Enhanced Cultivation of Actinomycetota Strains from Millipedes (Diplopoda) Using a Helper Strain-Assisted Method

Abstract

The limited cultivability of Actinomycetota strains restricts the exploration of their novel antibiotics, highlighting the need for improved isolation techniques. This study employed a helper strain-assisted cultivation method which utilizes culture supernatants from helper strains to isolate diverse members of the Actinomycetota from millipedes and compared its efficacy with a standard method. Using a preliminary dual-layer solid media assay and subsequent confirmation experiments, eight helper strains (M3, M9, M13, N3, N4, N6, N8, and N9) were identified, whose supernatants promoted the growth of Actinomycetota and other microbes. Application of this method to millipede samples established a novel cultivation strategy based on co-cultivation with helper strains. The new method enabled the isolation of 233 bacterial species in total, of which 143 were species of the phylum Actinomycetota, including 49 novel species. In contrast, the standard method yielded only 42 total bacterial species and 29 species of Actinomycetota, with merely 8 novel species. Comparative diversity analysis revealed that the helper strain-assisted method yielded Actinomycetota strains from 85 genera, which was 3.5 times higher than the standard method. This demonstrates that the helper strain-assisted approach is a highly effective strategy for accessing diverse and novel microbial majority. Among the isolated Actinomycetota strains, 75 strains predicted to have high biosynthetic gene clusters (BGCs) numbers or expected to be novel species were screened for antibacterial activity. Fourteen strains (17%) exhibited inhibitory effects against at least one indicator bacterium. One novel strain, Streptomyces sp. N8-31, was selected for whole-genome sequencing. AntiSMASH analysis predicted 40 biosynthetic gene clusters in N8-31, with 60% showing less than 70% similarity to known clusters; among these, 20 clusters showed less than 50% similarity. These findings indicate that strain N8-31 is a rich reservoir of novel genetic resources, and its broad-spectrum antibacterial activity is likely linked to these unique secondary metabolite gene clusters. Critically, this study confirms that helper strain-assisted cultivation is a powerful tool for unlocking the hidden biosynthetic potential of previously inaccessible Actinomycetota.

1. Introduction

Millipedes, as common soil arthropods, occupy unique habitats and possess unique physiological characteristics that not only endow them with diverse intrinsic ecological functions but also make them important carriers of microbial resources [1]. For example, chemicals secreted as a defensive mechanism by soil-dwelling millipedes have been shown to not only deter potential predators but also provide millipedes with chemical protection against pathogens and parasites, thereby enhancing their survival fitness in complex soil environments [2]. Over long-term evolutionary time, these highly variable and complex environments, in turn, have exerted significant evolutionary pressure on millipedes themselves and their associated symbiotic microbes [3]. Therefore, millipedes likely represent a significant reservoir of novel microbial resources, with a particularly high diversity of Actinomycetota (formerly known as Actinobacteria) [4,5,6].

Bacteria belonging to the phylum Actinomycetota are renowned as a premier source of clinically important antibiotics, producing a vast array of bioactive compounds including antimicrobial, antitumor, and immunosuppressive agents [7]. However, the over-exploitation of traditional sources has made the isolation of novel Actinomycetota strains from conventional environments like soil increasingly challenging. Therefore, exploring the Actinomycetota communities associated with millipedes holds considerable scientific and practical value.

The cultivation and isolation of microbial species are essential for understanding their characteristics and unlocking their biotechnological potential. However, most microbes from natural habitats cannot be cultivated on standard media. In recent years, the rapid development of biotechnology and high-throughput sequencing has enabled researchers to explore potentially microbial communities more extensively, enlarging the capacity to understand their diversity [8]. However, the inherent limitations of traditional cultivation techniques have restricted access to a mere fraction of the total microbial diversity, leaving the vast majority of microbial resources untapped. Therefore, developing new microbial cultivation techniques and expanding the range of cultivable microbial samples have become key objectives in contemporary microbiological research.

The “uncultivability” of a significant proportion of environmental microbes is largely attributable to the absence of essential nutrients or growth factors in standard laboratory media that are present in their natural habitats. To circumvent this limitation, novel in situ cultivation techniques, such as the diffusion chamber and the iChip, have been developed [9,10]. These methods facilitate the growth of previously uncultured microbes by more accurately simulating their natural conditions. Furthermore, subsequent research has indicated that microbial interactions are a critical component of these environments, demonstrating that growth stimulation by neighboring microbes can be necessary for isolating novel species [11,12,13].

