Novel Macrocyclic Amidinoureas: Potent Non-Azole Antifungals Active against Wild-Type and Resistant Candida Species

Abstract

Novel macrocyclic amidinourea derivatives 11, 18, and 25 were synthesized and evaluated as antifungal agents against wild-type and fluconazole resistant Candida species. Macrocyclic compounds 11 and 18 were synthesized through a convergent approach using as a key step a ring-closing metathesis macrocyclization reaction, whereas compounds 25 were obtained by our previously reported synthetic pathway. All the macrocyclic amidinoureas showed antifungal activity toward different Candida species higher or comparable to fluconazole and resulted highly active against fluconazole resistant Candida strains showing in many cases minimum inhibitory concentration values lower than voriconazole.

The epidemiology of fungal infections has been an evolving issue since the late 1960s when, as a consequence of the development of antibiotic therapies, a drastic rise of mycoses was observed. Today, fungal infections represent a major global health threat and the increasing incidence of invasive and opportunistic mycoses is often associated with excessive morbidity and mortality.¹ Fungal infections have increased in incidence in recent decades often as a result of advanced medical treatments and the increase in the number of immunocompromised patients. Patients who are at risk for the development of serious fungal infections include those undergoing blood and solid-organ transplantation, gastrointestinal surgery, patients with AIDS, and premature infants. In addition, cancer chemotherapy and allogeneic bone marrow transplantation are often associated with fungal disease, and up to 30% of patients with acute leukemia experience invasive fungal infections. Given the complexity of the population of patients suffering from a mycosis and the diverse and increasing array of pathogens, fungal infections pose considerable diagnostic and therapeutic challenges. Although several species of fungi are potentially pathogenic in humans, Candida, and in particular Candida albicans, is the organism responsible for most fungal diseases.^2,3 Defense against Candida infections has relied on the use of a limited number of chemotherapeutic agents, including azoles, such as fluconazole and voriconazole, and polyenes, such as amphotericin B.⁴ In particular, azoles have been extensively used to treat a wide range of candidiasis because of their low toxicity and high degree of bioavailability after oral administration.⁵ However, despite they still constitute the preferred first line therapy, the emergence and the spread of drug resistant fungal species^6,7 draw the line at the use of these classical antifungal drugs. Today, fluconazole is poorly or not effective against mutant Candida species making the development of new and more active drugs a priority.⁸ Recently, we reported the discovery and synthesis of a new class of potent antifungal agents, namely, guanidines and macrocyclic amidinoureas,⁹⁻¹³ which proved to be strongly active against Candida species. Macrocycle 1, bearing a crotyl moiety on the alkylguanidine side chain, emerged as the most interesting compound with a biological activity higher than fluconazole against several Candida spp.^14,15 Only few structure–activity relationships (SAR) arose from our previous studies, indicating that the nature of the alkyl group bound to the guanidine moiety plays a key role in the biological activity.

However, no investigations were carried out on the macrocyclic core, both in terms of size as well as regarding the effects that the incorporation of polar/apolar moieties in to the ring might have on the antifungal activity. Hence, with the aim to identify new macrocyclic amidinoureas active both on wild-type as well as mutant fungal species and to further investigate their SAR, we planned the synthesis of two series of amidinourea derivatives A and B, both preserving some structural characteristics of the lead 1 (Figure 1). The first series of derivatives A was planned with the aim to modulate the antifungal activity introducing an ester moiety (polar group) or fusing a benzene (apolar group) with the macrocyclic core, thus potentially favoring the interaction of novel compounds with target enzymes through H-bond or π-interactions. Moreover, the synthesis of 14- and 15-membered ring amidinoureas was also planned in order to further explore the chemical space and evaluate the effects that the macrocycle size might have on the biological activity.

Figure 1.

Structure of hit compound 1.

