Microparticles enhance the formation of seven major classes of natural products in native and metabolically engineered actinobacteria through accelerated morphological development

Actinobacteria provide a rich spectrum of bioactive natural products and therefore display an invaluable source towards commercially valuable pharmaceuticals and agrochemicals. Here, we studied the use of inorganic talc microparticles (hydrous magnesium silicate, 3MgO·4SiO2·H2O, 10 µm) as a general supplement to enhance natural product formation in this important class of bacteria. Added to cultures of recombinant Streptomyces lividans, talc enhanced production of the macrocyclic peptide antibiotic bottromycin A2 and its methylated derivative Met‐bottromycin A2 up to 109 mg L−1, the highest titer reported so far. Hereby, the microparticles fundamentally affected metabolism. With 10 g L−1 talc, S. lividans grew to 40% smaller pellets and, using RNA sequencing, revealed accelerated morphogenesis and aging, indicated by early upregulation of developmental regulator genes such as ssgA, ssgB, wblA, sigN, and bldN. Furthermore, the microparticles re‐balanced the expression of individual bottromycin cluster genes, resulting in a higher macrocyclization efficiency at the level of BotAH and correspondingly lower levels of non‐cyclized shunt by‐products, driving the production of mature bottromycin. Testing a variety of Streptomyces species, talc addition resulted in up to 13‐fold higher titers for the RiPPs bottromycin and cinnamycin, the alkaloid undecylprodigiosin, the polyketide pamamycin, the tetracycline‐type oxytetracycline, and the anthramycin‐analogs usabamycins. Moreover, talc addition boosted production in other actinobacteria, outside of the genus of Streptomyces: vancomycin (Amycolatopsis japonicum DSM 44213), teicoplanin (Actinoplanes teichomyceticus ATCC 31121), and the angucyclinone‐type antibiotic simocyclinone (Kitasatospora sp.). For teicoplanin, the microparticles were even crucial to activate production. Taken together, the use of talc was beneficial in 75% of all tested cases and optimized natural and heterologous hosts forming the substance of interest with clusters under native and synthetic control. Given its simplicity and broad benefits, microparticle‐supplementation appears as an enabling technology in natural product research of these most important microbes.


| INTRODUCTION
Natural products are chemically diverse molecules that are synthetized by living organisms (Harvey, 2000). Many of them have potent pharmacological activities, which has enabled the development of antibiotics (Fleming, 2001), anticancer drugs (Arcamone et al., 1969), immunosuppressants (Vézina et al., 1975), infectants (Campbell et al., 1983), and various agrochemicals, including herbicides (Bayer et al., 1972), insecticides (Butterworth & Morgan, 1968), and fungicides (Takeuchi et al., 1958). During the last century, natural products and their derivatives have greatly helped to double human life expectation (Demain, 2006) and display more than 50% of today's approved drugs (Cragg & Newman, 2013). However, drug development from natural products is declining, partly caused by unreliable access and supply, leading inter alia to cost and profit concerns among pharmaceutical companies (Li & Vederas, 2009).
Actinobacteria display the most important microbial source of natural products and therefore offer an immense treasure (Barka et al., 2016). The genus Streptomyces provides more than two-third of all known antibiotics of microbial origin (Bibb, 2013) and more than half of the FDA-approved antibacterial natural products (Patridge et al., 2015) but also related actinobacteria such as Micromonospora, Actinoplanes, and Amycolatopsis (Barka et al., 2016) have emerged as producers of potent bioactive molecules. Unfortunately, actinobacteria produce natural products often in minute amounts, making even structural identification almost impossible, or do not form them at all under laboratory conditions (Ren et al., 2017). Moreover, natural products exhibit complex, highly diverse biosynthetic pathways and structures (Figure 1), which complicates efforts to streamline and enhance production (Hanson, 2003). Towards optimized supply, general strategies that allow to enhance natural product formation in actinobacteria are highly desired.
At this point, a promising line of research seems to exploit the link between natural product formation and the unique morphological life cycle of actinobacteria (Chater, 1984), which differentiates them from most other bacteria (Barka et al., 2016). In liquid culture, their morphological development starts from germinated spores that grow into a vegetative mycelium by linear tip extension and hyphae branching (Chater & Losick, 1997;van Dissel et al., 2014), followed by the formation of an aerial mycelium, which finally differentiates into uninucleoid cells that develop again into spores (Angert, 2005).
Recently, we discovered that supplementation with inorganic talc microparticles (hydrous magnesium silicate, 3MgO·4SiO 2 ·H 2 O) reprogrammed recombinant Streptomyces albus J1074/R2 to live a life of accelerated morphogenesis, which enabled a three-fold enhanced production of the antituberculosis polyketide pamamycin .
Here, we took this approach further. Exemplified for recombinant Streptomyces lividans TK24 DG2-Km-P41hyg + , the effects of talc were elucidated using transcriptomics and metabolomics together with detailed analysis of strain physiology and morphology. The recombinant producer, representing one of the most prestigious Streptomyces species used industrially (Sevillano et al., 2016), formed the ribosomally and posttranslationally modified peptide (RiPP) bottromycin under control of synthetic promoters, supposed to uncouple production from the morphological cell cycle.
It displayed the leading cell factory to produce bottromycin (Vior et al., 2020). The cyclopeptide was first isolated from cultures of Streptomyces bottropensis (Waisvisz et al., 1957) and is active against methicillin-resistant Staphylococcus aureus and vancomycin-resistant Enterococci (Kobayashi et al., 2010). Its complex biosynthesis involves a unique posttranslational macrocyclization step (Huo et al., 2012) and has posed enormous challenges on bottromycin research and development over the past decades (Kazmaier, 2020).
Finally, the microparticle approach was scaled down to the microtiter plate scale to investigate its potential in overproducing twelve natural compounds of commercial interest from ten major structural classes in native and metabolically engineered actinobacteria, including various Streptomyces but also other genera and families ( Figure 1).