Based on these findings, this study was designed to establish new cultivation strategies that utilize helper strains. These helper strains, selected based on prior microbial interactions tests, were employed to facilitate the isolation of previously uncultured and phylogenetically diverse Actinomycetota from millipede specimens. Culture supernatants from these growth-promoting helper strains were applied to enhance the recovery of novel Actinomycetota. The isolated Actinomycetota strains were screened for antimicrobial activity. One strain exhibiting potent activity and those suspected to be novel species were selected for whole-genome sequencing. Genomic analysis of biosynthetic gene clusters (BGCs) and functional protein annotations were conducted to assess their genetic potential for secondary metabolite production. The series of experimental procedures used in this study, including a new cultivation method using helper strains, provides an efficient way for the discovery of microbial based antibiotics in unique environments such as millipedes.

2. Materials and Methods

2.1. Candidates for Helper Strains

The experimental design for screening helper strains is outlined in Figure 1. Candidate helper strains consisted of 19 strains isolated from millipede samples (Prospirobolus joannsi) and 10 strains from a strain bank (Figure 1a). We collected two millipedes from Xiala Mountain (Ningbo, Zhejiang, China) in October 2023 and transported them alive to the laboratory for bacterial cultivation. Millipedes were surface-sterilized by brief rinsing with 70% ethanol, followed by three washes with sterile phosphate-buffered saline (PBS) to remove external contaminants. Under sterile conditions, each sample was individually homogenized in 1 mL of PBS using a vortex mixer for 15 min. The resulting homogenate was serially diluted from 10⁻³ to 10⁻⁵. An aliquot of each dilution was spread onto 1:10-diluted R2A agar plates and incubated at 28 °C for 7 days. After incubation, distinct colonies were purified for identification. From the 74 bacterial strains ultimately isolated and identified from the millipedes, 19 Actinomycetota strains were randomly selected as candidate helper strains. In addition, 10 bacterial strains known to have growth-promoting activity in the Ningbo Institute of Marine Medicine (NIMM) strain bank were included as helper strain candidates (Table S1).

2.2. Selection of Type Strains

To validate the effectiveness of the selected 19 candidate helper strains, interaction tests were performed together with eight type strains: four from the phylum Actinomycetota and four from other phyla. Three of these strains, obtained from the Ningo Institute of Marine Medicine (NIMM) strain bank, were Kitasatospora cheerisanensis (NIMM 60144), Kitasatospora phosalacinea (NIMM 60145), Kitasatospora gansuensis (NIMM 60205). Five strains, obtained from the Marine Culture Collection of China (MCCC), were Ornithinimicrobium laminariae (MCCC 1K06093T), Erythrobacter luteus (MCCC 1F01227T), Maribacter cobaltidurans (MCCC 1K03318T), Luteolibacter marinus (MCCC 1K04772T), and Oceanobacillus pacificus (MCCC 1K01074T).

2.3. Preliminary Screening for Helper Strains Using a Double-Layer Plate Method

A double-layered plate method was employed for the preliminary screening of helper strains, as previously described with modifications (Figure 1b) [14,15]. This method assessed the growth-promoting effects of 29 candidate helper strains on eight type strains (dependent strains). Each of 29 candidate helper strains was pre-cultured in 5 mL of diluted R2A broth at 28 °C with shaking at 160 rpm for 3–5 days until the early stationary phase. Subsequently, 2–3 µL of each culture was spotted onto surface of diluted R2A agar (first layer) in large square plates (130 mm × 130 mm). The plate was divided into 12 sections, allowing for a maximum of 11 candidate spots and one uninoculated control section per plate. The plates were then incubated for 7–14 days to allow for colony formation of the helper strain candidates (Figure 1b). Each of the eight dependent type strains was cultured in diluted R2A broth for 2–3 days. The cultures were then empirically diluted to an appropriate density. For each assay, 0.5 mL of a diluted dependent strain culture was mixed with 40 mL of warm, liquefied and diluted R2A agar (held at 45 °C) and immediately poured as a uniform second layer over the first layer containing the pre-grown helper candidate colonies. The double-layered plates were incubated at 28 °C for 3 days. A positive interaction was recorded when a clear growth circle (halo) of the dependent strain formed around a candidate helper colony. Strains exhibiting such positive effects were selected for further confirmation experiments.

