Madurastatins with Imidazolidinone Rings: Natural Products or Side-Reaction Products from Extraction Solvents?

Madurastatins are a group of pentapeptides containing an oxazoline moiety, and, in a few cases, an imidazolidinone ring as an additional structural feature. In our search for new potential antiparasitic metabolites from natural sources, we studied the acetone extracts from a culture of Actinomadura sp. CA-135719. The LC/HRMS analysis of this extract identified the presence of the known madurastatins C1 (1), D1 (4), and D2 (5) together with additional members of the family that were identified as the new madurastatins H2 (2) and 33-epi-D1 (3) after isolation and spectroscopic analysis. The planar structures of the new compounds were established by HRMS, ESI-qTOF-MS/MS, and 1D and 2D NMR data, and their absolute configuration was proposed using Marfey’s and bioinformatic analyses of the biosynthetic gene cluster (BGC). A revision of the absolute configuration of madurastatins D1 and D2 is proposed. Additionally, madurastatins containing imidazolidinone rings are proved to be artifacts originating during acetone extraction of the bacterial cultures.

due to limitations including suboptimal efficacy, drug resistance, and adverse effects, poses serious threats to human health, and new therapeutic solutions are required [9,10].Natural products are unique sources of chemical diversity and have historically constituted rich sources of new bioactive compounds, with many drugs originated from natural products being used today in clinical practice.Thus, microbial natural products remain a unique and untapped source for the identification of novel bioactive compounds [11].As part of an ongoing partnership between Fundación MEDINA, Institut Pasteur Korea (IPK), and Drugs for Neglected Diseases initiative (DNDi) to discover natural products active against T. cruzi and L. donovani, a subset of the MEDINA natural products extracts Leishmaniasis and American trypanosomiasis (Chagas disease) are two neglected tropical diseases (NTDs), caused by the parasites Leishmania spp.and Trypanosoma cruzi, respectively, that lead to thousands of deaths worldwide every year.These diseases are also emerging as a health problem in developed countries.The lack of field-adapted treatments, due to limitations including suboptimal efficacy, drug resistance, and adverse effects, poses serious threats to human health, and new therapeutic solutions are required [9,10].Natural products are unique sources of chemical diversity and have historically constituted rich sources of new bioactive compounds, with many drugs originated from natural products being used today in clinical practice.Thus, microbial natural products remain a unique and untapped source for the identification of novel bioactive compounds [11].
As part of an ongoing partnership between Fundación MEDINA, Institut Pasteur Korea (IPK), and Drugs for Neglected Diseases initiative (DNDi) to discover natural products active against T. cruzi and L. donovani, a subset of the MEDINA natural products extracts libraries was tested against both parasites in the IPK phenotypic screening platform.LC-HRMS analysis of the active acetone extract from a culture of Actinomadura sp.CA-135719 in FPY-6 medium revealed, together with madurastatin C1, the presence of other compounds with molecular formulae corresponding to potentially new madurastatins that prompted us to a more detailed investigation of their structures.

LC-UV-HRMS Analysis of the Extract from Actinomadura sp. CA-135719 and Identification of Components
The acetone extract from a culture in FPY-6 medium of Actinomadura sp.CA-135719 displayed the LC-UV chromatographic profile at 210 nm shown in Figure 2. The major peaks (1 and 5) together with the minor metabolites (2)(3)(4) were isolated, and their structures were established by HRMS and 1D and 2D NMR analysis.The major components of the extracts were identified as madurastatins C1 (1) and D2 (5) libraries was tested against both parasites in the IPK phenotypic screening platform.LC-HRMS analysis of the active acetone extract from a culture of Actinomadura sp.CA-135719 in FPY-6 medium revealed, together with madurastatin C1, the presence of other compounds with molecular formulae corresponding to potentially new madurastatins that prompted us to a more detailed investigation of their structures.