| Strains
All strains used in this study are listed in Table 1. For strain maintenance, spores were collected from agar plate cultures after five-day incubation at 30°C, resuspended in 20% glycerol, and kept at −80°C.

| Cultivation
One loop of spores was scratched from a 5-day-old plate culture (incubated at 30°C) and used to inoculate a liquid pre-culture, which was grown overnight in a 500 ml baffled shake flask, filled with 50 ml medium and 30 g soda-lime glass beads (5 mm, Sigma-Aldrich). When the pre-culture had reached the late exponential phase, cells were collected (8500×g, room temperature, 5 min), resuspended in 10 ml fresh medium, and used to inoculate the main cultures, either containing talc or not. Main cultures were grown in shake flasks and in deep-well flower plates, respectively. Shake flask cultures were incubated in 500 ml baffled shake flasks (50 ml medium) on a rotary shaker (28°C, 230 rpm, 75% relative humidity, 5 cm shaking diameter, Multitron, Infors AG). For miniaturized screening, cells were grown in 48-well flower plates (1 ml medium) using a benchtop incubator (28°C, 1300 rpm, 75% relative humidity, BioLector m2p labs). All experiments were conducted as biological triplicate.

| Quantification of cell concentration
The cell dry weight (CDW) of S. lividans was measured as follows.
Cells were collected (10,000×g, 4°C, 10 min), washed twice with 15 ml deionized water, and freeze-dried . Subsequently, the dry biomass was gravimetrically determined . In microparticle cultivations of S. lividans, biomass data were corrected for the amount of talc added  2.5 | Quantification of glucose Glucose was quantified by HPLC (1260 Infinity Series, Agilent), using an Aminex HPX-87H column (300 × 7.8 mm; Bio-Rad) as stationary phase and 7 mM H 2 SO 4 as mobile phase (55°C, 0.7 ml min −1 ). Refraction index measurement was used for detection, and external standards were used for quantification .

| Microscopy
The 5 µl culture broth was transferred onto a glass for bright-field microscopy (Olympus IX70 microscope). The software ImageJ 1.52 (Schneider et al., 2012) was used to automatically determine the size T A B L E 1 Strains and natural products investigated in this study and corresponding culture conditions | 3079 of pellets formed during growth . At least 150 aggregates were analyzed per sample.

