Microparticles globally reprogram Streptomyces albus toward accelerated morphogenesis, streamlined carbon core metabolism, and enhanced production of the antituberculosis polyketide pamamycin

Streptomyces spp . are a rich source for natural products with recognized industrial value, explaining the high interest to improve and streamline the performance of in these microbes. Here, we studied the production of pamamycins, macrodiolide homologs with a high activity against multiresistant pathogenic microbes, using recombinant Streptomyces albus J1074/R2. Talc particles (hydrous magnesium silicate, 3MgO·4SiO 2 ·H 2 O) of micrometer size, added to submerged cultures of the recombinant strain, tripled pamamycin production up to 50 mg/L. Furthermore, they strongly affected morphology, reduced the size of S. albus in response to talc appears of general relevance to further explore and upgrade the concept of microparticle enhanced cultivation, widely used for filamentous microbes.

It is easy to understand that strategies to activate and enhance the synthesis of natural products in Streptomycetes have been of a broad interest from early on and still display a topic of major relevance (Ahmed et al., 2020;Horbal, Marques, Nadmid, Mendes, & Luzhetskyy, 2018;Kallifidas, Jiang, Ding, & Luesch, 2018;Lopatniuk et al., 2019;Zhang et al., 2020).
Members of the genus are well known for a complex morphology linked to their multicellular life cycle, which starts with the germination of a single spore that grows into a vegetative mycelium by linear tip extension and hyphae branching (Chater & Losick, 1997; van Dissel, Claessen, & van Wezel, 2014), then forms an aerial mycelium, and finally differentiates into uninucleoid cells that further develop again into spores (Angert, 2005). In submerged culture, more relevant for industrial production, morphogenesis comprises primary and secondary mycelial networks, pellets, and sporulation (van Dissel et al., 2014).
Notably, morphological development and natural product formation are closely linked, and various efforts have been made to increase production through an altered morphology (Chater, 1984).
Here, we studied the use of talc microparticles for the production of pamamycins (Figure 1), a family of 16 macrodiolide homologs that are highly active against multiresistant pathogenic microbes, using recombinant Streptomyces albus J1074/R2 (Rebets et al., 2015). Carefully conducted cultures with analysis of growth, product formation and cellular morphology enabled us to specifically study the impact of the microparticle addition on production performance. In addition, transcriptome and intracellular CoA thioester analyses provided insights into the cellular response of S. albus and provided a systems-level picture on how the particles reprogrammed morphogenesis and streamlined metabolism for enhanced production and a notable shift toward heavier pamamycin homologs.

| Strain
S. albus J1074/R2 expressing the heterologous pamamycin gene cluster was obtained from previous work (Rebets et al., 2015). For strain maintenance, spores collected from agar plate cultures after 5-day incubation were resuspended in 20% glycerol and kept at −80°C.

| Cultivation
One loop of spores was scratched from a 5-day old plate culture and used to inoculate a liquid preculture, which was then grown overnight in a 500-ml baffled shake flask with 50 ml medium and 30 g soda-lime glass beads (5 mm; Sigma-Aldrich). When the preculture reached the late exponential phase, an appropriate amount of cells was collected (8,500g, room temperature, 5 min), resuspended in 10 ml fresh medium, and used to inoculate the main-culture (50 ml medium in 500-ml baffled shake flasks). Main cultures (with and without talc) were inoculated from the same preculture to enable identical starting conditions. All cultivation experiments were conducted in triplicate on a rotary shaker (28°C, 230 rpm, 75% relative humidity, 5-cm shaking diameter, Multitron, Infors AG, Bottmingen, Switzerland).

| Quantification of substrates
Before analysis, starch was hydrolyzed to glucose monomers for 3 hr using 3 M HCl at 100°C. Glucose and glycerol were quantified by high-performance liquid chromatography (HPLC; 1260 Infinity Series; Agilent, Waldbronn, Germany) using an Aminex HPX-87H column (300 × 7.8 mm; Bio-Rad, München, Germany) and 7 mM H 2 SO 4 as mobile phase (55°C, 0.7 ml/min). Refraction index measurement was used for detection and external standards were used for quantification. Phosphate was analyzed by HPIC (Dionex Integrion; Thermo Fisher Scientific, Karlsruhe, Germany) using a Dionex IonPac AS9-HC column (2 × 250 mm; Thermo Fisher Scientific) and 9 mM Na 2 CO 3 as F I G U R E 1 Chemical structure of the polyketide pamamycin family and the building blocks of its different derivatives succinyl-CoA, malonyl-CoA, methylmalonyl-CoA, and ethylmalonyl-CoA. The individual combination of different CoA-esters results in a different decoration of the product (-H, -CH 3 , -C 2 H 5 ) at specific positions and determines the molecular weight (MW). Except for Pam 607, several isomers exist for each pamamycin, differing by the exact position of the side chains (Hanquet, Salom-Roig, & Lanners, 2016). For every pamamycin mass derivative, one exemplary structure is shown [Color figure can be viewed at wileyonlinelibrary.com] mobile phase (35°C, 0.25 ml/min). Conductivity measurement was used for detection and an external standard was used for quantification. All measurements were performed in triplicate.

