Abstract
Leupyrrins are highly potent antifungal agents. A structure–activity‐relationship study of natural and synthetic derivatives is reported which reveals important insights into the biological relevance of several structural subunits leading to the discovery of highly potent but drastically simplified leupylogs that incorporate a stable and readily available aromatic side chain. For their synthesis a concise strategy is described that enables a short and versatile access.
Keywords: antifungal agents, Leupylogs, natural products, SAR studies, total synthesis
Novel simplified leupyrrins retaining powerful antifungal potencies were efficiently obtained by a concise sp2–sp3 cross coupling and macro‐lactonization.

The leupyrrins present structurally complex macrodiolides from Sorangium cellulosum. [1] As shown for leupyrrins A1 (1) and B1 (2, Figure 1) they are characterized by an 18‐membered nonsymmetric cyclic core with an unusually substituted γ‐butyrolactone ring together with a pyrrole and an oxazoline in combination with a unique dihydrofuran side chain containing two exocyclic alkylidenes. They exhibit not only very potent antifungal activities but also antiproliferative and anti‐HIV‐properties.[ 1 , 2 ] So far the drug target has not been identified but conventional drug targets are not affected, which may suggest that an unusual biochemical interaction site(s) may be involved. [1] However, the further exploration has been severely hampered by the low availability, in combination with their pronounced structural complexity and a lability of the diene residue in the side chain that is prone to double bond shifts. We have recently isolated natural leupyrrins, determined their full stereochemistry and developed a modular synthetic strategy which culminated in the total synthesis of leupyrrins A1 and B1 (1 and 2, see Figure 1).[ 3 , 4 , 5 ] Herein, we report the first detailed SAR studies with five natural and six synthetic analogues that vary in the side chain, the furan, the oxazoline, the pyrrole moieties and the diacid fragment. Dramatically simplified leupylogs of type 3 and 4 were developed that are characterized by a simplified aromatic side chain and that retain potent biological activities (3).
Figure 1.

Leupyrrin A1 (1) and structure related leupylogs 3 and 4. The structural modifications are highlighted in red.
Despite considerable progress in leupyrrin total synthesis,[ 3 , 4 , 5 ] preparation of the bis‐alkylidene substituted furane moiety has remained a challenge, due to the lability of this subunit and the high number of synthetic steps required, despite the advancement of a concise a one‐pot process. Therefore, a prime focus of this study was placed on simplification and stabilization of this segment. As shown in Figure 1, it was envisioned that the alkyliden units may be stabilized by incorporation into an aromatic moiety. Furthermore, biological activity tolerance for side chain variants of natural leupyrrins (see below) also suggested that the aliphatic residue may also be simplified. In combination, this led to design of analogues 3 and 23 (Scheme 1).
Scheme 1.

Concise three‐step synthesis of simplified Eastern building blocks 12 and 13 and completion of the synthesis of leupylogs 3 and 23.
For their synthesis, the basic strategy developed during leupyrrin total syntheses was applied, involving an sp2–sp3 cross coupling to forge the C6−C7 bond, a cyclodehydration for oxazolin condensation (C13−C14) in combination with an esterification (C4′−C19) and a final macrolactonisation (C1′−C3). For implementation of this approach, Eastern fragments 10 and 11 were required. They were obtained in a short approach from known epoxyalcohols 5 [6] and 6, [7] by coupling with diacid 7 [5] using the Shiina protocol [8] and regio‐ and stereoselective epoxide opening of derived esters 8 and 9 under Brønsted acidic conditions with NH4Cl and NaN3 as a nitrogen source (Scheme 1). [9] Small amounts of likewise obtained C2‐esters could be removed by chromatography. Subsequent azide reduction by hydrogenation in the presence of Pd/C gave amino alcohols 12 and 13 which may be used without further purification. Alternatively, also a Staudinger reduction with polymer‐bound triphenylphosphine may be applied, leading to quantitative yields for 12 (see Supporting Information). For further elaboration, azabenzotriazole 16 was prepared from pyrrole 14 and γ‐lactone 15 following our previously reported procedure,[ 3 , 4 , 5 ] involving a well‐scalable tandem sequence for 15. Subsequent amide formation of 16 with either 12 and 13 under basic conditions followed by cyclodehydration with DAST [10] gave oxazolines 17 and 18 in high yield (90 and 87 %). For conversion to seco‐acids 19 and 20, cleavage of the tert‐butyl esters was realized with TMSOTf/lutidine [11] followed by selective acetate saponification using K2CO3 in a THF/MeOH/H2O mixture. Macrolactonization was then realized by applying the Shiina protocol [8] under high dilution conditions to give macrolactons 21 and 22, and finally TASF mediated deprotection yielded the desired leupylogs 3 and 23.