2.4. Confirmation of Growth-Promoting Effects Using Helper Strain Supernatants

Based on preliminary screening, 15 helper strains that exhibited positive interactions were selected for a confirmatory growth promotion assay (Table S1). To quantitatively assess their effect, the colony formation efficiency of dependent strains was compared between culture media supplemented with helper strain supernatants and untreated control media (Figure 1c). Helper strains and dependent strains were prepared as described for the preliminary screening assay. Cell-free supernatants were prepared from helper strain cultures via centrifugation (3000 rpm, 10 min, room temperature), repeated in triplicate. The resulting supernatant was then filter-sterilized using a 0.2-μm pore-size membrane filter (Merck-Millipore, Burlington, MA, USA) and added to diluted R2A agar medium at a final concentration of 1% (v/v). Each dependent strain was serially diluted from 10⁻¹ to 10⁻⁵ in sterile PBS. The dilutions were inoculated in triplicate onto two sets of plates: (1) diluted R2A agar supplemented with 1% helper strain supernatant, and (2) standard diluted R2A agar as a control. All plates were incubated at 28 °C for 3–5 days. After incubation, colony-forming units (CFU) were enumerated only from plates containing 30 to 300 colonies. The colony formation ratio was calculated by dividing the average CFU/plate on the supernatant-supplemented media by the average CFU/plate on the control media for each dependent strain.

2.5. Application of Helper Strain Supernatants for the Cultivation of Millipede-Associated Bacteria

Based on the quantitative results from the confirmatory assays, eight helper strains with significant growth-promoting effects were selected to facilitate the cultivation of bacteria from millipede samples (Table S1). The millipede specimens were processed as previously described, including surface sterilization, homogenization, and serial dilution. Aliquots of the diluted homogenates were spread onto diluted R2A agar plates supplemented with 1% (v/v) supernatant from each selected helper strain. Control plates containing standard diluted R2A agar without supernatant were prepared in parallel. All plates were incubated at 28 °C for 2 to 3 weeks. After the incubation, 60 colonies were randomly picked from each type of supernatant-supplemented medium and control for subsequent purification and identification. To construct an unbiased isolate library, bacterial colonies were picked using a sterile disposal needle at evenly spaced intervals across the entire agar plate, without selection based on colony morphology.

2.6. 16S rRNA Gene Sequencing and Phylogenetic Identification

Single colonies were directly used as templates for PCR amplification. The 16S rRNA gene was amplified using the universal bacterial primers 27F and 1492R, as previously described by Lane with a BBL Taq polymerase (Sangon Biotech, Shanghai, China) [16]. The PCR products were purified and commercially sequenced by Sangon Biotech (Shanghai, China) using the fluorescent dye-terminator method. The resulting sequences were compared to the EzBioCloud database (https://www.ezbiocloud.net accessed on 20 November 2024) to identify the nearest taxonomic relatives.

2.7. Comparison of Microbial Diversity via Illumina HiSeq Sequencing

To compare the diversity of cultivated bacteria with the total bacterial community in millipede samples, 16S rRNA gene amplicon sequencing was performed using the Illumina HiSeq platform. The pretreated millipede homogenates, prepared as described above, were stored at −80 °C prior to DNA extraction. Genomic DNA was extracted, amplified, and sequenced by Shanghai Lingen Biotechnology Co., Ltd. (Shanghai, China) following their standard protocols. Microbial diversity was assessed at the genus level by comparing the composition of three groups: (1) the original millipede microbiota (via HiSeq sequencing), (2) bacterial isolates obtained from media supplemented with helper strain supernatants, and (3) isolates from control media without supernatant.

2.8. Prediction of BGCs Profiles by PSMPA

The 16S rRNA gene sequences of bacterial isolates obtained through the helper strain-assisted cultivation method were compiled into a FASTA file. This file was subsequently analyzed using the PSMPA (https://www.psmpa.net/ accessed on 20 June 2025) (Predicting the Secondary Metabolism Potential using Amplicon) web platform to predict the potential for secondary metabolite production. PSMPA [17], a tool developed by Zhejiang University, that predicts the profiles of biosynthetic gene clusters (BGCs) based on 16S rRNA gene amplicon data.