The synthesis of two additional derivatives B bearing different alkyl substituents on the guanidine moiety was also planned with the aim to evaluate the importance of an unsaturated double bond for the biological activity. Compounds 11a–c and 18a–b were synthesized through a convergent approach using as a key step a ring-closing metathesis (RCM) macrocyclization,¹⁶⁻²⁰ as described in Scheme 1. Aldehyde 2 was first O-alkylated with the appropriate alkenylbromide and in turn condensed with NH₂OH. The resulting oximes were then reduced in the presence of Zn affording the free amines 3a–c, which were guanylated to give the guanidine building blocks 4a–c. These latter compounds were refluxed in THF together with 6 affording amidinoureas 7a–c. Secondary amine 6 was synthesized in a two amidation–reduction sequence starting from Cbz-aminooctanoic acid 5, which was first coupled with allylamine followed by reduction of the resulting amide with diisobutylaluminium hydride (DIBAL-H). Macrocyclization of 7a–c was then carried out using Grubbs’ catalyst second generation at 40–80 °C leading to macrocycles 8a–c in good-high yields,^17,21 as a mixture of E/Z isomers. Hydrogenation of 8a–c allowed the reduction of the double bond and the Cbz cleavage in one single step, affording amines 9a–c. Primary amines were then guanylated with N-crotyl-S-Me-isothiourea-N,N′-diBoc and then treated with trifluoroacetic acid (TFA), leading to desired compounds 11a–c as trifluoroacetic salts. A similar approach was used for the synthesis of derivatives 18a–b bearing an ester moiety on the macrocyclic ring. Beta-alanine 12 was first guanylated in the presence of trimethylsilyl chloride (TMSCl) to protect in situ the carboxylic moiety,²² and the resulting acid was esterified using the appropriate alcohol to afford guanidines 13a–b. Allylamine 6 was then reacted with 13a–b in refluxing tetrahydrofuran (THF) leading to dienes 14a–b in high yields. RCM macrocyclization of these latter compounds was accomplished with Grubbs’ catalyst in DCM at 40 °C, affording a mixture of E/Z alkenes 15a–b. Hydrogenation of 15a–b followed by guanylation and TFA mediated Boc deprotection led finally to desired derivatives 18a–b. In our previous paper, we described the synthesis of lead 1 starting from the commercially available 1,17-diamino-9-azaheptadecane. However, because of the removal of this building block from the market, we were forced to design a new synthetic pathway. Thus, macrocyclic derivatives 25a–b were synthesized as outlined in Scheme 2. The coupling between amine 19 and acid 20 led to the formation of amide 21, which was in turn reduced with DIBAL-H affording the secondary amine 22.²³ Refluxing of 22 in THF led to the formation of the desired macrocycle 23 in good yield. The Cbz protecting group was then cleaved by hydrogenation affording the corresponding primary amine. This latter compound was then guanylated with two different S-Me-alkyl-thioureas 26a–b leading, after Boc cleavage, to the desired derivatives 25a–b. Compounds 11a–c, 18a–b, and 25a–b were then assayed against a total of 116 clinical isolates of seven different wild-type Candida species (oral, vaginal, anorectal, urine, stool, blood, central venous catheter, and respiratory tract specimen with each strain representing a single isolate from a patient). In Table 1, the MIC90 medium values for each species of Candida were reported. Macrocycles 25a and 25b proved to be the most active compounds with MIC values better than fluconazole against most of the Candida strains.

Scheme 1. Synthesis of Derivatives 11a–c and 18a–b.

Reagents and conditions: (i) Br(CH₂)_nCH=CH₂, K₂CO₃, DCM, reflux; (ii) NH₂OH, Py, EtOH, reflux; (iii) Zn, HCl, THF, reflux; (iv) (BocNH)₂C=NTf, Et₃N, DCM; (v) AllylNH₂, EDC, HOBt, DIPEA, DMF; (vi) DIBAL-H, DCM, r.t.; (vii) THF, reflux, 12 h; (viii) Grubbs’ Cat. second gen., toluene or DCM 2–10 mM, 40–80 °C; (ix) H₂, Pd/C, EtOH; (x) CrotylNBoc(C=NBoc)SMe, THF, reflux, 12 h; (xi) TMSCl, Et₃N, DCM, reflux; (xii) DMAP, DCC, HO(CH₂)_nCH=CHCH₃, DCM, rt 24 h; (xiii) 6, THF, reflux, 12 h; (xiv) Grubbs’ Cat. second gen., DCM 2 mM, 40 °C; (xv) H₂, Pd/C, EtOH; (xvi) CrotylNBoc(C=NBoc)SMe, THF, reflux, 12 h.

Scheme 2. Synthesis of Derivatives 25a–b.

Reagents and conditions: (i) EDC, HOBt, DIPEA, DMF; (ii) DIBAL-H, DCM, r.t.; (iii) THF, reflux, 12 h; (iv) H₂, Pd/C, EtOH; (v) 26a–b, THF, reflux, 12 h.

