Actinomycin Derivatives: Structural Diversification and Biological Activities

Abstract

Actinomycins constitute a class of bioactive compounds known for their potent cytotoxic properties. Among them, actinomycin D, isolated in the 1940s from Streptomyces antibioticus, is the most extensively studied derivative. Structurally, actinomycins are characterized by a planar phenoxazinone chromophore flanked by two cyclic pentapeptides. They exert their biological effects primarily through DNA intercalation and transcription inhibition. Despite its remarkable bioactivity, the clinical application of actinomycin D is limited by significant adverse effects, including hepatotoxicity and restricted selectivity. These limitations underscore the need for structurally optimized analogs with improved therapeutic profiles and reduced toxicity. Over the decades, natural product discovery, precursor‐directed biosynthesis, and synthetic modification have yielded more than 70 structurally distinct derivatives, incorporating variations in both the peptide rings and the chromophore core. These structural modifications have resulted in diverse antibacterial, antiviral, and cytotoxic activities. Thus, this review critically examines the historical development, chemical diversity, and in vitro and preclinical bioactivity of actinomycin derivatives, highlighting key structural modifications over time and discussing their implications for future drug development.

Actinomycins are chromopeptide antibiotics exhibiting remarkable structural diversity and broad bioactivity. This review traces their development from naturally occurring D‐, X‐, A‐, G‐, Y‐, and Z‐type analogs to precursor‐directed, synthetic, and semisynthetic derivatives. Historical nomenclature inconsistencies are clarified, and key biological activities, including anticancer and antimicrobial effects, are comparatively summarized to guide future research and therapeutic exploration.

1. Introduction

Actinomycins are a family of chromopeptide antibiotics produced predominantly by Streptomyces species, although they have also been reported from other actinomycetes, including Micromonospora. Structurally, actinomycins consist of a phenoxazinone chromophore linked to two cyclic pentapeptide lactones with variable amino acid compositions [1, 2, 3]. Further diversity is introduced into the peptide structure through rearrangements, such as β‐ring threonine modifications or the incorporation of heteroatoms. However, recent studies have also identified derivatives with modifications on the phenoxazinone chromophore, indicating that structural diversity extends beyond the peptide rings. Depending on whether the amino acid composition of the two cyclic pentapeptides is identical, actinomycins can be classified as iso‐ or aniso‐actinomycins. Collectively, these structural variations define the A‐Z‐type subclasses, which exhibit distinct biological activity profiles [4, 5]. Up to date, more than 70 structurally distinct actinomycin derivatives have been reported, along with several analogs obtained via precursor‐directed biosynthesis, synthesis, and semisynthesis [5, 6]. Among these, the best‐known member, actinomycin D (AMD), was discovered in the 1940s. AMD exerts its biological activity primarily through DNA intercalation and subsequent inhibition of RNA synthesis, which underlies both its potent antitumor efficacy and dose‐limiting toxicity [6, 7, 8]. The findings ultimately led to its clinical development as one of the earliest chemotherapeutic agents [9, 10, 11]. Given its clinical relevance, further studies have demonstrated that structural modifications profoundly influence biological activity. For example, replacing N‐methylvaline (MeVal) with N‐methylalanine (MeAla) markedly reduced antitumor activity, whereas substituting MeVal with N‐methylleucine (MeLeu) enhanced in vitro potency [12, 13].

Despite the identification of distinct actinomycin derivatives over seven decades, AMD remains the sole clinically approved member of this family. This lack of clinical translation, despite promising preclinical activities, reflects fundamental challenges in achieving sufficient therapeutic selectivity and acceptable toxicity profiles. Consequently, actinomycins are no longer considered a major focus of anticancer drug development, although they remain valuable tools for chemical biology and mechanistic studies.

This review summarizes actinomycin derivatives, focusing on their chemical structures and, where available, their biological activities, to provide an overview of their chemical diversity, biological activities, and translational potential. In addition, it outlines the historical progression of actinomycin discovery and clarifies inconsistencies in nomenclature and synonym usage across the literature, thereby consolidating fragmented reports into a coherent framework.

2. Naturally Occurring Actinomycins

2.1. Actinomycin Type D

Actinomycin D (AMD, 5) comprises a 2‐amino‐4,6‐dimethylphenoxazin‐3‐one‐1,9‐dicarboxylic acid chromophore linked to two cyclic pentapeptide lactones (Figure 1). The three‐dimensional structure of AMD has been elucidated by X‐ray crystallography and NMR studies, revealing a nonplanar phenoxazinone chromophore stabilized by intramolecular hydrogen bonding and conformational constraints imposed by the cyclic pentapeptides [14, 15]. AMD binds DNA through a dual recognition mechanism in which the phenoxazinone chromophore intercalates between adjacent base pairs while the cyclic pentapeptides occupy the minor groove. The planar chromophore intercalates preferentially at GpC steps within double‐stranded DNA [2, 16].

FIGURE 1.

Schematic overview of AMD‐mediated transcriptional inhibition and its mechanistic downstream effects in mammalian (green sector) and microbial cells (purple sector). The AMD–DNA complex was generated from the crystal structure of the d(GAAGCTTC)₂‐AMD complex (PDB ID: 173D) (created in BioRender. https://BioRender.com/het7ha0).

Elucidation of the AMD structure also facilitated characterization of its biosynthetic pathway (Figure 2). As the first step in biosynthesis, tryptophan is converted to the precursor 4‐methyl‐3‐hydroxyanthranilic acid (4‐MHA) via a multi‐step pathway shown in Figure 2a [17, 18, 19, 20, 21]. Actinomycin biosynthesis is orchestrated by a nonribosomal peptide synthetase (NRPS) system. During peptide assembly, integrated methyltransferase domains catalyze N‐methylation of amino acids, while epimerization domains convert L‐amino acids to D‐configuration. These modifications occur co‐translationally as the peptide chain is assembled on the NRPS, not as post‐assembly tailoring steps. The depsipeptide macrocycle is formed through intramolecular esterification between the threonine hydroxyl group and the thioester‐linked MeVal, followed by oxidative coupling of the two pentapeptide halves via phenoxazinone synthetase (PHS) [22, 23, 24, 25, 26] (Figure 2b,c).

FIGURE 2.

Biosynthesis of AMD. (a) Biosynthetic pathway for the formation of 4‐methyl‐3‐hydroxyanthranilic acid (4‐MHA). (b) Assembly‐line organization of the nonribosomal peptide synthetase (NRPS) system. (c) Proposed final step in actinomycin biosynthesis. Abbreviations: ACMS I, actinomycin synthetase I; ACMS III, actinomycin synthetase III; AcmACP, actinomycin acyl carrier protein; A, adenylation domain; T, thiolation domain; PCP, peptidyl carrier protein; C, condensation domain; E, epimerization domain; M, methyltransferase domain; TE, thioesterase domain; Thr, threonine; Val, valine; Pro, proline; Gly, glycine (created in BioRender. https://BioRender.com/xzy6iq4).

