFCs observed zole ( Fig. 2 C). Similar to the disk diffusion assays for were similar to the MIC analysis and confi rmed that the C. neoformans var. grubii, (Fig. 2B) the combination of inhibitory concentrations of these drug combinations fl uconazole and A1 or fl uconazole and A3 did not result

© 2010 ISHAM, Medical Mycology, 48, 255–262 Fungicidal activity of fl uconazole 259 in clear halos devoid of residual cells suggesting that A1 and A3 did not promote fungicidal activity of fl ucon- A2 azole against C. gattii ( Fig. 2 C). For C. var. grubii the combination of fl uconazole with astemizole or with A2 resulted in a fungicidal synergistic interaction as indicated by the FIC index (0.48 and 0.37, respectively) FIC index Ast Flu

( Table 1 ) [ 4, 10 ]. Our results have so far demonstrated an in vitro synergistic fungicidal activity of fl uconazole with astemizole and A2. In order to evaluate whether a similar synergism was present in vivo, these drugs were tested in the Galleria Downloaded from https://academic.oup.com/mmy/article/48/2/255/1013134 by guest on 27 September 2021

mellonela (wax moth caterpillars or larvae) system. This A2 Flu system has been shown to be a useful model to study the in vivo effi cacy of antifungal agents in larvae that had been previously challenged with C. neoformans [5 , 23 ]. A single combined a dose of each drug was administered alone or in combina- nations that resulted in fungicidal inhibition. MICs and 3 Ast Flu tion 48 h following the inoculation of larvae with 1.210 MIC CFU C. neoformans var. grubii ( Fig. 3 ). We observed a trend, albeit not statistically signifi cant, that suggested that Flu fl uconazole alone (P 0.079) was effective in prolonging

the survival of G. mellonela , consistent with a previous (89-610 5797). study [ 23 ]. However, the survival of the larvae improved signifi cantly when larvae were treated with fl uconazole grubii

and astemizole (P 0.019) or fl uconazole and A2 (P var. 0.002) ( Fig. 3 ). By the endpoint of the study (6.5 days) approximately 19% of caterpillars treated with fl uconazole

astemizole, survived compared to 0% of the controls and neoformans MFC the combination of fl uconazole A2 improved caterpillar uconazole survival by 28%. We further examined whether astemizole could promote the fungicidal activity of fl uconazole against strains with reduced susceptibility to fl uconazole. Halos from disk diffusion assays of two fl uconazole-resistant C. neoformans var. grubii strains (T1-5796 and 89-6105797) were exam-

ined microscopically at 2.5 magnifi cation ( Fig. 4 A). uconazole-resistant strain C. Microscopic analyses revealed that halos resulting from these strains exposed to a combination of fl uconazole (16 μg) and astemizole (32 μg) were not fungicidal since they were fi lled with cells ( Fig. 4 , middle and right panels). In stark contrast, the halo that resulted from the wild type of fl uconazole results in fungicidal activity C. neoformans var. grubii strain exposed to the same drug combination was clear and devoid of any residual cells, suggesting fungicidal activity of fl uconazole MIC alone (Fig. 4, left panel). The in vivo analysis of the fl uconazole- Strain #89-610 corresponds to the fl resistant strains supported the in vitro result. Caterpillars b challenged with a fl uconazole-resistant strain of C. neoformans (89-6105797) did not demonstrate improved 32 64 — 32 64 — 8/32 — 0.75 — survival when treated with a combination of fl uconazole and astemizole (P 0.1383). Furthermore, the calculated FIC index for fl uconazole-resistant strains T1-5796 and Combination of astemizole and A2 with fl Combination of astemizole and

89-6105797 in the presence of fl uconazole and astemizole b was 1.00 and 0.75 respectively, suggesting a lack of synergy Combined MICs, expressed as [Fluconazole]/[Astemizole] or [Fluconazole]/[A2] represent the minimum concentrations of drug combi Combined MICs, expressed Strain Flu Ast A2 Flu Ast A2 gattiigrubiiT1-579689-610 16 16 5.3 64 54 64 128 128 — 16 5.3 16 64 54 64 128 128 — 4.5/4.7 1/16 8/32 2.6/32 2/32 — 0.35 0.48 1.00 0.41 0.37 — Table Table 1 a ( Table 1 ). as mg/l. MFCs are expressed