Encouraged by the potent biological activities of these two analogues (see below), we opted for further modifications and studied the biological role of the pyrrole by replacing it with a furan and a thiophen, giving analogues 4 and 36. As shown in Scheme 2, their synthesis was based on initial cross coupling of bromides 24 and 25 [12] with 15. While the conditions developed for leupyrrin synthesis [5] involving Pd(PtBu3)2, Cs2CO3 in THF/H2O were directly applicable to access 26 in 99 % yield, coupling with thiophene 25 required adjustment of parameters. Finally, optimal results to obtain 27 in quantitative amounts were realized with Pd(PPh3)4 in presence of K3PO4 in THF/water at 60 °C. [13] Subsequently, removal of the tert‐butyl ester with TMSOTf/lutidine, [11] conversion of derived crude acids to active esters 28 and 29 and cyclodehydration [3] with amine 12 gave oxazolines 30 and 31 in high yield. In contrast to previously reported route (Scheme 1), application of conditions for cleavage of tert‐butyl ester (TMSOTf) and acetate esters (K2CO3) were unreliable or led to decomposition of 30 and 31. After considerable attempts, a deprotection sequence incorporating magnesium methoxide [14] for acetate cleavage at 0 °C and zinc bromide [15] in dichloromethane at room temperature for tert‐butyl removal enabled a reliable access to seco‐acids 32 and 33. Subsequent Shiina macrolactonization gave macrocycles 34 (73 %) and 35 (94 %) in a good and excellent yield, respectively. Finally, removal of the remaining primary TBS‐group proceeded smoothly by utilizing TBAF in THF at 0 °C to obtain leupylogs 4 in 96 % and 36 in 61 % yield, which proved to be more reliable as compared to TASF.
Scheme 2.

Synthesis of Western fragments 26 and 27 and completion of the synthesis of leupylogs 4 and 34.
Inspired by a certain remaining activity of 4 in combination with a high biological tolerance of the diester fragment observed for natural leuyprrins, more truncated leupylog 51 (Scheme 3) was targeted, which contains an oxazol instead of the oxazoline and a simplified diacid building block. For its synthesis, a modified Northern building block (43) was required. Its synthesis was initiated by treatment of α‐chloro‐β‐keto‐ester 38 [16] with a mixture of commercially available acid (37) and Hünig's base in acetonitrile [17] to give β‐keto‐ester 39. Subsequent microwave assisted heating in acetic acid in presence of ammonium trifluoroacetate [18] gave the desired oxazole 40 together with imidazole 41, which could be removed by chromatography. After reduction of the ester 40 with DIBAL‐H, resulting alcohol was protected as acetate 42. Selective side chain bromination with NBS then gave furfuryl bromide 43 (80 %) together with small amount of 44 (7 %), derived from double bromination. [19] This side product could be removed at a later aldehyde stage (46, see Supporting Information).
Scheme 3.

Synthesis of Northern building block 43 and completion of the synthesis of leupylog 51.
The next step required a challenging sp2–sp3 cross coupling with boronate 15. Gratifyingly, by applying our initially optimized conditions [5] (see above), the desired product 45 could be isolated in 58 % yield. For introduction of symmetric diacid a one‐pot esterification‐macrolactonization was then evaluated, requiring diol 47. Subsequent saponification studies to access 47 from 45 using K2CO3 in MeOH indicated the primary alcohol to be cleaved first, followed by slower liberation of the secondary hydroxyl group. However, as extended reaction times led to unexpected decompositions, it proved beneficial to keep them very short (6 min) and to stop the reaction before full conversion by addition of saturated NH4Cl. This strategy yielded mainly the desired diol 47 together with variable amounts of easily separable mono‐deprotected 48, which may be submitted to another round of ester cleavage. Gratifyingly, the pivotal one‐pot macrolactonization of diol 47 could then be realized with succinic acid anhydride (49) applying the Shiina protocol, [8] treating a highly diluted solution of anhydride 49, MNBA and DMAP in dichloromethane very slowly over several hours with a solution of diole 47 giving the desired macrodiolide 50 in respectable yield (44 %), considering the synthetic challenge of this tandem process. Finally, removal of the remaining TBS‐group proceeded smoothly with TBAF buffered with AcOH to yield truncated leupylog 51.