2.1.1. Molecular Mechanisms of AMD

The antitumor activity of AMD is primarily mediated through transcriptional inhibition following DNA binding. AMD binds strongly but reversibly to DNA by intercalating its phenoxazinone chromophore between adjacent base pairs, while the cyclic pentapeptides interact with the minor groove. This interaction preferentially occurs at GpC‐rich sites and impedes RNA polymerase progression, with transcription elongation being more sensitive than initiation. Among the three nuclear RNA polymerases, RNA polymerase I (RNAP I) is the most sensitive target, followed by RNAP II and RNAP III, resulting in pronounced inhibition of ribosomal RNA synthesis at nanomolar concentrations [27, 28, 29, 30]. Although transcriptional inhibition represents the principal mechanism of action, AMD has also been reported to inhibit DNA topoisomerases I and II in human cells. Further investigation revealed that AMD forms stable coordination complexes with Cu(II), which enhance oxidative DNA strand cleavage, particularly in the presence of H₂O₂, suggesting a redox‐mediated mechanism [31]. This interference is particularly damaging to rapidly proliferating cells, such as cancer cells, which rely heavily on efficient DNA and RNA synthesis for their uncontrolled growth. AMD ultimately leads to cell cycle arrest and apoptosis (programmed cell death) in malignant cells, thereby reducing tumor burden. However, this mechanism also affects healthy rapidly dividing cells such as bone marrow and gastrointestinal epithelium, leading to the side effects.

Inhibition of ribosomal RNA synthesis triggers nucleolar stress, a hallmark of AMD activity. Previous studies demonstrated reversible translocation of nucleophosmin (B23), accompanied by suppression of RNA synthesis and cell proliferation, in HeLa cells [32, 33].

Nucleolar disruption further promotes the sequestration of MDM2 by ribosomal proteins, thereby stabilizing and activating p53 in sensitive cells [34]. These signaling events contribute substantially to AMD's antiproliferative effects.

Downstream of transcriptional inhibition and nucleolar stress, AMD induces apoptosis through both p53‐dependent and p53‐independent mechanisms. In pancreatic ductal adenocarcinoma cells, AMD activated stress‐activated protein kinase/c‐Jun N‐terminal kinase (SAPK/JNK) signaling and increased Bax expression independently of p53 or Bad, leading to apoptosis in different pancreatic cancer lines (PANC‐1, CAPAN‐1, and T3M4) [35, 36]. In addition, AMD induced mitochondrial dysfunction characterized by mitochondrial membrane depolarization, cytochrome c release, activation of caspase‐9 and caspase‐3, and loss of anti‐apoptotic proteins, including MCL‐1, in several tumor models [37, 38, 39]. These findings highlight the dual p53‐dependent and p53‐independent apoptotic mechanisms of AMD.

2.1.2. Cytotoxic Activity and Combination Studies

Numerous studies have investigated the cytotoxic activity of AMD across diverse cancer cell types, either as monotherapy (Table 1a) or in combination with other agents (Table 1b). CellMiner (release 2025.3) data from the NCI‐60 drug screen (sulforhodamine B assay) were analyzed to evaluate AMD's cell line‐dependent cytotoxicity. CellMiner (release 2025.3) data from the NCI‐60 drug screen (sulforhodamine B assay) showed that AMD is among the most potent compounds in the panel, with GI₅₀ values predominantly in the picomolar to nanomolar range. AMD displayed broad cytotoxicity across most cell lines, with a mean GI₅₀ of approximately 0.0014 μM. Hematopoietic malignancies were particularly sensitive, with all leukemia cell lines exhibiting subnanomolar GI₅₀ values, whereas several solid tumor lines, including breast, colon, and melanoma, showed comparatively reduced sensitivity. For example, HL‐60 (TB) leukemia cells exhibited a GI₅₀ of approximately 0.0002 μM, whereas MCF‐7 breast cancer and SK‐MEL‐28 melanoma cells required concentrations of approximately 0.2–0.3 μM to achieve comparable growth inhibition. These findings indicate marked cell line–dependent differences in AMD sensitivity and suggest that rapidly proliferating hematopoietic tumors are particularly susceptible to transcriptional inhibition.

TABLE 1.

In vitro cytotoxic activities of AMD.

a. Monotherapy applications			b. Combinatorial therapy applications
Effect	Target/cell type	References	Compound/combinatorial effect	Target/cell type	References
↑ Hallmarks of apoptosis (e.g., chromatin condensation, nuclear fragmentation) ↑ Caspase‐3 activity	T98G glioblastoma cells	[40]	Nutlin‐3a
			In NCI‐H28 cells; Total p53 ↑ ∼4‐fold p53‐Ser46 ↑ ∼40‐fold ↓ Phospho‐p70S6K (Thr389) in mTOR pathway In A549 cells; ↑ PML protein levels	Multiple cancer cell lines/cell‐type dependent response	[41]
↓ Nucleolar coefficients (nucleophosmin, RH‐II/Gu) (p < 0.03) ↑ incidence of micronucleoli and nucleolar remnants (p < 0.005)	HeLa cells	[42]	Histone deacetylase inhibitor (SAHA)
			↓ Viability by >20% with IC50 values of about 1 µM for SAHA and 0.5 nM for AMD	SK‐N‐JD neuroblastoma cells	[43]
↑ DNA damage response via γ‐H2AX induction	HeLa and MCF‐7	[44]	Cisplatin
			Mitochondria‐mediated apoptosis (↑ P53, ↑ PUMA, ↑ BAX, ↑ cytochrome c release, ↑ PARP cleavage, ↓ BCL‐XL)	KB cells	[45]
↑ Mitochondria‐dependent apoptosis (cytochrome c release, ↑ Puma mRNA) ↓ Procaspase‐9 levels ↑ Caspase‐3 activity ↑ p53 protein ↔ Bcl‐2 protein	Human peripheral blood lymphocytes	[46]	Dimethylamino‐parthenolide
			↑ Caspase‐positive populations, c‐Jun phosphorylation ↓ NF‐κB p65 phosphorylation	PANC‐1 cells	[47]
AKT phosphorylation at Ser473 At high concentrations ↓ MDM2 levels → p53 stabilization At low concentrations ↑ Ribosomal stress → ↓ MDM2 activity → p53 activation	Multiple cancer cell lines	[48]	Nutlin‐3a
			↑ p53 stabilization and activation (↑ Ser392 phosphorylation)	A549 cells	[49]
↓ Downregulated Sox2 in vitro and in vivo ↓ stemness markers (Sox2, Oct4, Nanog and c‐Myc)	Patient‐derived recurrent glioblastoma stem‐like cells (TD2, TS9)	[50]	Telmisartan
			Anticancer initiating stem cells effects ↑ ROS, caspase‐3 ↓ Expression of stemness markers (e.g. Sox2 and ALDH)	Lung cancer tumoroids	[51]
↓ Mitochondrial membrane potential loss in dose‐dependent way ↓ cytochrome C levels ↓ GSH levels ↑ Phosphorylation of ERK1/2, p38, JNK, AKT, and P70S6K	Hepatocellular carcinoma cell line (Hep‐G2)	[52]	Bortezomib
			↑ Apoptotic signaling and tumor regression	Wilms tumor models (WT‐CLS1, WiT49, and anaplastic WT 17.94)	[53]
↑ p53 and Ser15 dose‐ and time‐dependent ↑ p21 expression In A549 cells; ↑ PUMA mRNA and ↑ PUMA protein In H460 cells; ↔ PUMA mRNA and ↑ PUMA protein	Head/neck and lung cancers (A549 and H460)	[54]	Doxorubicin
			↑ p53, PUMA, BAX↓ MDM2, BCL‐2A	Triple‐negative breast cancer cell lines (HCC1937 and MDA‐MB‐436)	[55]

To overcome the narrow therapeutic index of AMD, several combination strategies have been explored (Table 1b). In a B16‐F10 melanoma model, co‐administration of tributyrin and AMD significantly enhanced tumor growth inhibition compared with either treatment alone, an effect attributed to butyrate‐induced TNFα production by stromal or immune cells [56]. Similarly, a combination of AMD with the recombinant immunotoxin RG7787 produced synergistic activity against mesothelin‐positive cancer cells and was associated with enhanced NF‐κB signaling [57]. Synergistic interactions have also been reported between AMD and echinomycin in DNA mismatch repair‐deficient HCT116 colon cancer cells (combination index ≈0.5) [58]. In addition, co‐treatment with resveratrol potentiated antiproliferative activity by activating p53 signaling and upregulating p21, PUMA, and GDF15, accompanied by reduced Ki‐67 expression in vivo [59].