| Natural compound extraction and quantification
The different natural products investigated in this study (Table 1) were extracted from culture broth using a two-step process . In short, 500 µl broth was mixed with 500 µl acetone and incubated for 15 min (1000 rpm, room temperature, Thermomixer F1.5, Eppendorf Raw data sets (sequenced reads) as well as processed data sets (input matrix and normalized read counts from DESEQ. 2) are available from GEO (GSE168044). For statistical analysis, Student's t-test was carried out and the data were filtered for genes with a log 2 -fold change ≥ 1 (p ≤ 0.05) . RNA extraction and sequencing were conducted as biological triplicates for each strain.  (van Wezel et al., 2000). The production of the latter started after 48 h and continued until the end of the process.
In comparison, the microparticle supplemented culture achieved significantly increased product titers. The final level of bottromycin A2 (60 mg L −1 ) was 43% higher than that in the control, while the methyl-bottromycin A2 titer (46 mg L −1 ) was increased by 31%. Increased product levels already occurred after 48 h. In addition, undecylprodigiosin production was increased by the particle supplementation (particularly during later stages of the cultivation) and resulted in a 64% increased final titer. The enhanced accumulation of the red pigment was even visible by the eye (Figure 3).
Talc-supplemented cells appeared red, while cells from the control culture were less-intensively colored.
Interestingly, biomass formation and glucose consumption remained rather unchanged between the talc-supplemented culture and the control. However, talc had a major impact on cellular morphology. Grown without talc, the recombinant producer formed larger pellets with an average diameter of 390 ± 118 µm ( Figure 3), while cellular aggregates in the talc supplied culture were significantly smaller and exhibited a 40% reduced pellet diameter of 279 ± 49 µm. In later stages of the cultivation, the morphology differed also in inner structure ( Figure S1). The cell aggregates in the control reached up to 1 mm in diameter and pellets appeared decomposed inside. Pellets of the talc culture were less than half in size and exhibited an intact and denser structure. Further tests revealed a production optimum for total bottromycins (109 ± 24 mg L −1 ) at 15 g L −1 talc ( Figure S2). Higher talc levels were found detrimental.
3.2 | Microparticles do not affect bottromycin precursor availability during initial ribosomal translation but significantly change the abundance of intermediates of the posttranslational modification pathway We hypothesized that the increased bottromycin titers could result from a higher precursor availability, as recently shown for other natural products Kuhl et al., 2020;Ser et al., 2016;Tang et al., 1994;Thykaer et al., 2010). Therefore, we analyzed the level of intracellular amino acids during growth (12 h, botAH was specifically increased in the talc culture (2.4-fold) and botC and botA remained constant, while the latter two genes were found upregulated in the control (2.1-fold and 2.3-fold, respectively).
Other genes equally behaved under both conditions: botRMT1 was downregulated (0.4-fold), whereas botCD, botH, botT, botRMT2, and botRMT3 remained unaffected over the entire process. It was interesting to note that the expression of the cluster was only weakly driven by the synthetic promoters ( Figure S5a), but largely relied on the promoter of the hygromycin marker gene hygR (P hygR ), inserted in between them ( Figure S5b). In addition, a so far unknown but apparently even stronger promoter downstream of hygR, designated P AS , transcribed the cluster in the opposite (antisense) direction ( Figure S5c).