Antifungal activities were first assayed for natural leupyrrins 1, 2, 52–55 and triol 56 (Figure 2), which was obtained by saponification of leupyrrin B1 (see Supporting Information).
Figure 2.

Overview of natural leupyrrins 52–55 and triol 56.
As outlined in Table 1, all natural leupyrrins show similarly highly efficacious fungal cell growth inhibitory effects in the low μg mL−1 range towards the yeast Rhodotorula glutinis, and the two molds Botrytis cinerea and Mucor hiemalis (both of which are important plant pathogens) in agreement with the data originally published for 1. [1] In addition, none of them showed activity against the yeast Saccharomyces cerevisiae or the Gram‐positive bacterium Staphylococcus aureus.
Table 1.
Biological activity of natural leupyrrins (1, 2, 52–55) and Triol 56.
|
|
MIC [μg mL−1] |
||||||
|---|---|---|---|---|---|---|---|
|
Compound |
A1 [a] (1) |
A2 (53) |
B1 (2) |
B2 (54) |
C (52) |
D (55) |
Triol (56) |
|
Rhodotorula glutinis |
0.25 |
0.47 |
0.47 |
0.94 |
0.78 |
0.47 |
>40 |
|
Botrytis cinerea |
0.80 |
1.25 |
1.25 |
1.25 |
1.25 |
1.25 |
>40 |
|
Mucor hiemalis |
0.30 |
0.24 |
0.31 |
0.47 |
0.31 |
0.24 |
>40 |
|
Saccharomyces cerevisiae |
n.d. |
>40 |
>40 |
>40 |
>40 |
>40 |
>40 |
|
Staphylococcus aureus |
n.d. |
>40 |
>40 |
>40 |
>40 |
>40 |
>40 |
[a] Value from ref. [1].
These data suggest that slight modifications of the diacid fragment and aliphatic side chain have only a minor effect on antifungal activity. On the other hand, triol 56 showed no activity against all tested bacterial and fungal organisms, revealing the whole macrocycle to be essential.
We then evaluated antifungal activities of leupylogs 3, 4, 23, 36 and 51 in direct comparison to leupyrrins A1 (1) and B1 (2), as well as with Amphotericin B (AmB) as a control using the yeasts Candida albicans, a human pathogen, or Rhodotorula glutinis as test organisms. As shown in Table 2, aryl‐substituted leupylogs 3 and 23 displayed potent IC50 values in the same range as the parent natural products against R. glutinis, demonstrating that substitution of the dihydrofuran side chain by an aryl group is well‐tolerated. In contrast, furan leupylog 4 was less active and the thiophene analogue 36 was essentially inactive, demonstrating that the original pyrrole may not be easily substituted. Notably, the pyrrole N−H of the original heterocycle exerts a characteristic hydrogen bond to the neighboring oxazoline, leading to an N,N‐syn orientation of the two respective nitrogen atoms, which has a pronounced effect on the macrocyclic conformation in solution and solid state. [3] Potentially, a hydrogen bond donor like the pyrrole is also be necessary for the correct conformation of the molecule at the target. The entire exchange of the pyrrole‐oxazoline structure by a furan‐oxazole unit as in leupylog 51 led to complete loss of activity. These data suggest that the pyrrole‐oxazoline subunit and the macrocyclic ring constitute important parts of the pharmacophore.
Table 2.
Biological activity of natural leupyrrin A1 (1) and B1 (2) and synthetic leupylogs 3, 4, 23, 36 and 51 with Amphotericin B (AmB) as a control.