Table 1. Antifungal Activity (MIC90, μg/mL)^a.

species (No. strains tested)	Fb	11a	11b	11c	18a	18b	25a	25b	1c
C. albicans (22)	2	16	4	4	32	16	1	1	2.5
C. guillermondii (10)	4	16	2	2	32	8	2	2	n.d.
C. krusei (13)	256	32	8	4	64	32	4	0.5	5
C. parapsilosis (24)	0.5	8	2	2	32	8	2	4	5
C. tropicalis (11)	2	8	2	1	32	16	0.5	0.5	1.25
C. kefyr (10)	1	4	4	2	32	16	4	4	n.d.
C. glabrata (26)	16	64	8	16	64	32	16	32	20

^aMIC90 values were determined at 24 h both visually and spectrophotometrically.

^bFluconazole.

^cActivity values expressed as MIC50.

In particular, 25b showed to be strongly active against C. krusei, a species resistant to common azole drugs, and against C. tropicalis. Compounds 11 and 18 retain antifungal activity as well, in many cases with MIC values better than fluconazole. In general, compounds 11a–c possessing an aromatic substitution on the macrocyclic core proved to be more active when compared to 18a–b bearing an ester moiety. The increase of the ring size in compounds 11 from 13 to 15 carbon atoms evidenced an enhancement of activity toward all Candida species. However, it is noteworthy that the 14-membered ring 11b showed the highest activity toward C. albicans, C. guillermondii, and C. parapsilosis, while the larger macrocycle 11c proved to be more active against C. tropicalis and C. kefyr. In all cases, 11b–c resulted more than or as active as fluconazole, while the 13-membered ring 11a proved to be less active. However, compounds 11 bearing a benzene fused with the macrocyclic core resulted less active than the unsubstituted analogue 25. In general, the introduction of an apolar moiety leads to the decrease of the antifungal activity, with the exception of compound 11b, which showed a stronger activity against C. glabrata, a species resistant to common azoles, with a minimum inhibitory concentration (MIC) = 8 μg/mL. Compounds 18a–b present a lower activity than 25 and 11 toward all Candida species. Again the larger macrocycle 18b resulted more active than 18a. As a general rule, the introduction of a substituent on the macrocyclic ring results in a slight decrease of antifungal activity. However, the presence of a fused aromatic ring seems to be more beneficial than the introduction of an ester moiety while larger rings lead to an increase of biological activity. Azole drugs are known to act as antifungals inhibiting the fungal enzyme 14α-demethylase, which produces ergosterol, an important component of the fungal plasma membrane. At the molecular level, different mechanisms contribute to resistance against azole agents.²⁴ The principal mechanism regards the alterations of 14-α-demethylase due to the overexpression and mutations of the gene ERG11 coding for the enzyme. As a consequence, a higher intracellular azole concentration is needed to complex the increased number of demethylase enzymes present in the fungal cells.²⁵ In addition, it has been demonstrated that mutations in ERG11 prevents binding of azoles to the enzymatic site. A second major mechanism leading to azole resistance is the overexpression of plasma membrane efflux pumps. This mechanism is mediated by two types of transporters, the major facilitators (encoded by MDR genes in C. albicans) and the ABC transporters (encoded by CDR genes in C. albicans).²⁶ Upregulation of the CDR genes confers resistance to multiple azoles in C. albicans, whereas upregulation of MDR1 alone leads to fluconazole resistance exclusively. Compounds 11, 18, and 25 were thus assayed against mutant C. albicans and C. glabrata strains, and voriconazole, an azole active on fluconazole resistant strains, was chosen as the reference compound. All the tested derivatives proved to be active against C. albicans and C. glabrata fluconazole-resistant strains (Table 2). Among them macrocycles 25a–b showed better activity values against resistant C. albicans strains than voriconazole (MIC = 1–2 μg/mL) with the exception of DSY348 and DSY751 strains. Compound 11b resulted more active than voriconazole itself against mutant strains DSY289, DSY296, DSY284, and DSY775 as well as 11c against DSY289. Derivatives 18 were more active than fluconazole but in general less active than 11 and voriconazole, except for DSY289 strain. Analogues 11a, 11c, and 18a–b resulted inactive against C. glabrata resistant strains. However, the 14-membered macrocycle 11b, as well as compounds 25a–b, showed a good activity proving to be more active than fluconazole against all C. glabrata strains and showing a MIC value against DSY756 close to voriconazole. In addition all these compounds are fully active against Candida isolates reported as resistant to amphotericin B (Table 3).²⁷

Table 2. Antifungal Activity on C. albicans and C. glabrata Fluconazole Resistant Strains.