2.1.3. Clinical Applications and Limitations

AMD remains the only clinically approved member of the actinomycin family. It is used primarily in pediatric oncology for the treatment of Wilms tumor, rhabdomyosarcoma, and Ewing sarcoma, as well as in gynecologic oncology for gestational trophoblastic neoplasia and in selected patients with metastatic nonseminomatous germ cell tumors. In clinical practice, AMD is rarely administered as monotherapy and is generally incorporated into multi‐agent chemotherapy regimens, including VAC (vincristine‐AMD‐cyclophosphamide) for rhabdomyosarcoma and vincristine/AMD‐based protocols for Wilms tumor and other pediatric sarcomas [60, 61, 62]. In addition, AMD has been employed as a component of regional perfusion protocols for the palliative or adjunctive treatment of locally recurrent or locoregional solid malignancies. Because of poor gastrointestinal absorption, AMD is administered exclusively by the intravenous route. Pharmacokinetic studies indicate that the drug undergoes limited metabolism, preferentially accumulates in nucleated cells, exhibits poor penetration across the blood–brain barrier, and has a terminal plasma half‐life of approximately 36 h. About 30% of an administered dose is recovered in urine and feces within 1 week. Despite its clinical efficacy, the therapeutic application of AMD is constrained by dose‐limiting toxicities, particularly myelosuppression, hepatotoxicity, nephrotoxicity, and gastrointestinal adverse effects. Moreover, resistance mechanisms, including drug efflux and enhanced DNA repair, may reduce therapeutic efficacy in certain tumor types [63, 64, 65]. Additionally, AMD can cause a “radiation recall” phenomenon, where skin reactions (erythema, blistering) potentially occur in previously irradiated areas, even months or years after radiation therapy. Long‐term treatment regimens incorporating AMD have also been associated with an increased risk of secondary malignancies, further limiting its therapeutic window.

2.1.4. Other Bioactivities

2.1.4.1. Antibacterial and Anti‐Virulence Activity

AMD exhibits antibacterial activity against both Gram‐positive and Gram‐negative pathogens and has additionally been reported to interfere with biofilm formation, quorum sensing, and bacterial virulence mechanisms. In various research, AMD exhibited bioactivity against both methicillin‐sensitive and methicillin‐resistant S. aureus strains, inhibiting bacterial growth, suppressing biofilm formation, and modulating the expression of virulence‐associated genes [66, 67, 68, 69, 70, 71, 72, 73, 74]. Resistance in MRSA could be partially overcome by treatment with D‐serine (Ser) or D‐threonine (Thr), which alter the composition of the peptidoglycan precursor, or by subinhibitory concentrations of tunicamycin, an inhibitor of teichoic acid biosynthesis, thereby enhancing bacterial susceptibility to AMD [72]. In addition, AMD inhibited biofilm formation in Staphylococcus epidermidis ATCC 35984 in a dose‐dependent manner, reducing biofilm biomass by more than 90% at 1.5 µg/mL by disrupting the ica operon [75].

AMD has also demonstrated activity against several other bacterial pathogens. Although only weak inhibition of Mycobacterium tuberculosis pantothenate synthetase was initially reported [76], subsequent studies showed antibacterial activity against M. tuberculosis H37Ra and H37Rv strains, with protein kinase B (PknB) identified as a potential molecular target [77, 78]. Additional antibacterial effects have been described against Klebsiella pneumoniae [79], whereas proteomic analyses identified the flagellar motor switch protein FliN as a potential target in Vibrio parahaemolyticus [80]. Beyond growth inhibition, AMD exhibited pronounced anti‐virulence effects. Treatment reduced hemolysin production, motility, and biofilm formation in Aeromonas hydrophila [81]; inhibited the alternative peptidoglycan biosynthesis enzyme MurD2 in Xanthomonas oryzae [82]; and interfered with quorum sensing pathways in Pseudomonas aeruginosa, resulting in reduced motility phenotypes and altered expression of quorum sensing–associated genes [83, 84].

Despite these promising in vitro activities, the antibacterial application of AMD is severely limited by its pronounced cytotoxicity toward mammalian cells. Thus, it has not been developed as a clinically useful antibacterial agent.

2.1.4.2. Antifungal and Antiparasitic Activity

Antifungal activity has also been reported for AMD‐producing Streptomyces isolates. For example, AMD was obtained from the endophytic Streptomyces sp. Tc022, isolated from Alpinia galanga roots, and exhibited weak activity against Colletotrichum musae and Candida albicans [85]. Streptomyces sp. IA1 was isolated from Saharan soil and produced antifungal activity against a broad range of fungi, exhibiting inhibition zones of 19–45 mm against Fusarium culmorum, Fusarium equiseti, Fusarium moniliforme, Fusarium proliferatum, Aspergillus carbonarius, Aspergillus niger, and Umbelopsis ramanniana. Notably, AMD significantly reduced disease symptoms of chocolate spot (Botrytis fabae) in field bean by up to 91.7% and Fusarium wilt (Fusarium oxysporum) in flax by nearly 60% [86]. Lately, AMD was identified as an inhibitor of polysaccharide deacetylase (Agd3) in Aspergillus fumigatus, weakly reducing biofilm biomass at a concentration of 100 µM [87].

AMD has also demonstrated antiparasitic activity, although available studies remain limited. In Leishmania major, AMD showed activity against promastigotes, with an EC₅₀ of 1.90 µg/mL. However, the corresponding cytotoxic concentration (CC₅₀ = 0.10 µg/mL) resulted in a very low selectivity index (SI = 0.053), indicating poor selectivity toward the parasite relative to mammalian cells and thereby limiting its therapeutic potential [88].

2.1.4.3. Antiviral Activity

AMD has demonstrated antiviral activity against several RNA and retroviruses. Early studies showed that AMD inhibited replication of Pichinde virus in BHK‐21 fibroblast cells in a dose‐dependent manner at concentrations ranging from 0.4 to 4 µg/mL. Notably, administration up to 12 h post‐infection still resulted in a 50% reduction in viral yield. The data indicated that AMD did not inhibit viral RNA synthesis but rather interfered with a late stage of viral maturation, likely involving host membrane glycoproteins required for virion assembly [89]. Similarly, AMD inhibited human coronavirus 229E replication, with a 50% inhibitory dose (ID₅₀) of 0.1 µg/mL [90]. In simian rotavirus (SA11)–infected cells, treatment with 2.5 µg/mL AMD reduced virus‐specific fluorescence by more than 50% and decreased cytoplasmic metachromasia by approximately 25%, without affecting viral RNA replication [91].