Fig. 4 Astemizole does not promote the fungicidal activity of fl uconazole against strains of Cryptococcus neoformans with reduced susceptibility Fig. 3 The combination of fl uconazole with astemizole or A2 prolongs to fl uconazole. (A) Microscopic analysis of halos from disk diffusion the survival of Galleria mellonella caterpillars following challenge with assays of C. neoformans var. grubii and two fl uconazole-resistant strains Cryptococcus neoformans . Caterpillars treated with fl uconazole and (T1-5796 and 89-6105797) exposed to drug combinations. Clear halos astemizole (P 0.019) or fl uconazole and A2 (P 0.002) demonstrated devoid of residual cells and microcolonies resulting from C. neoformans signifi cant increases in survival compared to controls. Approximately var. grubii (H99) exposed to fl uconazole and astemizole were observed 19% and 28% of caterpillars treated with fl uconazole + astemizole or (left panel). In contrast the two strains with reduced susceptibility to fl uconazole A2, respectively, survived compared to controls. This fl uconazole resulted in turbid halos fi lled with cells (middle and right combination of drugs was more effective than fl uconazole alone (P panels). Disk diffusion assays were performed as described in Fig. 1. 0.079). Each drug was administered as a single dose to larva previously Plates were incubated at 30°C for 4 days and observed at 2.5 challenged with 1.2 103 CFU of C. neoformans strain H99. Estimation magnifi cation. (B) Caterpillars challenged with a fl uconazole-resistant of differences in survival (log rank and Wilcoxon tests) was analyzed by strain of C. neoformans var. grubii (89-6105797, indicated at Fluc-R the Kaplan-Meier method using GraphPad Prism software. strain in graph) did not demonstrate improved survival when treated with fl uconazole and astemizole (P 0.1383). Each drug was administered as a single dose to larva previously challenged with 1.2 103 CFU of C. Discussion neoformans strain 89-6105797. Estimation of differences in survival (log rank and Wilcoxon tests) was analyzed by the Kaplan-Meier method We sought to identify existing compounds that could poten- using GraphPad Prism software. tially promote fungicidal activity of fl uconazole, which is generally fungistatic against strains of C. neoformans . The combination of astemizole and fl uconazole resulted especially since the effects of astemizole or A2 on fungal in fungicidal activity against C. neoformas var. grubii and cells are not known. The lack of synergy observed with the C. gattii . When tested in vivo , astemizole or A2 in com- combinations of these drugs against strains with reduced bination with fl uconazole signifi cantly improved the sur- susceptibility to fl uconazole does not provide much vival of G. mellonella compared to control. Although the insight into the potential mechanisms. Numerous studies in vivo results presented here are encouraging it is important have reported increased activity of azole drugs when they to note that the in vivo activity of the drug combinations are combined with other antifungals in the treatment of should be confi rmed in murine models of cryptococcosis C. neoformans infections [ 20 – 22 ]. For example, the inhibi- and cryptococcal meningitis. It is likely that these same tory action of fl ucytosine was signifi cantly enhanced by molecules may have a synergic effect with the newer the addition of fl uconazole and the resulting synergy is triazole agents, such as voriconazole, posaconazole and thought to have been caused by an increased uptake of ravuconazole [ 25 ]. fl ucytosine as a consequence of the effects of fl uconazole Because the mechanisms by which astemizole and A2 on the cell membranes [20, 22 ]. Interestingly, in strains enhance the fungicidal activity of fl uconazole were not that demonstrated reduced susceptibility to azoles, the investigated, it is diffi cult to speculate on their activity positive effects caused by the combination of drugs were