2.9. Screening for Antibacterial Activity

The antibacterial activity of 75 Actinomycetota strains (Table S2), of which 8 strains came from the control group and 67 strains were isolated via helper strain-assisted cultivation, was tested. They were evaluated against four model strains: Staphylococcus aureus CMCC 26003, Bacillus subtilis JMC 1465, Escherichia coli CMCC 44102, and Pseudomonas aeruginosa NIMM 10828. Primary antibacterial screening was conducted using a 24-well plate assay. Each well contained 1 mL of agar (1.5% agar in 30% R2A broth) seeded with a model strain. A 5 µL aliquot of each Actinomycetota strain culture was spotted onto the agar surface. The plates were incubated for 24 to 48 h, and the formation of inhibition zones was assessed visually on a daily basis [18]. Susceptibility testing was performed on selected strains using the Kirby-Bauer disk diffusion method [19]. Crude extracts were prepared according to the protocol described by Saar et al. (2025) [20]. Sterile filter paper disks (6 mm diameter) were impregnated with the crude extract and placed onto 30% R2A agar plates that had been pre-inoculated with a model strain. After incubation at 28 °C for 24–48 h, the diameters of the inhibition zones were measured. All experiments were performed in triplicate.

2.10. Genomic Analysis of Strain N8-31

The putative novel strain N8-31 (deposited as NIMM 61070) was selected for whole-genome sequencing based on its high predicted potential for biosynthetic gene cluster (BGC) production, as indicated by prior in silico analysis. For DNA extraction, the strain was cultivated in 50% R2A broth at 28 °C for 48 h. Genome sequencing, de novo assembly, and functional annotation were performed by Shanghai Biozeron Biotechnology Co., Ltd. (Shanghai, China) following their standardized protocols.

3. Results

3.1. Preliminary Screening of Helper Strain Candidates

To identify helper strains capable of promoting bacterial growth, we investigated the interactions between 29 candidate helper strains and 8 type strains (dependent strains) using a double-layer plate assay. A total of 30 positive growth-promoting interactions were observed, originating from 16 distinct candidate strains against the 8 type strains (Figure 2). The specific growth conditions are shown in Figure S1a,b. The selected representative helper strains significantly promoted the growth of type strains. In the agar medium, the closer the type strain is to the helper bacteria, the denser the number of strains. The type strain forms a distinct halo around the candidate helper colony. Notably, 15 of these 16 candidate strains enhanced the growth of Actinomycetota type strains (K1, K2, K3, and TA8). These 15 strains were therefore selected for further confirmation of their growth-promoting activity. Only one candidate strain, M12, exhibited a positive effect exclusively on non- Actinomycetota type strains and was set aside for other studies.

3.2. Confirmation of Growth Promoting of Selected Helper Strains

The growth-promoting effects of the 15 selected helper strains were quantitatively confirmed by comparing the colony-forming efficiency of dependent type strains on media with and without helper strain supernatants. Eight of the 15 helper strains enhanced the growth of their corresponding type strains, resulting in a ≥1.5-fold increase in colony-forming units (CFU) compared to the control (Figure 3). The specific growth conditions are shown in Figure S1c,d. The selected representative helper strains significantly promoted the growth of type strains. In the agar medium, the number of type strains in the experimental group was noticeably higher than in the control group. According to the result, we selected 8 helper strains, M3, M9, M13, N3, N4, N6, N8, and N9 (Table S1), for further cultivation experiments.

3.3. Enhanced Cultivation of Diverse Actinomycetota Strains Using Helper Strain Supernatants

To evaluate the efficacy of the eight selected helper strains in facilitating the isolation of diverse Actinomycetota strains, their supernatants were incorporated into agar media used to cultivate bacteria from millipede samples; a control medium without supplementation was prepared in parallel. In media supplemented with the supernatants from helper strains, we consistently observed a greater diversity and abundance of colonies compared to control (Figure S2a–d). A total of 540 bacterial isolates were identified, representing 244 operational taxonomic units (OTUs), 78 of which were novel species (<98.5% 16S rRNA similarity to the closest known relative among the isolates in EzBioCloud) [21]. The new method resulted in the isolation of 233 different bacterial species (OTUs), including 72 novel species. In contrast, the standard method only led to the isolation of 42 OTUs, including 11 new species (Table S3).