| Microparticles accelerate morphological development and aging of S. lividans
On a global level, talc-supplied and non-supplied cells massively changed their gene expression during the process ( Figure S3). In the control culture, 5% of all 7751 genes (414)  For the talc-amended culture, the transcription profile changed faster, and the dynamics involved more genes ( Figure S3). The observed changes were related to the cultivation stage ( Figure 5).
Notably, the microparticles accelerated the shift in expression, leading to specific differences already during early production which indicated an accelerated morphological development. The number of genes, altered in expression after 24 h in the microparticle culture (1094 genes, 14%), was almost three times higher than that in the control. Key morphology genes and regulators, including ssgA (SLIV_18635, 3.3-fold), ssgB (SLIV_30050, 2.0-fold), wblA (SLIV_20395, 2.8-fold), sigN (SLIV_18240, 2.5-fold), and bldN (SLIV_21180, 3.2-fold, among others, were exclusively activated after 24 h in the presence of talc (Tables 2 and 3). In addition, various native biosynthetic genes for secondary metabolites were upregulated by talc at this early stage ( Figure S6). During the main production after 48 h, the expression level of genes associated to morphology and secondary metabolism appeared rather similar with and without talc, although slight differences remained and totally more genes (38%, 2979) were modulated in the talc process. Interestingly, not even one single gene showed a significantly altered expression with talc supply during the growth phase ( Figure S4).  (Crone et al., 2016;Vior et al., 2020). The data reflect the production start (24 h) and the major production phase (48 h), and the control culture (without talc), set to 100%, is shown as dashed line (*p ≤ 0.05, **p ≤ 0.01). The intracellular amino acid pools reflect the growth phase (12 h, b) and the major production phase (48 h, c). The amino acids that are incorporated into mature bottromycins (L -glycine, L-proline, L-valine, L-phenylalanine, L-aspartate, and L-cysteine) are highlighted in blue. The intracellular amino acid pools during early production were not found significantly changed (data not shown) [Color figure can be viewed at wileyonlinelibrary.com] F I G U R E 5 Statistical evaluation of gene expression profiles of Streptomyces lividans TK24 DG2-Km-P41hyg + using PCA. Global transcription profiling of the cultures was conducted using RNA sequencing during growth (12 h) and bottromycin production (24 h, 48 h) in the presence of talc (10 g L −1 ) and without talc (control). For calculation of normalized read counts, the raw read count data were processed by DESeq. 2 (Love et al., 2014), including regularized log transformation (with blind dispersion estimation enabled). Subsequently, PCA was performed and visualized using ggplot2 (Wickham et al., 2016).  (Table 1), followed by culture harvesting, natural product extraction and HPLC-ESI-MS analysis. As shown, the production of the bottromycins and undecylprodigiosin by S. lividans TK24 DG2-Km-P41hyg + , as well as the production of pamamycins by S. albus J1074/R2 was significantly enhanced by talc ( Figure 7). Obviously, the miniaturized scale yielded the same picture as observed before for the 50 mL scale in shake flasks, although titers were somewhat lower, and it seemed appropriate to screen for the microparticle-based effects.
Subsequently, we investigated the production of different classes of natural products in a variety of Streptomyces species at the miniaturized scale including cinnamycins, another family of RiPPs (in addition to bottromycin), alpiniamides, another family of polyketides (in addition to pamamycins), the tetracycline derivative oxytetracycline, usabamycins of the anthramycin-type, the tetrahydroisochinoline-type perquinolines, and nybomycin, exhibiting a so far unique structure F I G U R E 6 Hierarchical cluster analysis of expression dynamics of bottromycin biosynthetic pathway genes in Streptomyces lividans TK24 DG2-Km-P41hyg+. Samples were taken from a control and a talc supplied culture (10 g L −1 ) during growth (12 h), production start (24 h), and major production phase (48 h). The expression level of the control during growth (12 h) was set as reference. The bottromycin cluster comprised the genes botA, encoding the precursor peptide; botP, leucyl-aminopeptidase; botC, YcaO domain protein; botRMT1, radical SAM; botRMT2, radical SAM; botRMT3, radical SAM; botCD, YcaO domain protein; botAH, aminohydrolase; botH, hydrolase; botCYP, CYP450 enzyme; botOMT, O-methyl transferase; botT, multidrug transporter; botR, transcriptional regulator (Huo et al., 2012). n = 3 [Color figure can be viewed at wileyonlinelibrary.com] T A B L E 2 Gene expression profiling of recombinant Streptomyces lividans TK24 DG2-Km-P41hyg + during growth (12 h) and bottromycin production (24 h and 48 h) in the presence of talc (10 g L −1 ) and without talc (control)  usabamycins. In each case, a production optimum could be identified.
Interestingly, the optimal talc concentration strongly differed for each strain and compound. Among all products, the highest increase was observed for the alkaloid undecylprodigiosin (thirteen-fold, at 20 g L −1 talc). For certain products, the microparticles not only influenced the overall yield, but specifically altered the spectrum of the formed derivatives. This resulted in notable changes, including enhancing and diminishing effects on derivatives that were minor side products in the control without talc. As example, usabamycin C production was more than doubled by the addition of 10 g L −1 talc, whereas the deoxycinnamycin level was reduced three-fold, when 20 g L −1 talc was present.