| Morphology analysis
Five microliter culture broth was transferred onto a glass for brightfield microscopy (Olympus IX70 microscope, Hamburg, Germany).
The software ImageJ 1.52 (Schneider, Rasband, & Eliceiri, 2012) was used to automatically determine the size of pellets formed during growth (Krull et al., 2013). The diameter of a pellet was assumed as the smallest circle into which the complete aggregate fitted (Martin & Bushell, 1996). At least 150 aggregates were analyzed per sample.

| Natural compound extraction and quantification
Pamamycin was extracted from culture broth using a two-step process. First, 200 µl broth was mixed with 200 µl acetone and incubated for 15 min (1,000 rpm, room temperature, Thermomixer F1.5; Eppendorf, Wesseling, Germany). Subsequently, 200 µl ethyl acetate was added, and the mixture was incubated for further 15 min under the same conditions. Afterward, the organic phase was collected (20,000g, room temperature, 5 min). The solvents were evaporated under a laminar nitrogen stream. The extract was redissolved in 2 ml methanol, clarified from debris (20,000g, 4°C, 10 min) and analyzed,

| Extraction and quantification of intracellular CoA-esters
The analysis of CoA-esters was conducted as recently described (Gläser et al., 2020). In short, a broth sample (8 mg CDW) was transferred into a precooled tube, which contained quenching and extraction solution (95% acetonitrile, 25 mM formic acid, −20°C) at a volume ratio of 1:4 followed by repetitive mixing and cooling on ice for 10 min, clarification from cell debris (15,000g, 4°C, 10 min) and the addition of 10 ml supercooled deionized water. The cell pellet was washed twice with 8 ml supercooled deionized water. All supernatants were combined, followed by freezing in liquid nitrogen and lyophilization. Before analysis, the obtained dry extract was dissolved in 500 µl precold buffer (25 mM ammonium formate, 2% methanol, pH 3.0, 4°C) and filtered (Ultrafree-MC 0.22 µm; Merck, Millipore, Germany). Analysis of the CoA-esters was performed on a triple quadrupole MS (QTRAP 6500+; AB Sciex, Darmstadt, Germany) coupled to an HPLC system (Agilent Infinity 1290 System). Separation of the analytes of interest was conducted at 40°C on a reversed phase column (Gemini 100 × 4.6 mm, 3 µm, 110 Å, Phenomenex, Aschaffenburg, Germany) using a gradient of formic acid (50 mM, adjusted to pH 8.1 with 25% ammonium hydroxide, eluent A) and methanol (eluent B) at a flow rate of 600 µl/min. The fraction of eluent B was as follows: 0-12 min, 0-15%; 12-16 min, 15-100%; 16-18 min, 100%; 18-20 min, 100-0%; 20-25 min, 0%. The first 3 min of the analysis were discharged to minimize the entry of salts into the mass spectrometer. CoA-esters of interest were analyzed in positive ionization mode, using multiple reaction monitoring. Analyte specific instrument settings such as declustering potential, collision energy, and collision cell exit potential were individually optimized for each CoA-ester, using synthetic standards. All measurements were done in triplicate. albus J1074/R2 genome sequence (CP059254.1) with Bowtie2 using standard settings (Langmead & Salzberg, 2012) except for increasing the maximal allowed distance for paired reads to 600 bases. For visualization of read alignments and raw read count calculation, ReadXplorer 2.2.3 was used (Hilker et al., 2014). Due to a high unspecific background over both strands, the raw read count for each CDS was corrected by subtracting the length-adjusted median read count calculated over all CDS from the respective noncoding strand.

| Transcriptome analysis
Using the resulting data, DESeq2 (Love, Huber, & Anders, 2014) was used to QC the datasets via, among others, calculation of the sample to sample distances ( Figure S9) and PCA ( Figure S10). In addition, DESeq2 was used to calculate DGE datasets. Raw datasets (sequenced reads) as well as processed datasets (input matrix and normalized read counts from DESeq2) are available from GEO (GSE155008). For statistical analysis, Student's t test was carried out KUHL ET AL.
| 3861 and the data were filtered for genes with a log 2 -fold change ≥ 1 (p ≤ 0.05). Hierarchical clustering was conducted, using the software package gplots (R Core Team, 2014;Warnes, Bolker, Bonebakker, & Gentleman, 2016). For visualization, Voronoi tree maps were created, using the Voronto tool (Santamaría & Pierre, 2012) for Java (version 8, update 231, Build 1.8.0_231-b11). RNA extraction and sequencing were conducted as biological triplicates, except for one of the controls, where one replicate was lost during processing. Given the excellent reproducibility of all analyzed samples (Figures S10 and S11), the available dataset was regarded acceptable to enable a robust and reliable evaluation of gene expression.