The antiviral activity of AMD has also been investigated against HIV‐1. AMD inhibited several stages of reverse transcription, including DNA‐dependent DNA synthesis and strand transfer reactions, primarily through binding to viral nucleic acids [92, 93, 94]. However, paradoxical enhancement of HIV‐1 replication has also been reported at low concentrations, accompanied by altered cytokine signaling and reduced susceptibility to certain nucleoside analogs [95, 96]. These findings highlight the complex and context‐dependent effects of AMD on HIV replication.

Collectively, the antiviral effects of AMD are highly context‐dependent. Because AMD primarily targets host transcriptional machinery rather than virus‐specific proteins, both antiviral and proviral effects have been reported depending on the virus, host cell, and experimental conditions.

2.1.5. Other D‐Type Analogs

Methylated actinomycin D (mAMD) displayed potent cytotoxicity against multiple cancer cell lines while exhibiting lower toxicity toward normal HL‐7702 liver cells, suggesting a modest improvement in selectivity relative to AMD. Mechanistic studies indicated that mAMD induces caspase‐dependent apoptosis through both mitochondrial and Fas‐mediated pathways [97]. In 2018, four analogs, actinomycins D₁–D₄, were isolated together with AMD (Figure 3). The compounds were generally less cytotoxic than AMD toward WI‐38 fibroblasts, although all retained antibacterial activity against MRSA, with actinomycin D₁ exhibiting an MIC of 0.125 μg/mL. In contrast, N‐acetylation (D₃) or a Val‐to‐Ala substitution (D₄) did not confer a clear biological advantage over AMD [98]. Overall, all four analogs were less cytotoxic, suggesting that the structural modifications introduced in D₁–D₄ did not improve the antiproliferative effect.

FIGURE 3.

Chemical structures of actinomycins D₁ (1), D₂ (2), D₃ (3), D₄ (4), and D (5); 7‐aminoactinomycin D (6); neo‐actinomycin D (7); and hemiactinomycin (8).

In addition, hemiactinomycin (8), a biosynthetic intermediate in actinomycin assembly rather than a mature analog, exhibited anti‐inflammatory activity by inhibiting nitric oxide (NO) production in LPS‐stimulated RAW 264.7 macrophages (IC₅₀ = 15.41 ± 0.66 μM) [99]. These findings demonstrate that even minor structural modifications within the actinomycin scaffold can substantially influence biological activity in vitro.

2.2. Actinomycin Type X

Among actinomycin subtypes, X‐type derivatives represent the second most studied group after D‐type analogs. Actinomycin X₂ (synonymous with actinomycin V; 9) is a widely studied analog that differs from AMD by the presence of a 4‐oxoproline (OPro) moiety. In contrast, actinomycin X_0β (synonymous with actinomycin A₁, 10) contains trans‐4‐hydroxyproline (Hyp) instead of proline (Pro) in the β‐ring (Figure 4). Early studies established that actinomycin X preparations consisted of several closely related analogs differing mainly in the oxidation state and stereochemistry of the proline residue. Subsequent structural studies identified additional members, including X_0α (11), X_0δ (12), and X_1a (13), thereby expanding the X‐type subclass (Figure 4) [100, 101]. Although Sharma and Manhas (2019) labeled one of their purified compounds as actinomycin V, the depicted structure and chromatographic characteristics are consistent with those of actinomycin X_0β. In the broader literature, actinomycin V is synonymous with actinomycin X₂, not with actinomycin X_0β, indicating that the compound assignment in the original publication was likely incorrect [102].

FIGURE 4.

Chemical structures of actinomycins X₂ (9), X_0β (10), X_0α (11), X_0δ (12), X_1a (13), and Bis(hyp)‐X₁ (14). Differences to AMD (5) are shown in bold letters. All amino acids are L‐configured except when indicated otherwise. Abbreviations: Thr = threonine, D‐Val = D‐valine, Pro = proline, Sar = sarcosine, MeVal = N‐methylvaline, Hyp = trans‐4‐hydroxyproline, OPro = 4‐oxoproline, αHyp = cis‐4‐hydroxyproline.

Cytotoxic evaluation of X‐type actinomycins revealed a clear potency hierarchy, with actinomycin X₂ consistently outperforming both AMD and X_0β in several leukemia and solid tumor models. Actinomycin X₂, isolated from Streptomyces sp. MITKK‐103, demonstrated pronounced activity against HL‐60 leukemia cells (LC₅₀ = 8 nM), markedly exceeding that of AMD (45 nM) and X0β (>100 nM) [103]. In lung cancer models, apoptosis induction by actinomycin X₂ was cell‐type dependent. A549 cells exhibited a substantial increase in apoptotic fraction (6.19% to 45.23%), whereas NCI‐H1299 cells showed minimal sensitivity (4.23% to 6.39%) [104]. Similarly, across various breast cancer cell lines, actinomycin X₂ displayed sub‐nanomolar potency (IC₅₀ = 0.83–4.16 nM), corresponding to 4.29–18.25‐fold greater cytotoxicity than AMD [105, 106]. Comparable trends were observed in colorectal cancer models (IC₅₀ = 6.38–8.65 nM), while substantially reduced toxicity was noted in nonmalignant QSG‐7701 and HEK‐293T cells (IC₅₀ = 68.3 and 82.6 nM, respectively), suggesting partial tumor selectivity. Mechanistically, actinomycin X₂ induced apoptosis via mitochondrial dysfunction, involving the activation of caspase‐9, caspase‐3, and PARP, and the inhibition of the phosphatidylinositol 3‐kinase (PI3K)/Akt pathway [107]. Notably, actinomycin X₂ from Streptomyces sp. ZZ338 exhibited nanomolar cytotoxicity, with IC₅₀ values ranging from 0.42 to 1.80 nM against glioblastoma and astrocytoma cell lines, and surpassed doxorubicin in selected models [108]. Collectively, available studies indicate that oxidation of the β‐ring proline residue, particularly to 4‐oxoproline, consistently enhances cytotoxic potency relative to AMD, potentially through altered steric and hydrogen‐bonding interactions within the peptide periphery. Despite consistently superior in vitro potency, actinomycin X₂ has not yet advanced to clinical evaluation.

Furthermore, X‐type actinomycins demonstrated broad antimicrobial activity, although potency varies considerably across species. Actinomycins X₂, X_0β, and AMD, from Streptomyces sp. MS449, inhibited Mycobacterium bovis BCG and M. tuberculosis H37Rv, with MICs ranging from 0.25 to 0.5 µg/mL and 1.0 to 8.0 µg/mL, respectively [109]. AMD and actinomycin X₂ demonstrated MICs ranging from 1.95 to 31.25 μg/mL against vancomycin‐resistant Enterococcus (VRE), Bacillus subtilis, E. coli (S1‐LF), MRSA, and K. pneumoniae subsp. pneumoniae [102]. Further, actinomycin X_0β and AMD inhibited the spore germination, hyphal growth, and biomass accumulation of Verticillium dahliae in a dose‐dependent manner [110]. Actinomycins X₂ and D exhibited IC₅₀ values of 17.6 and 29.6 μM, respectively, for inhibiting pyoverdine production in Pseudomonas cichorii SPC9018, while actinomycin X₂ further displayed antibacterial activity, with an MIC of 0.25 µg/mL against MRSA ATCC 43300 and 1.56–12.5 µg/mL against various other Gram‐positive pathogens [111, 112, 113]. In a study, actinomycin X₂ was reported with MIC values of 0.08, 0.12, and 0.83 µM against S. aureus, E. coli, and C. albicans, respectively [108]. In silico analyses proposed isoleucyl‐tRNA synthetase as a potential antimicrobial target, while penicillin‐binding protein‐1a was the least favorable target [113]. However, assigning specific antibacterial targets solely on the basis of docking studies should be further evaluated, as the physiological relevance of this interaction remains uncertain.