We further compared the diversity of Actinomycetota strains in particular (Figure 4). The untreated control media produced only 29 Actinomycetota OTUs including 8 novel species, while the helper strain-assisted cultivation yielded 143 Actinomycetota OTUs including 49 novel species, markedly exceeding those numbers of the control.

Comparison of 16S rRNA gene diversity revealed that the supernatant-based method enabled the isolation of significantly more bacterial genera than the control. The control method recovered isolates from 24 genera, representing only 14% of the total genera detected in the environmental samples via Hi-seq sequencing. In contrast, helper strain-assisted cultivation recovered isolates from 85 genera, accounting for 50% of the environmental diversity (Figure 5).

3.4. Antimicrobial Activity Screening in Selected Actinomycetota Strains

The biosynthetic potential of all isolates was initially assessed by predicting the number of biosynthetic gene clusters (BGCs) using the PSMPA platform (Table S4). We emphasize that the 16S rRNA-based phylogenetic analysis was used not to predict antibacterial activity, but as a preliminary diversity-based screening strategy. Its purpose was to select a phylogenetically representative subset of strains from our large isolate collection for subsequent empirical antibacterial testing, thereby improving overall screening efficiency.

Based on this analysis, 75 Actinomycetota strains were selected for antimicrobial activity assessment, of which 8 strains came from the control group and the remaining 67 strains came from a helper strain-assisted cultivation method. This subset included strains predicted to possess over 30 BGCs and those identified as putative novel Actinomycetota strains (Figure 6). Among these, 14 isolates (18.7%) exhibited antimicrobial activity against at least one of the test organisms, including Staphylococcus aureus, Bacillus subtilis, and Escherichia coli (

Table 1. Susceptibility of 4 model strains to the antibacterial activity of 14 Actinomycetota strains.

Isolate Number	Closest Species	Similarity	1	2	3	4
M3-11	Mycobacterium grossiae	98.29	+	++	−	−
M3-27	Streptomyces corynorhini	98.00	+	+	−	−
M3-28	Streptomyces glauciniger	99.10	++	++	−	−
M13-7	Homoserinibacter gongjuensis	97.89	−	+	−	−
M13-36	Streptomyces sodiiphilus	97.24	+	+	−	−
N3-6	Aldersonia kunmingensis	100.00	+	−	−	−
N4-5	Dactylosporangium tropicum	99.43	++	+++	−	+
N4-7	Kitasatospora cheerisanensis	99.01	−	+	−	−
N4-25	Streptomyces corynorhini	100.00	++	+	−	−
N8-28	Streptomyces alboniger	98.79	+	+	−	−
N8-31	Streptomyces cinnabarigriseus	98.07	+++	++	−	−
N9-6	Cellulomonas pakistanensis	97.55	++	+++	−	−
N9-23	Streptomyces cinnabarigriseus	98.71	++	++	−	−
M9-35	Streptomyces setonii	99.84	+	+	−	−

1: Staphylococcus aureus; 2: Bacillus subtilis; 3: Pseudomonas aeruginosa; 4: Escherichia coli; Positive control used 10 mg/mL Penicillin and 10 mg/mL chloramphenicol. Size (mm) of test strains suppression as an indicator of antibiotic activity intensity: «−»: no activity; «+»: ≤10, «++»: 11–15, «+++»: 16–20.

).

3.5. Genomic Analysis of a Putative Novel Species (N8-31) Reveals High Biosynthetic Potential

The putative novel strain N8-31 (NIMM 61070), which demonstrated antibacterial activity, was used for genomic analysis to elucidate its antibiotic production potential. Initially, we used the CARD for functional annotation and identified 134 antibiotic resistance genes [22], which were categorized into six major mechanisms: antibiotic efflux, target modification, target protection, inactivation, and target replacement (Figure 7). Genes associated with antibiotic efflux (66 genes) and target alteration (55 genes) were predominant, suggesting robust self-protection mechanisms that could facilitate the production of potent antibacterial compounds. This genomic evidence implies that the bacteriostatic substances produced by strain N8-31 may target essential pathways such as ribosome function or cell wall synthesis.