|
|
IC50 [μg mL−1] |
|||||||
|---|---|---|---|---|---|---|---|---|
|
Compound |
1 |
2 |
3 |
23 |
4 |
36 |
51 |
AmB |
|
Rhodotorula glutinis |
0.29 |
0.14 |
0.23 |
0.21 |
10 |
>16 |
>16 |
2.45 |
|
Candida albicans |
>16 |
>16 |
>16 |
>16 |
>16 |
>16 |
>16 |
0.33 |
|
Mouse fibroblasts (L929) |
0.95 |
2.09 |
9.82 |
5.08 |
>40 |
>40 |
5.69 |
n.d. |
Furthermore, none of the compounds tested were active against C. albicans. As 23 was the most active and potent leupylog, we evaluated it further and found it to be highly active against R. minuta and R. mucilaginosa and against Zygosaccharomyces bailii (MIC′s of 1 μg mL−1 and <0.125 μg mL−1, respectively). The growth of Zygosaccharomyces bailii was impaired, but not completely inhibited at concentrations up to 16 μg mL−1. As a control for unspecific toxicity, the yeast Candida glabrata that is resistant against natural leupyrrins, was also resistant to 23 (see Supporting Information). While the insensitive organisms are all capable of fermenting glucose, members of the genus Rhodotorula possess solely a respiratory (non‐fermentative) metabolism. These differences in metabolism might account for the variations in susceptibility observed. In order to assess whether leupylogs have a specific antifungal activity, or also affect the growth of mammalian cells, the natural leupyrrins 1 and 2 and synthetic leupylogs 3, 4, 23, 36 and 51 were tested in an antiproliferative assay with the murine fibroblast cell line L929. Natural leupyrrins 1 and 2 and antifungal leupylogs 3, 23 and 51 also exhibited antiproliferative potencies with similar IC50 values, which indicates that the molecular target is likely to be evolutionary conserved for these simplified analogs. In contrast, leupylogs 4 and 36 were inactive, but also had poor antifungal activity.
In summary, a simplified series of leupyrrin analogues has been developed which are characterized by a drastically simplified aromatic side chain. A first set of relevant SAR data for the leupyrrins has been obtained with natural and synthetic derivatives, focusing on variations of the macrocyclic ring, the pyrrole and oxazoline fragment, the side chain and the diacid building block. They revealed a high importance of the macrocyclic core, a certain flexibility of the diacid fragment and high tolerance of the side chain. Two of these leupylogs retain the potent biological activity of the parent natural products. This demonstrates that the overall structure can be simplified, which is quite remarkable as only few such examples have been described in the context of complex natural macrolides. [20] The most potent leupylogs 3 and 23 are characterized by a substitution of the synthetically challenging and chemically labile bis‐alkylidene diene‐substituted dihydrofuran side chain of the authentic leupyrrins by a mere aryl substituent, which also replace the fragment. These leupylogs are much more readily available as compared to the parent natural product (17 steps for 3 vs. 24 steps for 1 in the longest linear sequences from commercial material), involving a much more readily available Eastern fragment (6 steps for 12 from commercial material vs. 18 steps for the authentic Eastern fragment of leupyrrin A1). Furthermore, these studies suggest that the pyrrole‐oxazoline subunit seems to be a crucial element of their pharmacophore, which may be rationalized by solution conformation considerations. Due to the facilitated practical access, many more leupylogs may now be envisioned, enabling more detailed biological studies of these potent antifungal agents. The increased availability and the potential modifications will aid the drug target identification and will help to evaluate practical applications in basic research and medicine.
Conflict of interest
The authors declare no conflict of interest.
Supporting information
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.
Supplementary
Acknowledgements
Financial support by the DFG (ME 2756/7‐1, TRR261 (B07) and GK 850: scholarship to: S.D.) is gratefully acknowledged. We thank Andreas J. Schneider for excellent HPLC support and Maximilian Guhlke for excellent synthetic support (both University of Bonn). Open access funding enabled and organized by Projekt DEAL.
P. R. Wosniok, C. Knopf, S. Dreisigacker, J. M. Orozco-Rodriguez, B. Hinkelmann, P. P. Mueller, M. Brönstrup, D. Menche, Chem. Eur. J. 2020, 26, 15074.
References
- 1. Bode H. B., Irschik H., Wenzel S. C., Reichenbach H., Müller R., Höfle G., J. Nat. Prod. 2003, 66, 1203–1206. [DOI] [PubMed] [Google Scholar]
- 2. Martinez J. P., Hinkelmann B., Fleta-Soriano E., Steinmetz H., Jansen R., Diez J., Frank R., Sasse F., Meyerhans A., Microb. Cell Fact. 2013, 12, 85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Herkommer D., Thiede S., Wosniok P. R., Dreisigacker S., Tian M., Debnar T., Irschik H., Menche D., J. Am. Chem. Soc. 2015, 137, 4086–4089. [DOI] [PubMed] [Google Scholar]
- 4. Thiede S., Wosniok P. R., Herkommer D., Schulz-Fincke A.-C., Gütschow M., Menche D., Org. Lett. 2016, 18, 3964–3967. [DOI] [PubMed] [Google Scholar]
- 5. Thiede S., Wosniok P. R., Herkommer D., Debnar T., Tian M., Wang T., Schrempp M., Menche D., Chem. Eur. J. 2017, 23, 3300–3320. [DOI] [PubMed] [Google Scholar]
- 6. Gao Y., Klunder J. M., Hanson R. M., Masamune H., Ko S. Y., Sharpless K. B., J. Am. Chem. Soc. 1987, 109, 5765–5780. [Google Scholar]
- 7. Mahaney P. E., Kim C. Y., Coghlan R. D., Cohn S. T., Heffernan G. D., Huselton C. A., Terefenko E. A., Vu A. T., Zhang P., Burroughs K. D., Cosmi S. A., Bray J. A., Johnston G. H., Deecher D. C., Trybulski E. J., Bioorg. Med. Chem. Lett. 2009, 19, 5807–5810. [DOI] [PubMed] [Google Scholar]
- 8.