© 2010 ISHAM, Medical Mycology, 48, 255–262 Fungicidal activity of fl uconazole 261 signifi cantly reduced, consistent with the fi ndings reported have any clinical applications. This could be easily tested here [ 21 ]. Achieving synergy is a major justifi cation for by electrophysiological techniques that could examine using a combination of drugs to treat a disease. This if A2 blocked the HERK K + channel in which case the is particularly important when dealing with infecting action potential would be altered and heart arrhythmias organisms that no longer respond to current treatments would probably ensue. If however, A2 does not alter because of acquired resistance or for those patients who the function of HERK and is synergistic with fl ucon- fail to respond to standard treatment [ 24 ]. Among the azole in the murine models, then A2 might represent a advantages of combination therapy is the reduction in feasible adjuvant to fl uconazole therapy and perhaps a the amount of drug required for treatment thereby pos- starting point for the development of other similar mol- sibly reducing dose-related toxicities [ 20 – 22 ]. Synergis- ecules that might also promote fungicidal activity of fl u- tic combination therapy might be especially relevant in conazole. Determining whether existing FDA-approved treating cryptococcal meningitis since it occurs predomi- drugs have the ability to promote fungicidal activity of Downloaded from https://academic.oup.com/mmy/article/48/2/255/1013134 by guest on 27 September 2021 nantly in patients that lack an adequate T-cell dependent azoles may represent a reasonable approach for improv- immune response and cure rates with monotherapy remain ing the treatment regimes for fungal infections espe- low [ 23, 24 ]. cially in light of the economic challenges facing de novo Astemizole has recently been shown to have antima- drug development for neglected diseases like fungal larial activity, however its target(s) and the mechanism(s) meningitis. of activity are unknown [12 ]. Astemizole is a second- generation antihistamine drug that is used as an H1 antagonist in several countries but not in the US. It was withdrawn Acknowledgements from the market primarily because of a rare complication We thank E. Mylonakis for technical help with the experi- that led to arrhythmias in few patients [ 17 ]. The cardiac ments using Galleria mellonella . We are especially grate- toxicity is due to the ability of astemizole to block the ful to J. Heitman and his group for hosting and training HERG1 K+ channel and thereby prolong the QT interval K. Vu in G. mellonella techniques. We are also indebted (cardiac repolarization) in the cardiac action potential to E. Blumwald for discussions and critical reading of the [ 18 ]. Cardiac abnormalities have also been associated manuscript. Financial support was provided by a NIHAID with extreme astemizole overdoses and inappropriate grant to A.G. and a NIHAID training grant to K.V. co-medications [17 ]. Although these serious side effects are rare, it is unlikely that astemizole will be an accept- Declaration of interest: able adjunct for combination fl uconazole therapy in the The authors report no confl icts of United States. interest. The authors alone are responsible for the content Three closely related analogues of astemizole, i.e., A1, and writing of the paper. A2 and A3, were also identifi ed. Analogue A2 was the most effective in establishing a synergistic effect with fl u- References conazole. This was surprising considering that A1 is more 1 Martin CA. Invasive fungal infections in the critically ill patient. similar to astemizole than either A2 or A3. However, the J Pharm Prac 2005; 18: 9– 17. fi ndings reported here suggest that the introduction of a 2 Eisenman HC, Casadevall A, McClelland EE. New insights on the sole carbonyl group in A1 prevented it from potentiating pathogenesis of invasive Cryptococcus neoformans infection. Curr the growth inhibition caused by fl uconazole. This highly Infect Dis Reports 2007; 9: 457– 464. polar C O double bond in A1 presumably altered its 3 Chaayakulkeeree M, Perfect JR. Cryptococcosis. Infect Dis Clin North Am 2006; 20: 507– 544. electron density which may have distorted the shape of 4 Johnson MD, MacDougall C, Ostrosky-Zeichner L, Perfect JR, Rex JH. the molecule and prevented A1 from interacting with its Combination antifungal therapy. Antimicrob Agents Chemother 2004; target. A similar phenomenon was demonstrated in small 48: 693– 715. molecules designed to block a Ca2+ -activated K+ chan- 5 Bicanic T, Harrison TS. Cryptococcal meningitis. British Med Bulletin nel [19 ]. Unlike A1, the analogue A2 was effective in 2004; 72: 99– 118. 6 Aller AI, Martin-Mazuelos E, Lozano F, et al. Correlation of fl ucon- promoting synergistic fungicidal activity of fl uconazole. azole MICs with clinical outcome in cryptococcal infection. Antimicrob A2 also showed the greatest effect when administered to Agents Chemother 2000; 44: 1544– 1548. G. mellonella in combination with fl uconazole. The simi- 7 Kauffman CA. Clinical effi cacy of new antifungal agents. Curr Opin larity in the behavior of astemizole and A2 suggests that Microbiol 2006; 9: 483– 488. both molecules likely associate with the same target or 8 Odds FC, Brown AJP, Gow NAR. Antifungal agents: mechanism of action. Trends in Microbiol 2003; 11: 272– 279. pathway in cryptococcal cells. 9 Xu J, Onyewu C, Yoell HJ, et al. Dynamic and heterogenous mutations Whether A2 has a similar human toxicity profi le to to fl uconazole resistance in Cryptococcus neoformans . Antimicrob astemizole is not known since currently A2 does not Agents Chemother 2001; 45: 420– 427.