Analysis with antiSMASH predicted 40 biosynthetic gene clusters (BGCs) in the strain N8-31 genome [23]. Notably, 60% of these BGCs exhibited less than 70% similarity to known clusters, indicating a high degree of novelty (

Table 2. Predicted secondary metabolite gene cluster in strain N8-31 by anti-SMASH.

Region	Most Similar Known Species	Similarity (%)
Region 1	Other	100
Region 2	NRP	7
Region 3
Region 4	NRP	100
Region 5	Terpene	100
Region 6	NRP: Cyclic depsipeptide	57
Region 7	Other	100
Region 8	NRP	100
Region 9	Other	88
Region 10	NRP	100
Region 11	Polyketide	100
Region 12	NRP: Cyclic depsipeptide	13
Region 13	NRP + Polyketide	4
Region 14	Polyketide	23
Region 15	Other	57
Region 16	Polyketide: Iterative type I polyketide	10
Region 17	NRP + Polyketide	12
Region 18	Terpene	61
Region 19	Polyketide	16
Region 20
Region 21	NRP + Polyketide	4
Region 22	NRP + Polyketide	33
Region 23	Other	35
Region 24	Terpene	100
Region 25	Terpene	100
Region 26
Region 27	NRP	7
Region 28	Other	13
Region 29	RiPP	21
Region 30	NRP	10
Region 31	Polyketide	83
Region 32	Polyketide + Saccharide: Hybrid/tailoring saccharide	10
Region 33	NRP + Polyketide: Modular type I polyketide + Saccharide: Hybrid/tailoring saccharide	13
Region 34	RiPP: Lanthipeptide	85
Region 35	Other	13
Region 36
Region 37	Polyketide	13
Region 38	Polyketide	54
Region 39	Polyketide	21
Region 40	Polyketide	5

). The BGCs displayed considerable chemical diversity, encompassing non-ribosomal peptides (NRPS), type I polyketides (T1PKS), terpenes, ribosomal peptides (RiPPs), phosphonates, and nucleosides. Several clusters showed minimal similarity to known antibiotic BGCs (<15%), including a unique nucleoside cluster (Region 20; 0% similarity) and a type I polyketide cluster (Region 21; 4% similarity). Furthermore, T1PKS—a hallmark of potent antibiotics like erythromycin—was identified in eight distinct genomic regions (16, 17, 21, 27, 28, 37, 38, and 39), suggesting a high potential for novel antibacterial mechanisms. The genome also encoded complex potential ‘Trojan horse’ siderophores (Region 19), indicating the potential for entirely new modes of action. Collectively, these results position strain N8-31 as a highly promising source for discovering novel antibiotics to combat resistant bacterial infections.

4. Discussion

The disparity between standard laboratory growth conditions and natural environments is a major impediment to microbial cultivation [24,25,26]. We hypothesized that this uncultivability often stems from interdependent relationships among microbes, which necessitate specific signaling or cofactors for growth. This premise led us to adopt a helper strain-assisted cultivation approach, leveraging such microbial interactions to isolate strains of the phylum Actinomycetota from millipede samples. The success of this strategy not only validates the importance of microbial interdependence but also provides a robust platform for accessing unexplored microbial diversity and facilitating the discovery of novel antibiotics.

Previous studies have established that microbes in shared environments engage in complex interactions, and that manipulating these interactions can facilitate the cultivation of uncultured species [27,28]. Based on these findings, we screened 19 strains isolated from millipedes, along with ten previously established helper strains, for their ability to promote the growth of uncultivated bacteria. To target microbes with high biosynthetic potential, we selected four Actinomycetota strains—Kitasatospora spp. (K1, K2, K3) and Ornithinimicrobium sp. (TA8) and also four type strains from other phyla (TA3, TB5, TV6, TF7) as indicator (dependent) strains [29]. The screening of 232 co-culture pairs revealed 30 instances (12.9%) of enhanced colony formation compared to the control (Figure 2). This result provides compelling evidence for the growth-promoting role of specific inter-strain interactions.