LC-UV-HRMS Analysis of the Extract from Actinomadura sp. CA-135719 and Identification of Components
The acetone extract from a culture in FPY-6 medium of Actinomadura sp.CA-135719 displayed the LC-UV chromatographic profile at 210 nm shown in Figure 2. The major peaks (1 and 5) together with the minor metabolites (2-4) were isolated, and their structures were established by HRMS and 1D and 2D NMR analysis.The major components of the extracts were identified as madurastatins C1 (1) and D2 (5), whereas molecular formulae of C29H43N7O10 (2) and C28H39N7O9 (3 and 4) were assigned to the minor components of the extract.
carbon at δC 80.2 (Figure 4), indicating the connectivity and the presence of the imidazolidinone ring.Additionally, the fragment at m/z 171.1129 also confirms this assumption (Figure 5).Compound 2 was named madurastatin H2 based on its structural similarity with madurastatin H1.The molecular formula of C 28 H 39 N 7 O 9 assigned to compounds 3 and 4 by HRMS analysis was previously found in madurastatin D1 [4].Analysis of the 1D and 2D NMR data revealed both compounds to have the same planar structure as madurastatin D1, although noticeable differences in the 1 H and 13 C chemical shifts of C-22, C-23, C-26, C-32, C-33, and C-34 between both compounds were observed (Table 2).The NMR data of compound 4 are identical to those published for madurastatin D1 (Figure 3).The presence of a methyl group at δ H 1.12 and δ C 15.1 and the downfield chemical shift of H-33 to δ H 4.42 in compound 3 suggest a different configuration at C-33, which leads to the aforementioned differences in chemical shifts at the surrounding positions.Thus, compound 3 was confirmed to be the 33-epimer of compound 4 (Figure 3).
Once the planar structure of the compounds was established, Marfey's analysis was used to determine the absolute configuration of the constituent amino acids [12].Acid hydrolysis of 1-4 was followed by LC-MS analysis of the hydrolysates after derivatization with N-(2,4-dinitro-5-fluorophenyl)-D/L-valinamide (D/L-FDVA, Marfey's reagent) and comparison with the retention times and mass spectra obtained for standards.Since Nα-methyl-L-Orn was not available, its absolute configuration determination was based on the D/L elution order described in bibliography [6].Whereas the elution order of the doubly derivatized Orn-L-FDVA was determined to be D→L, this elution order was inverted in the case of Nα-methyl-Orn-L-FDVA, being L→D [6].Partial racemization was observed in all double ornithine adducts of the compound.Marfey's analyses confirmed the presence of D-Ser, Nδ-hydroxy-L-Orn, and Nα-methyl-Nδ-hydroxy-L-Orn in compounds 1-4 (Figures S1-S18).Thus, the absolute configuration of madurastatins was established as 9R, 23S, 26S.To our knowledge, all madurastatins described contain D-serine, Nα-methyl-Nδhydroxy-L-Orn, and Nδ-hydroxy-L-Orn as constituent amino acids except those reported by Yan and coworkers, who reported the isolation of ent-madurastatin C1, mainly based on a negative value of its specific rotation [4].The Marfey's analysis carried out by these authors [4] to confirm their stereochemical proposal was limited to comparing the retention times of D-and L-Ser standards derivatized with L-FDLA, where the difference found in retention time between the L-and D-Ser Mosher derivatives was less than 0.1 min.Poor chromatographic resolution of the D/L-Ser pair derivatized with L-FDLA using formic acid-based solvent systems was previously reported [13].A confirmation of the absolute configuration of other constituent amino acids of the molecule, together with the development of an analytical method with a better chromatographic resolution of the D/L-Ser Mosher derivatives pair, to unequivocally establish the absolute configuration of the Ser residue and hence that of the claimed ent-madurastatin C1 would have been desired.The absolute configuration of madurastatins D1 and D2, isolated in the same work, was assumed to be the same as that of ent-madurastatin C1.
As mentioned above, compounds 3 and 4 differ in their absolute configuration at C-33.Compound 4 was demonstrated to have the same NMR spectroscopic data as madurastatin D1 and hence the same relative configuration.However, based on a common biosynthetic origin with madurastatins C1 and H2 isolated from our CA-135719 strain, we propose that 4 has an opposite absolute configuration in all its chiral centers (9R, 23S, 26S, 33S) to that reported previously for madurastatin D1.The positive specific rotation measured for both compounds confirms that the absolute configuration initially proposed for madurastatin D1 is most probably wrong [4].