| Microparticles boost natural product formation across actinobacterial genera and families
Finally, the applicability of the microparticle approach was tested for other actinobacteria, outside of the genus of Streptomyces (Figure 7). Therefore, we examined the production of two glycopeptides T A B L E 3 Transcriptional changes of selected sigma factors in Streptomyces lividans TK24 DG2-Km-P41hyg + during growth (12 h) and bottromycin production (24 h and 48 h) in the presence of talc (10 g L −1 ) and without talc (control) in recombinant S. lividans TK24 DG2-Km-P41hyg + (Figure 2).
Generally, bottromycins are difficult to obtain and their research has developed into a "nightmare" over the past 65 years (Kazmaier, 2020). Various homologous as well as heterologous producers provide bottromycins only in the low milligram range and even below 1 mg L −1 (Crone et al., 2016;Horbal et al., 2018;Huo et al., 2012). Due to this, improvement of titer and yield is regarded crucial (Vior et al., 2020), especially as the pathway seems inefficient in laboratory conditions (Crone et al., 2016;Eyles et al., 2018). In this study, the refactored synthetic strain S. lividans TK24 DG2-Km-P41hyg + , apparently among the top bottromycin producers available (Franz et al., 2021), reached 109 mg L −1 bottromycins of the A2 type in the presence of 15 g L −1 talc, exceeding all previously reported efforts. This achievement will greatly support drug development and further biosynthetic studies of bottromycins and derivatives therefrom (Vior et al., 2020). It also demonstrates that the particle approach can be successfully used to boost the performance of even top-level microbial cell factories.

| Microparticle-enhanced production of bottromycins in recombinant S. lividans is supported by re-balanced expression of individual bottromycin cluster genes driven by a complex promoter architecture
The obtained genomic, transcriptomic, and metabolomic data sets were now integrated to understand the involved molecular processes on a systems level (Kohlstedt et al., 2014). Resequencing of the bottromycin cluster verified the previously designed architecture with a synthetic bidirectional promotor cassette in between the left and the right part of the bottromycin cluster .
While precursor availability on the amino acid level could be ruled out as a limiting factor (Figure 4b,c), the talc-induced increase of prebottromycins together with the simultaneous decrease of noncyclized shunt products from the upper pathway, indicated a higher efficiency of the posttranslational modification process (Figure 4a).
Notably, this was accompanied by a few significant changes in gene expression. Previous studies showed that the main limiting factor in bottromycin production, at least in the native host S. scabies, is the posttranslational maturation of the precursor peptide (Vior et al., 2020) and that the pathway inefficiently stalls at numerous biosynthetic steps (Crone et al., 2016). We conclude that the increased production efficiency in talc-supplied cultures largely resulted from the specifically increased expression of the aminohydrolase of the cluster (encoded by botAH, 2.4-fold, 48 h, Figure 6), responsible for the unique amidine-forming macrocyclodehydration of the molecule (Huo et al., 2012). Since this step follows after β-methylation of the valine, proline, and phenylalanine residues in the forming peptide, catalyzed by BotRMT1, BotRMT2, and BotRMT3 ( Figure 6), this link nicely explains that the titer of both bottromycin derivatives was increased simultaneously (Crone et al., 2016;Horbal et al., 2018).
It appears likely that also the lower abundance of botA, encoding for precursor peptide ( Figure 6) helped to balance biosynthesis, as previous attempts to increase precursor peptide levels were not found suitable to trigger bottromycin A2 synthesis, Vior et al., 2020). Taken together, fine-tuned re-balancing of the expression of individual cluster genes obviously mediated the improved biosynthesis. This picture differs largely from that obtained for talc-enhanced pamamycin production in S. albus, where the entire gene cluster was activated, up to 1000-fold . Likely, the difference originated from the different expression control strategies: the expression control of the bottromycin cluster in S. lividans was of synthetic nature , whereas the pamamycin cluster was expressed under its native control . Hereby, it appeared surprising that the bottromycin cluster was affected at all, as the synthetic design was expected to uncouple expression from the host metabolism. A close inspection of the RNA sequencing data revealed that the expression control was much more complex than expected ( Figure S5). It was only weakly driven by the synthetic promoters but largely relied on the highly active promoter of the hygromycin marker gene hygR, inserted in between of them, and transcribing the right cluster genes ( Figure S5b). In addition, a so far unknown but apparently even stronger antisense promoter sequence (P AS ) downstream of hygR, transcribed the cluster in the opposite direction and expressed the left cluster genes, plus an additional antisense transcript of hygR ( Figure S5c). These previously not considered promoters created a F I G U R E 7 Impact of talc microparticles on nine major classes of natural products across different actinobacterial families. RiPP-type cinnamycins using S. albus pCinCatInt (Lopatniuk et al., 2017), RiPP-type bottromycins using S. lividans TK24 DG2-Km-P41hyg + , alkaloid-type undecylprodigiosin using S. lividans TK24 DG2-Km-P41hyg + , angucyclinone-type simocyclinones using K. sp. (Bilyk et al., 2016), polyketide-type pamamycins using S. albus J1074/R2 (Rebets et al., 2015), polyketide-type alpiniamides using S. sp. IB2014/011-12 (Paulus et al., 2018), glycopeptide-type vancomycin using A. japonicum DSM 44213 (Stegmann et al., 2014), glycopeptidetype teicoplanin using A. teichomyceticus ATCC 31121 (Horbal et al., 2012), nybomycin using S. albus subsp. chlorinus NRRL B-24108 (Rodriguez Estevez et al., 2018), tetracycline-type oxytetracycline using S. rimosus ATCC 10970 (Pethick et al., 2013), tetrahydroisochinoline-type perquinolines using S. sp. IB2014/016-6 (Rebets et al., 2019), and anthramycin-type usabamycins using S. albus subsp. chlorinus NRRL B-24108 (nybomycin) (Rodriguez Estevez et al., 2018). All strains were incubated over 5 days using miniaturized microtiter plate cultures (1 ml) at different talc levels between 0 and 50 g L −1 . The natural product levels were determined after solvent extraction using HPLC-ESI-MS and reflect final titers after 5 days. The colored circles given beside strain names indicate the type of producers studied: native (blue circle) and heterologous strains (orange circle). The colored squares denote the underlying control type for cluster expression: native (blue square) and synthetic regulation (orange square). n = 3 [Color figure can be viewed at wileyonlinelibrary.com] complex control architecture and could play a major role in mediating the talc-effect.