Based on the interaction tests, eight strains (M3, M9, M13, N3, N4, N6, N8, and N9) that enhanced the colony formation (cell recovery) of other strains were selected for subsequent cultivation with millipede samples (Figure 3). The isolation efficiency was evaluated by comparing media supplemented with supernatants from these helper strains against a control. Culture collections from media containing supernatants of selected eight helper strains yielded a greater diversity of Actinomycetota strains (Figure 4) and bacterial genera (Figure 5) than the control. This finding is significant, as it not only expands the number of culturable Actinomycetota strains but also provides a scalable strategy to enhance microbial recovery. Unlike traditional methods focused solely on nutrient optimization, our supernatant-based approach leverages microbial metabolites—potentially including growth factors or signaling molecules—to activate the growth of previously uncultivable bacteria including diverse Actinomycetota strains.

Although the growth-promoting activity of cell-free supernatants from helper strains has been demonstrated in pure culture, its efficacy during the initial isolation stage from a sample remains unclear due to the presence of complex competitive microbiota. A successful isolation strategy should therefore incorporate not only positive growth factors (Figure S2a–d) but also antagonistic factors that can suppress fast-growing dominants [10]. In laboratory co-culture, faster-growing species often outcompete others for resources, even if these slower growers are dominant in their natural habitat [9,30]. Consequently, applying negative pressure (e.g., via antagonism) on these fast growers can shift the cultivated community’s composition, potentially enriching it with species that would otherwise be missed. The underlying mechanisms of these key growth factors (both positive and negative) require further experimentation, which should be a priority for follow-up studies.

A comparative analysis of culture-independent 16S rRNA gene sequencing and our cultivation results confirmed that the supernatant-supplemented media successfully isolated numerous bacterial taxa that were not accessible by standard cultivation media (Figure 5). Specifically, this approach recovered 85 genera, a substantial increase over the 24 genera obtained from the control. This outcome validates that simulating microbial interactions by modifying culture conditions is a powerful strategy for accessing previously uncultured microbial diversity.

However, a notable discrepancy was observed: several taxa that were highly abundant in the native millipede microbiome, such as the phylum Bdellovibrionota and the class Epsilonproteobacteria, were absent from our cultures. This suggests that these groups may require highly specific growth conditions—such as unique nutrient sources, precise oxygen availability, or the presence of specific host or microbial partners—that were not replicated in our experimental setup [31,32]. Consequently, while our method significantly expanded the recovery of Actinomycetota strains, these findings underscore the persistent challenge of cultivating certain fastidious microbes and highlight the need for further refinement of cultivation strategies to target these elusive lineages.

The functional potential of the isolated Actinomycetota strains further underscores the efficacy of our cultivation strategy. Among the 75 strains screened for bioactivity, 14 (18.7%) exhibited antibacterial activity, six of which are novel species. This high hit rate not only confirms that the helper strain-assisted method accesses greater phylogenetic diversity but also suggests it specifically enriches for Actinomycetota strains with significant biosynthetic potential.

To elucidate the genetic basis of its bioactivity, we selected Streptomyces sp. N8-31—a putatively novel strain exhibiting potent antibacterial activity—for whole-genome sequencing. Analysis against the Comprehensive Antibiotic Resistance Database (CARD) revealed no major pathogenic factors or concerning resistance genes. Conversely, genomic mining identified numerous biosynthetic gene clusters (BGCs) dedicated to secondary metabolite synthesis. This genetic profile not only indicates a favorable safety profile but also robustly supports the strain’s potential as a prolific source of novel antibiotics—a finding of critical importance in the context of the escalating antimicrobial resistance crisis [33,34,35].

In conclusion, this study establishes soil-dwelling millipedes as a rich reservoir of diverse and functionally significant Actinomycetota strains. The helper strain-assisted cultivation method not only expands the cultivability of Actinomycetota strains but also provides a new paradigm for microbial cultivation by leveraging ecological interactions for the identification of novel, bioactive, and biosynthetic gene cluster-rich strains such as Streptomyces sp. While our findings demonstrate that the eight helper strains promote the growth of the type strain through secreted metabolites, the chemical identities and biosynthetic origins of these compounds remain to be elucidated. Future work will therefore focus on isolating the active components via HPLC-MS and validating their biosynthetic pathways through genetic approaches such as targeted gene knockout. N8-31 lays a solid foundation for future antibiotic discovery. With further optimization of cultivation and characterization techniques, millipede-associated microbes are poised to become key resources in addressing global challenges in microbial ecology and drug development.