| Pamamycin production in S. albus J1074/R2
In a first set of experiments, the pamamycin production performance of S. albus J1074/R2 was assessed in liquid SGG medium, which contained starch and glycerol as carbon source ( Figure 2a). After inoculation, cells immediately started to grow into mycelial networks, which then aggregated into pellets as typically observed for actinomycetes. Growth lasted for about 12 hr. The production of pamamycins started after ∼9 hr at the end of the growth phase (when phosphate became limiting, Figure S11) and continued during the later stationary phase. The recombinant strain produced a rich spectrum of different pamamycins, which were attributed to derivatives with different side chains according to their molecular mass, that is, Pam 579, Pam 593, Pam 607, Pam 621, Pam 635, Pam 649, and Pam 663. Smaller pamamycins (Pam 579, Pam 593, and Pam 607) were most prominent. The total pamamycin titer was 18 mg/L. Interestingly, starch was the major carbon source until ∼24 hr ( Figure S11). Glycerol remained practically untouched during the initial process, but was consumed later when starch reached a lower level (although it was still present). During the cultivation, the pH value varied between 6.5 and 7.5. It decreased during the growth phase and increased again in later phases.
3.2 | The addition of talc to the culture of S. albus J1074/R2 increases the production of pamamycin up to threefold S. albus J1074/R2, cultivated in SGG medium with microparticles (2.5 g/L talc), revealed an increased pamamycin titer of 22 mg/L F I G U R E 2 Impact of talc microparticles on growth and pamamycin production in Streptomyces albus J1074/R2 using complex SGG medium with starch and glycerol as main carbon source. (a) Control culture without microparticles. (b) Microparticle-enhanced culture with 2.5 g/L talc. Growth phase (white) and major pamamycin production phase (gray) are indicated by color [Color figure can be viewed at wileyonlinelibrary.com] ( Figure 2b). Interestingly, talc did not generally enhance production, but specifically affected the pamamycin spectrum. The titer of low molecular weight derivatives (Pam 579, Pam 593, Pam 607) was found reduced, whereas higher molecular weight pamamycins were increased (Pam 635, Pam 649, Pam 663). This effect was most prominent for the two heaviest pamamycins (Pam 649, Pam 663). These two derivatives were increased almost twofold. The presence of the talc particles resulted in a slightly faster use of phosphate ( Figure   S11). The mode of substrate utilization was the same as for the control, mainly starch consumption during the initial phase and activation of glycerol utilization after ∼20 hr. Generally, glycerol uptake was more pronounced than in the control. It's uptake was faster, and a significantly lower amount of it was left at the end of the process ( Figure S11). The pH profile was like the control.
Further studies revealed a strong impact of the amount of talc on the product level ( Figure 3a). An optimum performance was observed for talc levels of 10 and 15 g/L. These concentrations provided 50 mg/L of total pamamycin. An even higher concentration of talc (20 g/L) resulted in a reduced titer (37 mg/L), slightly below the optimum. Notably, the stimulating effect of the microparticles on the formation of larger pamamycins was maintained even at the highest talc level: 20 g/L of talc specifically enhanced production of pamamycins Pam 649 and Pam 663. Altogether, a concentration of 10 g/L talc appeared optimal and was chosen for further studies.
3.3 | Talc microparticles reduce the pellet size of S. albus J1074/R2 more than sixfold The addition of talc caused substantial changes in cellular morphology ( Figure 3b). In control cultures without talc, the formed pellets exhibited an average diameter of ∼435 µm. Already low levels of talc (2.5 g/L) led to a drastic decrease to 150 µm. With an increasing microparticle concentration, this effect was even more pronounced.
Further increase of the talc concentration did not result in a further reduction of the pellet size.
Taken together, smaller pellets were obviously beneficial for pamamycin production (Figure 3a). Microscopic analysis of the cultures revealed that the microparticles attached to the cells, which obviously loosened the inner structure of the aggregates (Figure 4c,d). The pellets of talc enhanced cultures appeared of a similar loose structure during growth and production phase. In contrast, the central core of pellets of the control culture showed signs of decomposition during production phase (Figure 4b).
3.4 | Microparticles globally reprogram the metabolism of S. albus J1074/R2 To assess the response of the actinomycete to the added microparticles in more detail, global gene expression analysis was conducted. We compared the transcriptome during the growth phase (5 hr) and the production phase (21 hr) between a talc (10 g/L) supplied culture and a control culture without talc using RNA sequencing. Generally, the transition of S. albus J1074/R2 from the nonproducing growth phase to the production phase was linked to a wide readjustment of gene expression. The expression of 1,468 genes, representing 24% of the genomic repertoire, was significantly altered in the control culture, when cells shifted from growth to production mode (log 2 -fold change ≥ 1, p ≤ .05; Figures S1 and S7). On top of this general shift, talc supply induced a global change in the transcriptome. These talc-specific effects were observed for the growth as well as the production phase (Figures 5, S1, and S6). Altogether, 3,341 genes (56% of all genes) were specifically affected by the presence of talc, revealing a fundamental impact of the microparticles on the physiology of S. albus. During growth, the microparticles changed the expression of 2,133 genes (36%). This number  (Table S2).