In Silico Analysis of the Madurastatin Biosynthetic Gene Cluster
To confirm the difference in the absolute configuration between 4 and that previously reported for madurastatin D1, the genome of the strain CA-135719 was sequenced and assembled using a combination of PacBio and Illumina NovaSeq.A complete circular genome of ~9.5 Mb was obtained, and four identical rRNA 16S gene sequences showing a 98.80% similarity with Actinomadura darangshiensis DSLS-70T were identified.AntiSMASH analysis predicted 22 putative BGCs, including a 75 Kb cluster similar to the clusters encoding madurastatins C1, D1, and D2 (mad) [4] and madurastatins A1, A2, E1, F, and G1 (rene) [5] (Table S1).The cluster contained 61 genes (mds1-mds61), including four NRPS (mds12, mds21, mds24, and mds37), an L-ornithine-N-(5)-monooxygenase (mds39) involved in Nδ-hydroxy-ornithine biosynthesis, an aspartate 1-decarboxylase (mds23) involved in β-Ala biosynthesis, and a salicylate synthase (mds36) responsible of salicylic acid biosynthesis (Table S1, Figure 6a).It has been proposed that the biosynthesis of madurastatins begins with the loading of the salicylate moiety to the thiolation domain of Mds21 by Mds24 (Figure 6b) [5].The heterocyclization domain (C1) of Mds21 generates a phenyloxazoline ring from the condensation of salicylate and D-Ser (Figure 6b).Since no epimerases are present in the cluster, the conversion of L-Ser to D-Ser remains uncertain [5].Then, Mds37 incorporates glycine, β-alanine, and Nδ-hydroxy-acetyl-L-ornithine, according to the predictions of the first three NRPS modules (Figure 6b).The last module of Mds37, as in the case of ReneL [5], lacks the adenylation (A) and thioesterase (TE) domains that would incorporate Nδ-hydroxy-L-ornithine and release the NRP, respectively.While Yan et al. [4] described the presence of a D C L domain and an ornithine-specific A domain in the last module of the homologous NRPS Mad30, a reanalysis of the mad cluster shows that the last module shares the same organization as ReneL and Mds37 (presence of a L C L domain and absence of an A domain) (Figure S19), so no D-amino acids may be incorporated in that position.Moreover, except for the D-Ser from the phenyloxazoline ring, all the incorporated amino acids should be L-configured since no epimerization (E) or dual E/C domains are present in the NRPS (Figure 6b and Figure S19).A trans-acting A domain (Mds12 or Mds48) may activate the Nδ-hydroxy-L-ornithine incorporated in the last module of Mds37, similar to the proposed biosynthesis of madurastatin by the rene cluster [5].This should also occur in the mad cluster [4].The chain release might be catalyzed by the esterase Mds41, although a spontaneous nonenzymatic hydrolysis cannot be excluded [5].
Int. J. Mol.Sci.2024, 25, x FOR PEER REVIEW 8 of 17 be catalyzed by the esterase Mds41, although a spontaneous nonenzymatic hydrolysis cannot be excluded [5].
The biosynthesis of madurastatin C1 and the rest of the compounds isolated from CA-135719 has been clearly established, and BGC analysis confirmed the absolute configuration determined by Marfey's analysis.Similarities found between the mds and the mad and rene clusters allow us to conclude that the absolute configuration previously proposed for madurastatins D1 and D2 [4] is probably wrong and that the existence of ent-madurastatin C is questionable and might be explained due either to an error when measuring the specific rotation of the compound or to the specific rotation having been measured with a metal-chelated species of the molecule.
Finally, due to the absence in the mds cluster of genes related to the biosynthesis of the 4-imidazolidinone ring present in 2-5 and the fact that two different madurastatins D1 having opposite configuration at C-33 were isolated, we postulate that madurastatins 2-5 might be a consequence of the reaction of the natural madurastatins with the acetone used in the extraction.The biosynthesis of madurastatin C1 and the rest of the compounds isolated from CA-135719 has been clearly established, and BGC analysis confirmed the absolute configuration determined by Marfey's analysis.Similarities found between the mds and the mad and rene clusters allow us to conclude that the absolute configuration previously proposed for madurastatins D1 and D2 [4] is probably wrong and that the existence of ent-madurastatin C is questionable and might be explained due either to an error when measuring the specific rotation of the compound or to the specific rotation having been measured with a metal-chelated species of the molecule.
Finally, due to the absence in the mds cluster of genes related to the biosynthesis of the 4-imidazolidinone ring present in 2-5 and the fact that two different madurastatins D1 having opposite configuration at C-33 were isolated, we postulate that madurastatins 2-5 might be a consequence of the reaction of the natural madurastatins with the acetone used in the extraction.