To enhance potency against Gram‐negative pathogens, combinatorial antimicrobial strategies have also been explored. In a study using E. coli ATCC 25922, synergistic effects were observed when actinomycin X₂ and AMD were combined with polymyxin B and kanamycin (most synergistic area scores >10, as indicated by SynergyFinder results). In contrast, actinomycin X_0β showed no synergistic effect with the tested antibiotics [114]. Further, actinomycin X₂ inhibited Xanthomonas citri subsp. citri at 31.25 µg/mL (versus 62.50 µg/mL for AMD and X_0β) and demonstrated greater antifungal potency against Penicillium italicum, while X_0β was inactive [115]. Additional studies showed inhibition of spore germination and ROS induction in postharvest citrus pathogens [116]. In addition, actinomycins X₂ and X_0β displayed antimalarial activity against Plasmodium falciparum (IC₅₀ = 0.241 and 7.69 nM, respectively). Notably, both compounds exhibited higher selectivity indices over AMD [117].

Alternative applications beyond systemic therapy have explored actinomycin X₂ as an antimicrobial dye for silk materials and as a functional component of fiber‐blend membranes and multifunctional hydrogels, demonstrating retained or enhanced antibacterial activity, along with improved mechanical properties [118, 119, 120, 121, 122]. Actinomycin X₂‐containing wound‐healing films further promoted wound closure in rat models [123]. While these findings highlight physicochemical robustness, their translational relevance to systemic therapy remains unclear. Targeted delivery strategies have also been investigated. Folate‐modified liposomal formulations of actinomycin X₂ achieved a tumor inhibition rate of 77.2% in hepatocellular carcinoma models, representing a 4.17–fold enhancement relative to free compound [124]. Overall, antimicrobial activities are broad but highly strain‐dependent and, like AMD, may primarily reflect global transcriptional inhibition rather than selective target engagement. A major unresolved challenge remains the separation of desirable biological activity from the intrinsic toxicity associated with the actinomycin scaffold.

2.3. Actinomycins Type A, B, and C

The A‐, B‐, and C‐type actinomycins constitute relatively small subclasses whose distinguishing features arise primarily from modifications of the chromophore rather than extensive alterations of the peptide rings. Historically, actinomycin A₁ (synonymous with actinomycin X_0β) and A₂ (synonymous with actinomycin II and F₈) were identified [125]. More recently, structural diversification within the A‐ and B‐type subclasses was expanded by the discovery of neo‐actinomycin A (15) and neo‐actinomycin B (16) from the marine‐derived Streptomyces sp. IMB094. These analogs feature a tetracyclic 5H‐oxazolo[4,5‐b]phenoxazine chromophore formed via condensation of the classical actinomycin scaffold with tricarboxylic acid cycle intermediates. Neo‐actinomycin A contains a carboxyethyl‐substituted oxazole ring, while neo‐actinomycin B features a methyl‐substituted oxazole ring (Figure 5). Neo‐actinomycin A exhibited moderate cytotoxicity (IC₅₀ 38.7 nM in HCT116; 65.8 nM in A549) and antibacterial activity against MRSA and VRE (MIC 16–64 μg/mL), whereas neo‐actinomycin B showed ∼10‐fold lower cytotoxicity and negligible antibacterial activity (MIC > 128 μg/mL) [126]. These findings suggest that chromophore extension alone does not uniformly enhance potency and that substituent identity on the oxazole ring significantly influences activity.

FIGURE 5.

Chemical structures of neo‐actinomycins A (15), B (16), and C (17) and actinomycins C₂ (18) and C₃ (19). Structural variations relative to the classical actinomycin scaffold are highlighted.

Within the C‐type subclass, actinomycin C₁ (synonymous with AMD), C₂ (18), and C₃ (19) analogs exhibited potent antibacterial activity against both S. aureus and M. tuberculosis, with activities comparable to or exceeding those of AMD [127]. Consistently, an earlier study reported that actinomycin C₂, actinomycin VII (synonymous with actinomycin C₃), and AMD inhibited the Shc signaling adaptor protein/Grb2 (growth factor receptor‐bound protein 2) interaction while blocking MAP kinase (Erk1/Erk2) phosphorylation in B104‐1‐1 cells at 1 µM [128]. These observations suggest that certain actinomycins may influence signaling pathways beyond their canonical role as transcription inhibitors. Nevertheless, whether these effects represent direct target engagement or secondary consequences of transcriptional suppression remains unclear. In a recent investigation, neo‐actinomycins C (17) and D (7) were isolated from Streptomyces sp. S22. Structurally, both of the compounds retain the characteristic chromophore, as seen in neo‐actinomycins A and B; however, neo‐actinomycin D is distinguished by the presence of an N‐methyl‐N‐(p‐tolyl)formamide moiety attached to the chromophore. Neo‐actinomycin D demonstrated enhanced antibacterial activity compared to neo‐actinomycin C, with 4‐ to 32‐fold lower MICs against ESKAPE pathogens (E. faecium, S. aureus, K. pneumoniae, Acinetobacter baumannii, P. aeruginosa, and Enterobacter sp.) and cytotoxicity against the HepG2 liver carcinoma cell line [129], underscoring the likely impact of appended aromatic substituents on antimicrobial activity.

2.4. Actinomycins Type G, S, Y, and Z

G‐ and Z‐type actinomycins are characterized by the substitution of N‐MeAla in place of N‐MeVal at position 5 of the α‐peptidolactone unit, along with selective methylation and oxidation of the proline residues (Figure 6). Actinomycin G₁ (20), isolated in the early 2000s from Streptomyces sp. HKI‐0155, had a moderate activity against B. subtilis ATCC 6633 [130]. Furthermore, actinomycins G₂–G₆ were produced by Streptomyces iakyrus DSM 41873 isolated from the rhizosphere of an orange tree. These derivatives displayed distinct structural variations compared to AMD. In particular, the R‐ring incorporated trans‐3‐hydroxy‐cis‐5‐methylproline (HMPro), while the β‐ring contained MeAla and either chlorinated or hydroxylated threonine residues. Additional G₃–G₆ (21 –25) analogs differ through positional isomerism, hydroxylation patterns, chromophore‐ring cyclization, or conversion of one peptidolactone into a cyclopeptide, further expanding the structural diversity of the G‐type subclass [131].

FIGURE 6.

Chemical structures of actinomycins G₁ (20), G₂ (21), G₃ (22), G₄ (23), G₅ (24), G₆ (25), and S (26).