| Microparticles affect submerged culture morphology and accelerate morphological development and aging of S. lividans and S. albus
The insights into the impact of talc microparticles on morphological development (Figure 4) and the associated cellular program of S. lividans (Figures 5, 6, S4, and S6) and the picture recently obtained for the talc effects on the related strain S. albus , now allowed to draw a few important conclusions related to morphology and its control in these type of bacteria.
As shown, talc reduced the pellet size of both strains (Figure 3a) , and this response can be expected also in other actinobacteria (Ren et al., 2015), although exceptions cannot be excluded. From the bioengineering viewpoint, smaller pellets appear favorable, as they reduce problems with mass and oxygen transfer limitation, slow growth, and culture heterogeneity (Mehmood et al., 2012;Tamura et al., 1997;van Dissel & van Wezel, 2018;van Dissel et al., 2014).
The strength of the talc-effect was quite different for the two strains. While 10 g L −1 of the talc material reduced the pellet size of S. albus more than six-fold , the same amount of talc did much less to S. lividans, as its pellets still maintained 60% of the size of the control culture. The stronger resistance of S. lividans might be related to its more dense and compact aggerates, offering less attack points for fragmentation and segregation than the more open and loose structures of S. albus (Zacchetti et al., 2018). In addition, differences in cell wall composition, mechanical strength, and hydrophobicity of spores, hyphae, and mycelia could explain the different accessibility of these bacteria to talc-mediated morphology engineering and might be interesting to study further (Driouch, Roth, Dersch, et al., 2010a, 2010bDriouch et al., 2010;Driouch et al., 2012;Walisko et al., 2012).
Notably, the microparticles accelerated the morphological development of the studied Streptomyces in submerged culture.
S. lividans revealed a premature activation of regulator genes, such as sigN, bldN, ssgA, ssgB, and wblA (Tables 2 and 3), important control elements of morphology development in Streptomyces (Kang et al., 2007;Rebets et al., 2018;Traag & van Wezel, 2008). The same picture of accelerated aging was recently observed for talc-treated S. albus , although the microparticles affected the expression of a much higher number of genes in this microbe (56% of all genes) .
S. lividans, in contrast to S. albus, does not sporulate in liquid culture (Daza et al., 1989;Rebets et al., 2018). Differences in morphology control and development on the species level are likely the reason for the observed individual differences.
However, the overall response to the microparticle addition apparently was the same for both strains: accelerated morphology development and aging and it appears likely that other actinobacteria will behave in a similar way.