Effect of the Solvent Used in the Extraction of Madurastatins
To verify this hypothesis, small-scale fermentations of strain CA-135719 in FPY-6 medium for 7 days were extracted with an equal volume of five solvents, acetonitrile, n-butanol, ethyl acetate, methanol, and acetone, in triplicate.The extracts were analyzed by LC-HRMS, and ion extraction of each madurastatin was performed (Table 3).Whereas madurastatins C1 and H1 were detected in all extracts as major and minor compounds, respectively, madurastatins H2, 33-epi-D1, D1, and D2 were only detected at very low levels in acetone extracts.Consequently, the formation of madurastatins H2 and D2 can be explained as originating from the reaction of madurastatins H1 and C1, respectively, with acetone (Figure S21).Similarly, the isolation of the two madurastatin D1 epimers (3 and 4) might be a result of a nonenzymatic addition of acetaldehyde to madurastatin C1 (Figure S22).650.314 ± 0.005 1.9 ± 0.5 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.1 ± 0.0
Although this side reaction extensively occurs on the N-termini of peptides during dimethyl labeling using formaldehyde and significantly reduces the quality of proteomic analysis, it is frequently overlooked because of the limitations of current database search engines on the identification of unknown modifications [14].Other byproducts containing the imidazolidinone ring were observed in the synthesis of the precursors of discarines C and D and myrianthine A [15] and in nummularin G [16], sativanine B [17], and cambodines A, B, D, E, and F [18].Additionally, several vancomycin antibiotics also undergo spontaneous chemical modifications when kept at room temperature at physiological pH in aqueous solutions containing traces of formaldehyde or acetaldehyde.By using tandem mass spectrometry, the modification was unambiguously identified as a 4-imidazolidinone moiety at the N-terminus [19].The formation of 4-imidazolidinone moieties as formaldehyde adducts is not unprecedented and is indicated in several studies [20][21][22][23][24][25].

Obtainment of Madurastatin D2 from Madurastatin C1 Using Acetone
To confirm that madurastatin D2 can be originated from madurastatin C1 using acetone as solvent, the latter (Figure 7a) was incubated in acetone.As it can be observed after the incubation with acetone, the LC-MS profile revealed not only the presence of madurastatin C1 but also the presence of madurastatin D2 (Figure 7b).Thus, acetone can be incorporated into madurastatin C1 with the subsequent formation of the imidazolidinone ring to give madurastatin D2 (Figure S21).The extraction of the CA-135719 cultures with different solvents (Table 3) as well as the incubation of madurastatin C1 with acetone demonstrate that madurastatin D2 is an artifact produced by acetone.On the other hand, the origin of acetaldehyde to form compounds 3 and 4 is uncertain.Nonetheless, the absence of these compounds in extracts obtained in other solvents than acetone, together with the absence of genes in the mds cluster that explain the biosynthesis of the 4-imidazolidinone ring, together with the isolation of both epimers at C-33 of madurastatin D1 indicate that these compounds might not be the product of an enzymatic reaction.

Antiparasitic Activity
All compounds isolated were tested against Trypanosoma cruzi and Leishmania donovani, using the liquid microdilution method.The IC50 values against both parasites are indicated in Table 4. Madurastatins were not active against L. donovani at the highest concentration tested, showing IC50 values higher than 50 µM.Against Trypanosoma cruzi, madurastatin activities ranged from moderate to low, although the compounds also displayed some cytotoxicity.Madurastatin D2 displayed the lowest IC50 (8.93 µM) for the inhibition normalized by parasite number.and positive (Benz.treated-infected) control that measures the number of host cells infected with parasite in their cytoplasmic region.Inhibition norm.para.num.: Antiparasitic parameter normalized by negative (DMSO treated-infected) and positive (Benz.treated-infected) control that measures the number of parasites in the cytoplasmic region of host cells.Cell ratio (%): cytotoxicity parameter that count the number of host cells.The extraction of the CA-135719 cultures with different solvents (Table 3) as well as the incubation of madurastatin C1 with acetone demonstrate that madurastatin D2 is an artifact produced by acetone.On the other hand, the origin of acetaldehyde to form compounds 3 and 4 is uncertain.Nonetheless, the absence of these compounds in extracts obtained in other solvents than acetone, together with the absence of genes in the mds cluster that explain the biosynthesis of the 4-imidazolidinone ring, together with the isolation of both epimers at C-33 of madurastatin D1 indicate that these compounds might not be the product of an enzymatic reaction.