			MIC90 (μg/mL)a
strains	species	resistance mechanism	Fb	VORc	11a	11b	11c	18a	18b	25a	25b
DSY284	C. albicans	mutation in the ERG11 gene; increased expression of the CDR1 and CDR2 genes	256	4	16	2	9	16	8	1	1
DSY296	C. albicans	mutation in the ERG11 gene; increased expression of the CDR1 and CDR2 genes	128	8	16	4	10	16	16	2	2
DSY289	C. albicans	mutation in the ERG11 gene; increased expression of the CDR1 and CDR2 genes	256	8	8	2	5	8	8	1	1
DSY348	C. albicans	mutation in the ERG11 gene; increased expression of the CDR1 and CDR2 genes	32	0.25	8	2	5	8	8	2	1
DSY292	C. albicans	mutation in the ERG11 gene; increased expression of the MDR1	64	2	16	4	10	16	16	2	2
DSY735	C. albicans	increased expression of the CDR1 and CDR2 genes	64	2	8	2	5	8	16	1	1
DSY775	C. albicans	mutation in the ERG11 gene; increased expression of the CDR1 and CDR2 genes	128	8	16	4	10	16	16	2	2
DSY751	C. albicans	increased expression of the MDR1 gene; mutation in the ERG11 gene	256	0.25	16	4	10	16	16	2
DSY530	C. glabrata	increased expression of the CgCDR1 genes	64	0.5	64	8	36	64	32	16	16
DSY754	C. glabrata	increased expression of the CgCDR1 genes	64	2	64	16	40	64	32	16	32
DSY756	C. glabrata	increased expression of the CgCDR1, CgCDR2, and CgSNQ2 genes	128	4	64	8	36	128	64	16	16
DSY2254	C. glabrata	increased expression of the CgCDR1 and CgCDR2 genes	128	2	64	16	40	128	64	32	32
DSY2271	C. glabrata	increased expression of the CgCDR2 genes	64	0.125	64	8	36	64	32	16	32

^aMIC values were determined at 24 h both visually and spectrophotometrically.

^bFluconazole.

^cVoriconazole.

Table 3. Antifungal Activity on Amphotericin Susceptible and Resistant Candida Strains.

		MIC90 (μg/mL)a
strains	species	AmBb	11a	11b	11c	18a	18b	25a	25b
ATCC 200955	C. albicans	2	8	1	5	8	4	1	1
ATCC 200950	C. lusitaniae	1	8	2	5	8	8	4	2
ATCC 200951	C. lusitaniae	1	8	2	5	4	4	2	2
ATCC 200952	C. lusitaniae	2	8	2	5	4	4	1	2
ATCC 200953	C. lusitaniae	1	16	4	5	8	8	2	4
ATCC 24348	C. lusitaniae	0.125	8	2	5	4	8	2	1
ATCC 60247	C. lusitaniae	0.125	16	4	5	4	4	4	2
ATCC 200956	C. tropicalis	8	16	4	10	16	16	2	1

^aMIC values were determined at 24 h both visually and spectrophotometrically.

^bAmphotericin B.

In conclusion, novel macrocyclic amidinoureas 11a–c, 18a–b, and 25a–b were synthesized and biologically assayed as antifungal agents. Derivatives 11 and 18 were obtained through an innovative convergent synthetic pathway whose key step was represented by a RCM macrocyclization. All the new macrocycles showed potent antifungal activity against different wild-type Candida species and fluconazole- or Amphotericin B-resistant Candida strains. In particular, 25a–b and 11b resulted in the most interesting potential antifungal agents. Macrocyclic amidinoureas proved to be excellent non-azole lead compounds able to act both on classic as well as resistant fungal infections. Because of their innovative structure, it is reasonable to hypothesize for macrocyclic amidinoureas a mechanism of action different from that of classical azole and polyenic drugs. Further studies to disclose the mechanism of action are currently in progress in our laboratories.

Glossary

Abbreviations

DCM

dichloromethane

EDC

1-ethyl-3-(3-dimethylaminopropyl)carbodiimide

HOBt

hydroxybenzotriazole

DIPEA

N,N-diisopropylethylamine

DIBAL-H

diisobutylaluminium hydride

TMS

trimethylsilyl chloride

DMAP

4-dimethylaminopyridine

TFA

trifluoroacetic acid

DCC

dicyclohexylcarbodiimide

Cbz

benzyl carbamate

Boc

tert-butoxy carbamate

MIC

minimum inhibitory concentration

MDR

multidrug resistance

CDR

Candida drug resistance

RCM

ring closing metathesis

gen

generation

Supporting Information Available

Synthetic methods and characterization of compounds. This material is available free of charge via the Internet at http://pubs.acs.org.

The authors declare no competing financial interest.