| The use of microparticles displays a powerful approach to support natural product research in actinobacteria
As shown, microparticle-based cultures revealed a notably increased performance in producing natural products. First, the addition of talc beneficially increased the formation of 75% of the tested natural products, covering seven important classes with a highly diverse structure: polyketides, glycopeptides, RiPPs, alkaloids, tetracyclines, anthramycin-analogs, and angucyclinone-analogs with different structure (Figures 1 and 7). Stimulating production was observed for molecules of clinical relevance as antibiotics (vancomycin, oxytetracycline, and teicoplanin) (Elsayed et al., 2015;Fung et al., 2012;Liu et al., 2018) for several high-potential candidates presently under development into antibiotics (bottromycin, cinnamycin, pamamycin, and simocyclinone) (Buttner et al., 2017;Kobayashi et al., 2010;Lefèvre et al., 2004;Lopatniuk et al., 2017) and antitumor drugs (undecylprodigiosin and usabamycin) (Sato et al., 2011;Stankovic et al., 2014). Second, the microparticle-approach worked on actino-  Figure 7). Fourth, the approach is simple. All what is needed is to add a little bit of talcum powder into the medium. Although the optimum talc level differed between strains and products, a concentration of 10 g L −1 always revealed a significant effect. Altogether, microparticle-enhanced cultivation provides a valuable concept for natural product research in actinobacteria, complementing efforts to streamline eukaryotic filamentous fungi (Böl et al., 2020;Veiter et al., 2018).

| CONCLUSIONS
As shown, talc-supplementation boosted production of seven classes of commercially relevant natural products across strains from different actinobacterial families, including natural as well as heterologous producers. The improvement was up to 100-fold. Obviously, the talc-effect generally accelerated morphological development and aging, inter alia leading to enhanced expression  and re-balanced expression (this study) of the bacterial gene clusters of interest. It is well known that antibiotic production in Streptomyces correlates to morphological development (Chater, 1984) and certain natural products are even KUHL ET AL.
| 3089 part of programmed cell death (PCD) (Tenconi et al., 2018) so that the microparticles obviously trigger a sweet spot towards enhanced performance.
Without doubt, the microparticle concept appears useful to be systematically applied to the field of natural product research, where it offers several striking opportunities. Its successful application at the small scale shows potential for screening efforts to support research on major questions: discovery of novel molecules, activation of silent gene clusters, supply of sufficient amounts for structure elucidation, and exploration of biosynthetic control mechanisms, among others. The applicability to the microliter and microtiter plate scale provides a nice extension to recent developments on miniaturized physiological characterization of Streptomyces (Koepff et al., 2017). It appears promising to rescreen some of the libraries of isolates available worldwide, adding a bit of talc (Barka et al., 2016;Landwehr et al., 2016;Silva et al., 2020;Steele et al., 2019).
Regarding industrial manufacturing of natural products at large scale, talc could provide a straightforward drop-in solution. It is available in huge amounts, approximately 8 million tons per year according to recent market reviews. The material is cheap, in the range of 10 EUR cents per kg. As an example, supplementation of a 100 m 3 production process with 10 g L −1 talc (found effective) would add only costs in the range of 100 EUR. Moreover, co-harvested talc could upgrade fermentation biomass for post-process valorization, given its excellent performance as coating and baking agent in fertilizer formulations (Kaji et al., 2017) and its proven safety and efficacy as feed additive to all animal species (Mallet et al., 2005;Rychen et al., 2018).
Furthermore, it appears interesting to explore the interaction between microparticles and cells in more detail. On the particle side, talc prevented spore agglomeration during the initial culture phase of the fungus Aspergillus niger, whereas chemical (nutritional) effects by leached talc minerals played no significant role for the altered morphogenesis (Driouch et al., 2010). Moreover, particle size and shape were found critical. Even a small increase in particle diameter from 6 to 15 µm resulted in a severe loss of the morphology effects, and lamellar shaped talc particles were found more efficient to tailor morphology than round shaped materials (Driouch, 2009). Given the various sizes and shapes of talc, achievable from ball, rod, and autogenous mills, more systematic studies appear feasible. So far not explored but interesting seems the importance of the material softness, as talc is the softest known mineral.
Hereby, several other micro materials such as aluminum oxide (Driouch et al., 2010) and titanium silicate oxide (Driouch et al., 2012) await to be tested. On the cellular side, an inspection of the impact of cell wall composition, mechanical strength, and hydrophobicity of spores, hyphae, and mycelia seems relevant, as stated above. Moreover, more research will be needed to clarify the talc-effects in stirred tanks and at large scale, exhibiting different hydrodynamics and mixing regimes (Bliatsiou et al., 2020;Kowalska et al., 2020).

CONFLICT OF INTERESTS
The authors declare that there are no conflict of interests.

DATA AVAILABILITY STATEMENT
The data that supports the findings of this study are available in the supplementary material of this article.