Antiparasitic Activity
All compounds isolated were tested against Trypanosoma cruzi and Leishmania donovani, using the liquid microdilution method.The IC 50 values against both parasites are indicated in Table 4. Madurastatins were not active against L. donovani at the highest concentration tested, showing IC 50 values higher than 50 µM.Against Trypanosoma cruzi, madurastatin activities ranged from moderate to low, although the compounds also displayed some cytotoxicity.Madurastatin D2 displayed the lowest IC 50 (8.93µM) for the inhibition normalized by parasite number.Inhibition norm.inf.ratio: Antiparasitic parameter normalized by negative (DMSO treated-infected) and positive (Benz.treated-infected) control that measures the number of host cells infected with parasite in their cytoplasmic region.Inhibition norm.para.num.: Antiparasitic parameter normalized by negative (DMSO treated-infected) and positive (Benz.treated-infected) control that measures the number of parasites in the cytoplasmic region of host cells.Cell ratio (%): cytotoxicity parameter that count the number of host cells.

General Experimental Procedures
Optical rotations were measured using a Jasco P-2000 polarimeter (JASCO Corporation, Tokyo, Japan).UV spectra were obtained with an Agilent 1100 DAD (Agilent Technologies, Santa Clara, CA, USA).IR spectra were recorded on a JASCO FT/IR-4100 spectrometer (JASCO Corporation, Tokyo, Japan) equipped with a PIKE MIRacle TM single-reflection ATR accessory.NMR spectra were recorded on a Bruker Avance III spectrometer (500 and 125 MHz for 1 H and 13 C NMR, respectively) equipped with a 1.7 mm TCI MicroCryoProbe TM (Bruker Biospin, Falländen, Switzerland).Chemical shifts were reported in ppm using the signals of the residual solvent as internal reference (δ H 2.50 and δ C 39.51 for DMSO-d 6 ).LC-MS and LC-HRMS analyses were performed as described previously [26].

Taxonomic Identification of the Producing Microorganism
The producing strain CA-135719 was obtained from a rhizosphere sample associated to Veronica anagallis-aquatica, collected in the Baviaanskloof mountains in the Eastern Cape of South Africa.This rhizosphere soil was air-dried, heated at 100 • C for 1 h, and then suspended in sterile water.The soil suspension was serially diluted, plated on selective isolation media, and incubated at 28 • C for at least 6 weeks.The strain was isolated from an NZ-amine-based agar medium containing nalidixic acid (20 µg/mL).The colony was purified on yeast extract, malt extract, glucose medium (ISP2) and preserved as frozen agar plugs in 10% glycerol.
The morphological and 16S rRNA gene sequence data were indicative that strain CA-135719 was representative of members of the genus Actinomadura and the strain was referred to as Actinomadura sp.CA-135719.

Fermentation of the Producing Microorganism
A 3 L fermentation of the producing microorganism was generated as follows: the first seed culture of the strain CA-135719 was prepared by inoculating 10 mL of seed medium, which consists of soluble starch (20 g/L), dextrose (10 g/L), NZ amine EKC (Sigma) (5 g/L), Difco beef extract (3 g/L), Bacto peptone (5 g/L), yeast extract (5 g/L), and CaCO 3 (1 g/L), adjusted to pH 7.0 with NaOH before addition of CaCO 3 , in a 40 mL tube with 0.5 mL of a frozen inoculum stock of the producing strain and incubation of the tube at 28 • C with shaking at 220 rpm for about 48 h.A second seed culture was prepared by inoculating 50 mL of seed medium in two 250 mL flasks with 2.5 mL of the first seed.The content of both flasks was then mixed, and a 5% aliquot of the mixture was used to inoculate twenty-four 500 mL flasks containing 125 mL of the production medium consisting of fructose 10 g/L, glucose 10 g/L, bacto peptone 2 g/L, bacto yeast extract 5 g/L, NZ-Amine E (EKC) 5 g/L, amicase 5 g/L, deionized water 1000 mL, trace elements 1 mL, pH 7.0.Trace elements: FeSO 4 •7H 2 O 500 mg/L, ZnSO 4 •7H 2 O 500 mg/L, MnSO 4 •H 2 O 100 mg/L, CuSO 4 •5H 2 O 50 mg/L, CoCl 2 •6H 2 O 50 mg/L.The flasks were incubated at 28 • C for 7 days in a rotary shaker at 220 rpm and 70% humidity before harvesting.