- 8a. Shiina I., Kubota M., Oshiumi H., Hashizume M., J. Org. Chem. 2004, 69, 1822–1830; [DOI] [PubMed] [Google Scholar]
- 8b. Shiina I., Kubota M., Ibuka R., Tetrahedron Lett. 2002, 43, 7535–7539. [Google Scholar]
- 9. Cherian S. K., Kumar P., Tetrahedron: Asymmetry 2007, 18, 982–987. [Google Scholar]
- 10. Lellouche J.-P., Lafargue P., Guenot P., Heterocycles 1995, 41, 947. [Google Scholar]
- 11.
- 11a. Jones A. B., Villalobos A., Linde R. G., Danishefsky S. J., J. Org. Chem. 1990, 55, 2786; [Google Scholar]
- 11b. Trzeciak A., Bannwarth W., Synthesis 1996, 1433. [DOI] [PubMed] [Google Scholar]
- 12.Bromides 24 and 25 were obtained from 5-methylfuran-2-carboxylic acid and 5-methylthiophene-2-carboxylic acid, respectively by introduction of the tert-butyl-ester followed by bromination with NBS (56 %, 2 steps for 24 and 73 %, 2 steps for 25, see Supporting Information for details).
- 13. Shimizu M., Tomioka Y., Nagao I., Hiyama T., Synlett 2009, 3147–3150. [Google Scholar]
- 14.
- 14a. Xu Y.-C., Bizuneh A., Walker C., J. Org. Chem. 1996, 61, 9086–9089; [Google Scholar]
- 14b. Xu Y.-C., Bizuneh A., Walker C., Tetrahedron Lett. 1996, 37, 455–458. [Google Scholar]
- 15.
- 15a. Wu Y.-q., Limburg D. C., Wilkinson D. E., Vaal M. J., Hamilton G. S., Tetrahedron Lett. 2000, 41, 2847–2849; [Google Scholar]
- 15b. Kaul R., Brouillette Y., Sajjadi Z., Hansford K. A., Lubell W. D., J. Org. Chem. 2004, 69, 6131–6133. [DOI] [PubMed] [Google Scholar]
- 16.
- 16a. Matsuki K., Sobukawa M., Kawai A., Inoue H., Takeda M., Chem. Pharm. Bull. 1993, 41, 643; [Google Scholar]
- 16b. Perrone M. G., Santandrea E., Dell′Uomo N., Giannessi F., Milazzo F. M., Sciarroni A. F., Scilimati A., Tortorella V., Eur. J. Med. Chem. 2005, 40, 143. [DOI] [PubMed] [Google Scholar]
- 17. Altuna-Urquijo M., Gehre A., Stanforth S. P., Tarbit B., Tetrahedron 2009, 65, 975. [Google Scholar]
- 18. Trukhin D. V., Bagryanskaya I. Y., Gatilov Y. V., Mikhalina T. V., Rogozhnikova O. Y., Troitskaya T. I., Tormyshev V. M., Synlett 2005, 2072–2076. [Google Scholar]
- 19. Mayhoub A. S., Talukdar A., Cushman M., J. Org. Chem. 2010, 75, 3507. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.
- 20a. Maier M. E., Org. Biomol. Chem. 2015, 13, 5302–5343; [DOI] [PubMed] [Google Scholar]
- 20b. Halland N., Brönstrup M., Czech J., Czechtizky W., Evers A., Follmann M., Kohlmann M., Schiell M., Kurz M., Schreuder H. A., Kallus C., J. Med. Chem. 2015, 58, 4839–4844; [DOI] [PubMed] [Google Scholar]
- 20c. Scheeff S., Rivière S., Ruiz J., Abdelrahman A., Schulz-Fincke A.-C., Köse M., Tiburcy F., Wieczorek H., Gütschow M., Müller C. E., Menche D., J. Med. Chem. 2020, 63, 1684–1698. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.
Supplementary