10 Onyewu C, Blankenship JR, Del Poeta M, Heitman J. Ergosterol bio- 17 Taglialatela M, Timmerman H, Annunziato L. Cardiotoxic potential synthesis inhibitors become fungicidal when combined with calcineurin and CNS effects of fi rst-generation antihistamines. Trends Pharmacol inhibitors against Candida albicans , Candida glabrata and Candida Sci 2000; 21: 52– 56. krusei . Antimicrob Agents Chemother 2003; 47: 956– 964 . 18 Suessbrich H, Waldegger S, Lang F, Busch AE. Blockade of HERG 11 Marchetti O, Entenza JM, Sanglard D, et al. Fluconazole plus channels expressed in Xenopus ooctyes by the histamine recep- cyclosporine: a fungicidal combination effective against experimental tor antagonist terfenadine and astemizole. FEBS Lett 1996; 385: endocarditis due to Candida albicans . Antimicrob Agents Chemother 77– 80. 2000; 44: 2932– 2938. 19 Wulff H , Miller MJ , Hansel W , et al . Design of a potent selective 12 Chong CR, Chen X, Shi L, Liu JO, Sullivan DJ Jr. A clinical drug inhibitor of the intermediate-conductance Ca2+ -activated K + channel, library screen identifi es astemizole as an antimalarial agent. Nat Chem IKCa1 : a potential immunosuppressant . PNAS 9 7 : 8151– 8156 . Biol 2006; 8: 415– 416. 20 Hong Nguyen M, Barchiesi F, McGough DA, Yu VL, Rinaldi MG. 13 Fischer MJE, Paulussen JJC, Horbach DA, et al. Inhibition of In vitro evaluation of combination of fl uconazole and fl ucytosine

mediator release in RBL-2H3 cells by some H1 -antagonist derived against Cryptococcus neoformans var. neoformans . Antimicrob Agents

anti-allergic drugs: relation to lipophilicity and membrane effects. Chemother 1995; 39: 1691– 1695. Downloaded from https://academic.oup.com/mmy/article/48/2/255/1013134 by guest on 27 September 2021 Infl amm Res 1995; 44: 92– 97. 21 Barchiesi F, Gallo D, Caselli F, et al. In vitro interactions of itra- 14 Chamilos GMS, Lionakis ER, Lewis E, Kontoyiannis DP. Role of conazole with fl ucytosine against clinical isolates of Cryptococcus mini-host models in the study of medically important fungi. Lancet neoformans . J Antimicrob Chemother 1999; 44: 65– 70. Infect Dis 2007; 7: 42– 55. 22 Allendoerfer R, Marquis AJ, Rinaldi MG, Graybill JR. Combined 15 Myloakis ER, Moreno JB, Khoury EL, et al. Galleria mellonella as a therapy with fl uconazole and fl ucytosine in murine cryptococcal model system to study Cryptococcus neoformans pathogenesis. Infect meningitis. Antimicrob Agents Chemother 1991; 35: 726– 729. Immun 2005; 73: 3842– 3850. 23 Nucci M, Perfect JR. When primary antifungal therapy fails. Clin 16 Clinical and Laboratory Standards Institute/National Committee for Infect Dis 2008; 46: 1426– 1433. Clinical Laboratory Standards . Reference method for Broth Dilution 24 Cuenca-Estrella M. Combinations of antifungal agents in therapy – Antifungal Susceptibility Testing of Yeasts; Approved Standard- what value are they? J Antimicrob Chemother 2004; 54: 854– 869. Second Edition, M27-A2, 940 West Valley Road, Suite 1400, Wayne, 25 Pasqualotto AC, Denning D. New and emerging treatments for fungal PA 19087-1898 USA , 2004 . infections. J Antimicrob Chemother 2008; 61: ( Suppl. 1 ): i19– 130.

This paper was fi rst published online on Early Online on 01 February 2010.