Extraction and Isolation
After 7 days of fermentation, acetone was added to the fermentation flasks (1:1, 125 mL), and they were shaken in a Kühner at 220 rpm for 2 h.After that, the mixture was centrifuged (10 min, 8500 rpm) and filtered under vacuum, and the pellet was discarded.Acetone was evaporated under a nitrogen stream until the original volume of fermentation (3 L) to obtain an aqueous residue.
The aqueous residue was loaded onto an SP207ss resin column (76 g, 32 × 100 mm) and eluted with an H 2 O−acetone increasing gradient (90/10 to 0/100 for 50 min, and 100/0 for 10 min, 10 mL/min, 15 mL/fraction) in a Teledyne CombiFlash RF apparatus to give 40 fractions.Fraction 22 was subjected to preparative reversed-phase HPLC (XBridge C 18 , 19 × 250 mm, 5 µm, 14 mL/min, UV detection at 210 and 280 nm, 7 mL/fraction) using H 2 O (solvent A) and CH 3 CN (solvent B).Elution was carried out using isocratic conditions of 5% B for 1 min and then a linear gradient from 5% to 50% B for 34 min, followed by a linear gradient from 50% to 100% in 1 min and held at 100% B for 7 min, yielding 80 fractions.

Marfey's Analysis of Compounds 1-4
A sample of compound 2 (700 µg) was dissolved in 0.7 mL of 6 N HCl and heated at 110 • C for 16 h.The crude hydrolysate was evaporated to dryness under a N 2 stream, and the residue was dissolved in 100 µL of water.This solution was divided into two 50 µL aliquots.To each aliquot of the hydrolysate and to an aliquot (50 µL) of a 50 mM solution of each amino acid (D and L), 20 µL of 1 M NaHCO 3 solution and a 1% (w/v) solution (100 µL) of D or L-FDVA (Marfey's reagent, N-(2,4-dinitro-5-fluorophenyl)-L-valinamide) was added.The reaction mixture was incubated at 40 • C for 60 min.After this time, the reaction was quenched by the addition of 10 µL of 1 N HCl, and the crude mixture was diluted with 200 µL of acetonitrile and analyzed by LC/MS on an Agilent 1260 Infinity II single-quadrupole LC/MS instrument.Separations were carried out on an Atlantis T3 column (4.6 × 100 mm, 5 µm) maintained at 40 • C. A mixture of two solvents, A (10% acetonitrile, 90% water) and B (90% acetonitrile, 10% water), both containing 1.3 mM trifluoroacetic acid and 1.3 mM ammonium formate, was used as the mobile phase under a linear gradient elution mode (isocratic 20% B for 2 min, 20−45% B in 27 min, isocratic 45% B for 5 min) at a flow rate of 1 mL/min.
Samples of compounds 1, 3, and 4 (400, 400, and 200 µg, respectively) were dissolved in 6 N HCl (0.8, 0.8, and 0.4 mL, respectively) and heated at 160 • C for 7 h.The crude hydrolysates were evaporated to dryness under a N 2 stream, and the residues were dis-solved in 200 µL of water.To these solutions, 100 µL of 1 M NaHCO 3 solution and a 1% (w/v) solution (100 µL) of D-FDVA were added.The reaction mixtures were incubated at 40 • C for 60 min.After this time, the reactions were quenched by the addition of 40 µL of 1 N HCl, and the crude mixtures were diluted with 200 µL of acetonitrile and analyzed by LC/MS using the method described above.

Genome Sequencing and Biosynthetic Gene Clusters Prediction
High molecular-weight genomic DNA from strain CA-135719 was isolated following the protocol described in Practical Streptomyces Genetics.A Laboratory Manual [28] from 10 mL bacterial culture grown on ATCC-2 liquid medium (soluble starch 20 g/L, glucose 10 g/L, NZ Amine Type E 5 g/L, meat extract 3 g/L, peptone 5 g/L, yeast extract 5 g/L, calcium carbonate 1 g/L, pH 7) grown on an orbital shaker at 28 • C, 220 rpm, 70% relative humidity for 4 days.
Then the genome was analyzed with antiSMASH v7.0.0beta1-67b538a9 [34] and PRISM v4 [35] to identify putative secondary metabolite biosynthetic gene clusters.BLAST (Basic Local Alignment Search Tool) [36] was also employed to predict the function of the genes.