Regarding biological activities, actinomycin G₂ exhibited 15‐, 27‐, and 45‐mm inhibition zones against E. coli, S. aureus, and B. subtilis in a disc diffusion assay (50 µg/disc), while actinomycin G₃ showed no activity against E. coli but inhibited S. aureus and B. subtilis (21 and 33 mm). Actinomycin G₅ was only active against S. aureus (10 mm) and B. subtilis (20 mm), whereas actinomycin G₆ showed no detectable activity. Later, cytotoxicity assays indicated that actinomycin G₂ had an GI₅₀ of 0.0013 µM and a TGI (total growth inhibition concentration) of 0.0043 µM, which were better than those of AMD (GI₅₀, 0.0016 µM; TGI, 0.0064 µM) against HM02 gastric adenocarcinoma cells [131]. Thus, the hydroxylated and methylated proline derivatives can be accommodated within the actinomycin scaffold without loss of activity and may even enhance cytotoxic potency relative to AMD.

Later, biosynthetic investigations identified a dedicated acmG gene cluster containing accessory hydroxylase and halogenase enzymes responsible for the characteristic hydroxylated and chlorinated amino acid residues found in G‐type actinomycins [132].

Actinomycin S (26) was isolated from Streptomyces sp. S22. Structurally, it features an aspartic acid residue in place of Pro in the β‐peptidolactone ring. This substitution resulted in moderate antibacterial activity (MIC 1.25–80 μg/mL) against ESKAPE pathogens and induced G0/G1 arrest in HepG2 cells [129]. In an earlier study, actinomycins S₂ (synonymous with AMD) and S₃ were isolated from Streptomyces sp. 1048A. Structural characterization identified actinomycin S₃ as containing OPro residue instead of Pro. The presence of OPro in actinomycin S₃ suggests partial structural convergence with X‐type derivatives [133], indicating that β‐ring oxidation patterns may influence bioactivity.

Later, Bitzer et al. [134] reported that Y‐type actinomycins (Y₁–Y₅; 27 –31) (Figure 7) were isolated from Streptomyces sp. strain Gö‐GS12. Y‐type actinomycins feature unique amino acid compositions, such as Hyp or OPro, that distinguish them from Z‐ and G‐type analogs. Among the Y‐series, actinomycin Y₁ was the most potent analog, whereas replacement of OPro with Hyp (Y₂) substantially reduced activity. Structural simplification in Y₄ and Y₅ resulted in a marked loss of both cytotoxic and antimicrobial potency [134]. Further studies done using Streptomyces sp. strain Gö‐GS12 identified novel aniso‐actinomycins Y₆ (32), Y₇ (33), Y₈ (34), Y₉ (35), and iso‐actinomycin Z_P. Structural analyses revealed β‐ring rearrangements in actinomycin Y₆ and a rare 5‐methylproline (MPro) in actinomycins Y₉ and Z_p. Antibacterial activity tests showed Y₉ as the most bioactive one (MICs between 0.012 and 0.003 µg/mL), followed by Y₈ (0.031–0.003 µg/mL) and Z_p (0.047–0.006 µg/mL) against S. aureus, Micrococcus luteus, and B. subtilis. Strikingly, aniso‐actinomycins Y₉ and Z_p exhibited sub‐nanomolar cytotoxicity (EC₅₀ 0.16–0.74 nM) comparable to or exceeding AMD [4]. These data suggest that specific β‐ring rearrangements and incorporation of 5‐methylproline enhance potency, particularly when combined with halogenation.

FIGURE 7.

Chemical structures of actinomycins Y₁ (27), Y₂ (28), Y₃ (29), Y₄ (30), Y₅ (31), Y₆ (32), Y₇ (33), Y₈ (34), and Y₉ (35).

Z‐type actinomycins (Figure 8) form a structurally distinct subclass initially reported in 1958 from Streptomyces fradiae ETH‐20675, consisting of six derivatives (Z₀–Z₅) [135]. These compounds lack unmodified Pro but contain MeAla and rare Pro derivatives, such as MPro (Z₅), 4‐keto‐5‐methylproline in Z₁ and Z₅, and HMPro in Z₁ [136, 137, 138, 139, 140]. Some variants, including Z₁, also contain 4‐hydroxythreonine (HThr), an uncommon amino acid among actinomycins [141]. In a study, Z‐type actinomycins isolated from S. fradiae were structurally elucidated and denoted as actinomycins Z₁ (36), Z₂ (37), Z₃ (38), Z₄ (39), and Z₅ (40). These analogs contain rare amino acids, such as 4‐chloro‐L‐threonine (ClThr). Z₃ and Z₅ demonstrated particularly strong cytotoxicity (IC₅₀ ∼ 0.1 μM), surpassing AMD in HM02, HepG2, and MCF‐7 cell lines [142].

FIGURE 8.

Chemical structures of actinomycins Z₁ (36), Z₂ (37), Z₃ (38), Z₄ (39), Z₅ (40), and Z₆ (41).

More recently, actinomycin Z₆ (41), along with actinomycins Z₁, Z₃, and Z₅, was isolated from Streptomyces sp. KIB‐H714. Actinomycin Z₆ was distinguished by the presence of an additional oxygen‐bridged ring linking the chromophore to the β‐peptidolactone. Although Z₆ displayed cytotoxicity across HL‐60 leukemia, SMMC‐7721 liver carcinoma, A549 lung carcinoma, MCF‐7 breast carcinoma, and SW480 colon carcinoma cell lines, the strongest activities were observed for actinomycins Z₃ and Z₅, both exhibiting IC₅₀ < 0.064 µM, surpassing the potency of cisplatin (1.91–13.26 µM) [143]. Notably, the most potent Z‐type analogs consistently contained modified proline residues and halogenated amino acids, further supporting the importance of peptide ring modifications rather than chromophore alterations as drivers of activity enhancement.

Accordingly, structural diversification across A–Z subclasses highlights that halogenation and proline oxidation consistently enhance cytotoxic potency, whereas chromophore extension alone does not guarantee improved activity. Modifications within the β‐peptidolactone ring appear particularly influential in modulating both antimicrobial and antitumor effects, underscoring the peptide periphery as a primary driver of SAR within the actinomycin family. Despite decades of investigation and the identification of analogs with improved in vitro potency, altered selectivity, or additional biological activities, these advances have not yet translated into clinical applications.

2.5. Actinomycins From Precursor‐Directed Biosynthesis

Precursor‐directed biosynthesis has been widely employed to expand the structural diversity of actinomycins by incorporating non‐native amino acids or chromophore analogs into the peptide framework (Figure 9). When L‐isoleucine (Ile) was added to the fermentation medium, two new derivatives, actinomycins E₁ (42) and E₂ (43), were produced, each incorporating one or two MeIle residues into its peptide cores [144, 145]. In contrast, supplementation with sarcosine yielded a broader series of compounds, including actinomycins F₁ (44), F₂ (45), F₃ (46), and F₄ (47), which reflect more extensive structural variation. All F‐series actinomycins contained threonine, sarcosine, and MeVal, but differed in Val, Ile, and Pro content, indicating substantial tolerance of the NRPS assembly line toward alternative imino acid incorporation. Notably, some analogs (e.g., F₁) lack proline entirely, thereby differing from the canonical actinomycin architecture [146].

FIGURE 9.