Extraction of CA-135719 Cultures
Strain CA-135719 was cultured in fifteen Erlenmeyer flasks (500 mL containing 125 mL of culture medium) using the conditions described in Section 3.3.The flasks were extracted by addition of an equal volume of five different solvents, acetonitrile, n-butanol, ethyl acetate, methanol, and acetone, in triplicate.After addition of the solvent, flasks were shaken in a Kühner at 220 rpm for 2 h.The mixture was centrifuged (10 min, 8500 rpm) and filtered under vacuum, and the pellet was discarded.Organic solvents were evaporated under a nitrogen stream until half original volume to obtain the crude extracts.All crude extracts were analyzed by LC-HRMS.

Antiparasitic Activity on L. donovani
Human monocyte cell line THP-1 (ATCC TIB-202) was purchased from the American Type Culture Collection (ATCC).The assay for activity against L. donovani was performed as previously described [38].L. donovani MHOM/ET/67/HU3 was cultivated as promastigotes at 28 • C in M199 (Sigma-Aldrich, St. Louis, MO, USA) with 40 mM HEPES, 0.1 mM adenine, 0.0001% biotin, and 4.62 mM NaHCO 3 , supplemented with 10% heat-inactivated FBS and incubated 6 days before infection to enrich the proportion of metacyclic promastigotes.THP-1 was cultivated in RPMI medium (Thermo Fisher Scientific) supplemented with 10% heat-inactivated FBS and differentiated with 50 ng/mL of phorbol 12-myristate 13-acetate (PMA, Sigma-Aldrich) before infection.Purified L. donovani parasites were seeded with 1:20 infection ratio in 384-well plates and incubated for 24 h, then followed by the treatment of 4 µM amphotericin B (Sigma-Aldrich) as positive control, 0.5% DMSO as negative control, and tested samples.After additional incubation for 3 days at 37 • C, 5% CO 2 , the cells and parasites were fixed with 4% PFA and stained with 5 µM DRAQ5.Fluorescent images were acquired from each well using an automated confocal microscope (Operetta CLS) with a 635 mm laser at 20× objective.

Data Analysis
The acquired images were analyzed with Columbus software (v2.3) (PerkinElmer) to quantify cell numbers, parasites numbers, and infection ratio.Data were normalized by positive and negative controls.The activity of the tested samples was determined in terms of IC 50 and CC 50 calculated with XLfit software (v5) (IDBS, Boston, MA, USA).The quality of the assay data was assessed with Z' factor, assay windows, and CV [37].

Conclusions
In conclusion, five madurastatin derivatives were isolated from an acetone extract of the strain CA-135719 of Actinomadura sp.Four of these compounds, containing 4-imidazolidinone rings, H2 (2), 33-epi-D1 (3), D1 (4), and D2 (5), were shown to be artifacts formed during the acetone extraction of the culture broths.Madurastatin C1 is therefore the only natural madurastatin produced by the strain Actinomadura sp.CA-135719 in significant titers.Based on the similarity between the mad and mds BGCs, we revised and corrected the absolute stereochemistry of the previously reported madurastatins D1 and D2 [4] (also isolated from an acetone extract), establishing their stereocenters' absolute configuration as 9R, 23S, 26S, 33S and 9R, 23S, 26S, respectively, and based on our results and the biosynthetic evidence provided, we question the existence of ent-madurastatin C1 isolated in the same work.

Figure 1 .
Figure 1.Chemical structures of known madurastatins.* Indicates those compounds whose names were reported earlier for compounds having different structures.

Figure 1 .
Figure 1.Chemical structures of known madurastatins.* Indicates those compounds whose names were reported earlier for compounds having different structures.
Int. J. Mol.Sci.2024, 25, x FOR PEER REVIEW 4 of 17carbon at δC 80.2 (Figure4), indicating the connectivity and the presence of the imidazolidinone ring.Additionally, the fragment at m/z 171.1129 also confirms this assumption (Figure5).Compound 2 was named madurastatin H2 based on its structural similarity with madurastatin H1.
, whereas molecular formulae of C 29 H 43 N 7 O 10 (2) and C 28 H 39 N 7 O 9 (3 and 4) were assigned to the minor components of the extract.

Table 2 .
Cont.Carbon chemical shifts obtained from HSQC spectrum.Chemical shifts affected by the different absolute configuration at C-33 are shown in bold.

Table 3 .
Ion extraction by LC-HRMS of the targeted compounds (1-5) using five different extraction solvents.