Chemical compositions of actinomycins E₁ (42), E₂ (43), F₁–F₄ (44 –47), A_u5 (48), A_u6a (49), A_u6b (50), A_u7a (51), A_u7b (52), K_1c (53), K_2c (54), K_1t (55), K_2t (56), Pip_1α (57), Pip_1β (58), Pip_1γ (59), Pip_1δ (60), Pip_1ε (61), Pip₂ (62), and azetomycins I and II (63, 64) are given while indicating differences to AMD (5) shown in bold letters. All amino acids are L‐configurated except when indicated otherwise (abbreviations: MeVal = N‐methylvaline, aIle = allo‐isoleucine, Sar = sarcosine, MPro = cis‐5‐methylproline). (*) The marked compounds were not examined to determine whether incorporation occurred in the α‐ or β‐ring and were likely mixtures containing both possibilities.

Supplementation with L‐leucine (Leu) or Ile generated a series of aurantin‐related actinomycin congeners (48 –52), demonstrating that branched‐chain amino acid precursors can be incorporated into the peptide scaffold during biosynthesis [147, 148, 149]. Biological evaluation of these analogs has been limited. A more targeted precursor approach involved culturing Streptomyces parvulus, a known producer of AMD, with cis‐5‐methylproline (MPro) isomers. Supplementation with cis‐4‐methylproline led to the production of actinomycins K_1c (53) and K_2c (synonymous with actinomycin K_2t, 54), in which the modified amino acid substituted one or both Pro residues. Similarly, feeding the system with trans‐4‐methylproline yielded K_1t (55) and K_2t (56) derivatives. Replacement of proline with methylproline derivatives reportedly reduced antibacterial potency. Compared to AMD (MIC 0.03 µg/mL against M. luteus), actinomycin K_2c exhibited > fivefold lower activity, while reductions against S. aureus reached 17‐ to 20‐fold. These data suggest that subtle steric alterations at the imino acid position critically affect DNA intercalation efficiency and antibacterial activity [150, 151].

Streptomyces antibioticus produces novel actinomycin derivatives when cultured with DL‐pipecolic acid, yielding three compounds: actinomycins Pip_1α (57), Pip_1β (58), and Pip₂ (62). These analogs differ from typical actinomycins by incorporating pipecolic acid and its derivative 4‐oxopipecolic acid (OPip) at the imino acid positions, alongside standard residues such as threonine, valine, sarcosine, and N‐MeVal. Specifically, actinomycin Pip₂ contains two pipecolic acid residues, while actinomycin Pip_1β has one pipecolic acid and one proline. Nevertheless, actinomycin Pip_1α contains pipecolic acid linked to OPip, an uncommon amino acid found in several peptide antibiotics. Biological activities of actinomycins Pip_1α and Pip₂ had reduced antibacterial activity compared to AMD, particularly against penicillin‐resistant S. aureus, where MIC values increased from 0.1 µg/mL (AMD) to 5.0 µg/mL (actinomycin Pip_1α) and 1.5 µg/mL (actinomycin Pip₂) [152, 153]. Therefore, the expansion of the five‐membered proline ring to a six‐membered pipecolic acid ring was found to be unfavorable for antibacterial activity.

Another study in the 1960s published that feeding Streptomyces antibioticus with non‐proteinogenic amino acids, such as azetidine‐2‐carboxylic acid (AzC) and pipecolic acid, yielded novel actinomycins in which AzC or pipecolic acid replaced one or both proline residues [154]. Consistently, another study reported that S. antibioticus, when cultured with AzC, produced novel actinomycins, named azetomycins I (synonymous with azetomicin I, 63) and II (64), in which one or both proline residues are replaced by AzC, respectively. Unlike most proline replacements, azetidine substitution largely preserved antibacterial potency while increasing total antibiotic production. This observation reflects that ring size reduction is better tolerated than ring expansion and highlights the flexibility of the actinomycin biosynthetic machinery [155].

Another example of biosynthetic manipulation was observed when Streptomyces chrysomallus and S. parvulus were cultured in media supplemented with 3‐hydroxyanthranilic acid (3‐HA), a structural analog of the chromophore precursor 4‐MHA. This led to the formation of C‐demethylactinomycins, namely mono‐C‐demethylactinomycin D and di‐C‐demethylactinomycin D. Although C‐demethylactinomycins constituted only 7%–8% of total actinomycin production, their natural occurrence (up to 0.8%) in non‐supplemented cultures suggests 3‐HA formation via a minor diversion from the 4‐MHA biosynthetic pathway [156]. In a mutasynthesis approach, 10 new halogenated derivatives of actinomycin D were generated by feeding halogenated (F⁻, Cl⁻, Br⁻, I⁻) or methoxy‐substituted 4‐MHA to a Streptomyces mutant strain. Among the resulting analogs, compounds 5 (4,6‐dichloro), 7 (4,6‐dibromo), 10 (4,6‐diiodo), and 12 (6‐iodo) showed the antibacterial activity against eight different S. aureus strains, M. luteus ML01, E. faecium ATCC 35682, E. faecalis ATCC 29212, and Bacillus thuringiensis BT01 with MIC values ranging <0.06 to 64 µg/mL [157]. These findings demonstrate that the chromophore tolerates extensive halogen substitution without complete loss of activity. Together with observations from naturally occurring Y‐ and Z‐type analogs, these results identify halogenation as one of the most consistently beneficial modifications reported within the actinomycin family.

A more recent discovery involved the actinomycin L‐type diastereomers (Figure 10) and actinomycins L₁ (65) and L₂, isolated from Streptomyces sp. MBT27, following fermentation in the presence of anthranilamide. These derivatives diverge from classical actinomycin scaffolds by forming a unique spirolinkage between the OPro residue of actinomycin X₂ and an appended anthranilamide moiety, a modification previously unobserved in natural actinomycins. Antibacterial tests revealed at least twofold less activity than AMD against Gram‐positive strains, with no inhibition observed against E. coli ATCC 25922 or K. pneumoniae ATCC 700603 (MICs > 128 µg/mL) [158]. Thereafter, precursor‐directed biosynthesis reveals limited but exploitable tolerance of the actinomycin scaffold to structural modification, delineating the boundaries of its structural flexibility. The spirolinked architecture of actinomycins L₁ and L₂ demonstrates that noncanonical chromophore modifications can be achieved through precursor‐directed approaches; however, the resulting compounds showed only moderate antibacterial activity and no clear advantage over established actinomycins.

FIGURE 10.

Chemical structure of actinomycin L₁ (65).

3. Synthetic and Semisynthetic Actinomycins

Synthetic and semisynthetic modifications have been instrumental in defining the structural features required for actinomycin activity and in assessing whether potency improvements can be achieved beyond those of AMD. Using a modular total synthesis approach, novel Bis(hyp)‐X₁ (14) and naturally occurring actinomycins X_0β and X_0δ were obtained. Actinomycin Bis(hyp)‐X1 contains trans‐4‐hydroxyproline residues in both peptide rings in place of proline [101], demonstrating that the imino acid position tolerates certain hydroxylated substitutions. However, available data suggest that hydroxylation alone does not confer a substantial pharmacological advantage over naturally occurring X‐type analogs. Furthermore, a study evaluated the SAR of 13 synthetic AMD analogs. It indicated that modifications at the cyclic peptide backbone's Val and/or MeVal positions significantly influenced both Grb2‐SH2 domain binding and cellular antiproliferative activity. Among synthetic analogs, compounds containing tyrosine displayed the strongest IC₅₀ values of 0.5 and 0.8 µM against B104‐1‐1 cell line, respectively. However, they showed weaker cell growth inhibition, likely due to poor cellular permeability. In contrast, analogs containing phenylalanine, O‐methyltyrosine (Ome), or Val showed potent antiproliferative effects, with IC₅₀ values ranging between 0.9 and 2 µM against HeLa cells. By comparison, analogs with threonine substitutions were significantly less active [159]. This infers aromatic or hydrophobic side chains are generally preferred at these positions, whereas polar substitutions compromise activity, likely through unfavorable effects on DNA binding and/or cellular permeability.

Subsequently, in 2008, a novel analog, L‐MeLeu⁵AMD ([D‐Val², L‐MeLeu⁵]₂‐AMD), was synthesized, in which MeVal was replaced with MeLeu. Further bioactivity assays demonstrated antiproliferative activity against SGC‐7901 human gastric carcinoma cells, with an IC₅₀ of 10 nM. Mechanistic studies confirmed that L‐MeLeu⁵AMD induced apoptosis via the mitochondrial pathway, mitochondrial membrane potential disruption, and downregulation of Bcl‐2 expression. In vivo testing in S‐180 sarcoma‐bearing mice showed dose‐dependent tumor suppression, with tumor weight inhibition (TWI%) reaching 63.42% at 102 μg/kg, exceeding AMD (53.8% TWI at 140 μg/kg) [160]. Notably, L‐MeLeu⁵AMD is among the few synthetic analogs reported to outperform AMD in both in vitro and in vivo models, highlighting the importance of N‐methyl amino acid substitutions within the peptide ring. Another study investigated AMD analogs generated by replacing the D‐Val and MeVal residues within the cyclic depsipeptide rings with alternative amino acids, including D‐Phe, L‐ and D‐N‐MeVal, N‐methylisoleucine (MeIle), MeLeu, N‐methylphenylalanine (MePhe), MeAla, and Sar. Most analogs carrying substitutions at the D‐Val position exhibited substantially reduced antimicrobial and cytotoxic activities relative to AMD, indicating that the native D‐Val residue is critical for biological activity. The acute toxicity of L‐MeLeu⁵‐AMD was comparable to that of AMD, whereas L‐MeIle⁵‐AMD was approximately 1.25‐fold more toxic. The three analogs exhibited cytotoxicity against prostate carcinoma (PC‐3), bladder carcinoma (Biu‐87 and EJ), and cervical carcinoma (HeLa) cells, with IC₅₀ values ranging from 0.017 to 0.121 μM, demonstrating that replacement of the MeVal residue is generally well tolerated and may modestly improve potency in selected cases, although no clear advantage in therapeutic profile was established. In antimicrobial assays, L‐MeIle⁵‐AMD and D‐MeLeu⁵‐AMD retained potent antibacterial activity, exhibiting submicromolar MIC values against E. coli, S. aureus, and B. subtilis [161]. Together, these studies revealed that substitutions at the MeVal position are among the most productive modifications explored within the actinomycin family. Replacement with larger hydrophobic residues such as MeLeu or MeIle can preserve or enhance biological activity, whereas smaller or more polar residues generally reduce potency. Unlike peptide modifications aimed at improving pharmacological properties, certain chromophore modifications have primarily been used as research tools rather than as therapeutic leads.

7‐Aminoactinomycin D is a semisynthetic derivative of AMD and is widely used as a fluorescent marker in viability assays to discriminate between apoptotic and necrotic cell death [162, 163, 164]. Overall, synthetic studies have provided valuable SAR insights and consistently indicate that modifications within the peptide periphery are better tolerated than extensive chromophore alterations. However, despite occasional improvements in potency, none of the reported derivatives has demonstrated a sufficiently improved therapeutic index to warrant clinical development.

4. Summary and Outlook

Actinomycins constitute one of the most extensively investigated families of chromopeptide natural products. To date, more than 70 naturally occurring, precursor‐directed, semisynthetic, and synthetic derivatives have been reported, revealing considerable structural diversity within both the phenoxazinone chromophore and cyclic peptide rings.

Several consistent structure–activity relationship trends emerge from these studies. Modifications within the peptide periphery generally exert a greater influence on biological activity than alterations of the chromophore. In particular, oxidation, methylation, and halogenation of certain amino acid residues frequently enhance cytotoxic or antimicrobial potency, whereas extensive replacement of proline or the introduction of highly polar residues typically reduces activity. Likewise, substitution of N‐methyl amino acid residues, especially with larger hydrophobic residues such as MeLeu or MeIle, can preserve or improve biological activity. In contrast, extensive chromophore modifications rarely confer clear pharmacological advantages and are often associated with reduced activity.

The biological activity of actinomycins remains fundamentally linked to DNA intercalation and the subsequent inhibition of transcription. Although numerous additional molecular targets and biological effects have been proposed, many reported activities are likely secondary consequences of this primary mechanism. Several derivatives, including actinomycin X₂, Leu⁵AMD, and selected Y‐ and Z‐type analogs, exhibit enhanced preclinical potency relative to AMD. Nevertheless, these improvements have not translated into clinical application, and AMD remains the sole clinically approved member of the family. Importantly, no structural modification reported to date has successfully uncoupled antitumor efficacy from the intrinsic toxicity associated with DNA intercalation and transcriptional inhibition.

The absence of new clinically useful actinomycins after more than seven decades of investigation highlights fundamental limitations of the scaffold. DNA intercalation and transcriptional inhibition are inherently nonselective processes that affect both malignant and normal proliferating cells, resulting in a narrow therapeutic window. Consequently, structural modifications that enhance potency rarely improve tumor selectivity, whereas those that reduce toxicity frequently compromise efficacy.

Future progress is therefore unlikely to arise solely from the discovery of additional analogs. Instead, greater translational potential may lie in strategies designed to overcome the intrinsic limitations of the actinomycin scaffold. Preliminary studies have demonstrated that targeted delivery approaches for AMD, including cell‐penetrating peptides, antibody‐mediated delivery platforms, and nanoparticle‐based formulations, can enhance intracellular delivery, modify biodistribution, or improve antitumor efficacy in preclinical models [165, 166, 167]. However, these approaches remain at an early stage of development, and whether they can provide sufficient improvements in therapeutic selectivity and safety to enable broader clinical application remains uncertain. In parallel, actinomycins are expected to remain valuable chemical biology tools for investigating transcriptional regulation and other DNA‐dependent processes.

Conflicts of Interest

The authors declare no conflicts of interest.

Biographies

Özge Can received her B.Sc. in Molecular Biology and Genetics from Konya Food and Agriculture University and M.Sc. in the Dept. of Bioengineering from Izmir Institute of Technology. Her master's thesis focused on isolation of actinomycin and collismycin derivatives from Streptomyces parvus 35M1 and their antibacterial and synergistic activities under the supervision of Prof. Erdal Bedir. Currently, she is a PhD student and a research assistant at the Dept. of Bioengineering from Izmir Institute of Technology, studying on bioactive fungal secondary metabolites.

Erdal Bedir, Pharm.D., Ph.D., conducted postdoctoral research at the National Center for Natural Products Research, University of Mississippi (1999–2003). He is currently a Full Professor in the Dept. of Bioengineering at İzmir Institute of Technology. His expertise lies in medicinal chemistry and pharmacognosy, with research focused on the discovery, isolation, and structural elucidation of bioactive natural products.

Data Availability Statement

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.