| Talc microparticles affect the expression of morphology regulators
Since the microparticles obviously affected the morphology of S. albus (Figures 3b and 4), we searched within the transcriptome data for genes involved in morphology and secondary metabolism. Based on their similarity to known morphogenetic genes identified in other Streptomyces, we could identify 55 genes which were affected by the particles (Figures S2 and S3). As prominent example, the addition of microparticles resulted in an upregulation (log 2 -fold change = 2.0) of the sporulation and cell division protein SsgA, encoded by XNRR2_5315, already during growth (Table 1). The upregulation was even higher during production (log 2 -fold change = 3.0). A similar picture was observed for other prominent morphology genes, that is, the signaling protein Factor C (XNRR2_2306), the chaplin (chp) and rodlin (rdl) hydrophobic sheath proteins and a neutral zinc metalloprotease (XNRR2_1391), a homolog to sgmA in S. griseus, and a wellknown morphology regulator (Table 1). In all cases, the microparticles caused an overexpression, which was most pronounced during the pamamycin production phase but partly started already during growth.

| Microparticles drive the expression of the pamamycin biosynthetic cluster
Causally linked to the enhanced production, S. albus J1074/R2 responded to the microparticles by a strong overexpression of the pamamycin cluster (log 2 -fold change up to 10, Figure 6). The activation was most pronounced for the production phase, where 19 out F I G U R E 4 Impact of talc microparticles on the cellular morphology of Streptomyces albus J1074/R2. Control culture without microparticles during growth (5 hr) (a) and production (20 hr) (b). Microparticle-enhanced culture with 10 g/L talc during growth (5 hr) (c) and production (20 hr) (d). The arrow indicates the physical attachment of the mycelium to the microparticles [Color figure can be viewed at wileyonlinelibrary.com] 3864 | F I G U R E 5 Global gene expression pattern for pamamycin producing Streptomyces albus J1074/R2 at different stages of the cultivation and under the impact of talc microparticles (10 g/L). Kyoto Encyclopedia of Genes and Genomes (KEGG)-orthology tree maps for the average gene expression of different functional classes of growing (a) and producing (b) S. albus J1074/R2 as compared to the control culture during as a reference. Each cell represents a functional class of the orthology: Carbohydrate metabolism, amino acid metabolism, lipid metabolism, nucleotide metabolism, energy metabolism, glycan biosynthesis and metabolism, xenobiotics biodegradation and metabolism, metabolism of cofactors and vitamins, metabolism of terpenoids and polyketides, biosynthesis of other secondary metabolites, metabolism of other amino acids, signal transduction, membrane transport, replication and repair, transcription, translation, folding, sorting and degradation, cellular community, and cell motility with their subclasses. The annotation was taken from the genome (entry T02545 in KEGG) of S. albus (Zaburannyi et al., 2014). In addition, the right and left core of the pamamycin gene cluster are shown [Color figure can be viewed at wileyonlinelibrary.com] of the 20 cluster genes were overexpressed as compared to the control. As exception, only the regulator gene pamR1 remained relatively unaffected. It was interesting to note that the cluster activation was not fully balanced among the "two cores" of the gene cluster. In fact, genes of the "right core" were induced stronger than genes of the "left core." Most of the "right core" genes, that is, 11 out of 15, were upregulated with a log 2 -fold change of more than 8 by talc. A few genes of the pamamycin cluster (pamC, pamD, pamE, pamF, pamH, pamK, pamO, pamS, and pamX) were activated by the microparticles already during the growth phase. In contrast, the pamamycin transporter gene (pamW) and the regulator inside the "left core" (pamR2) were downregulated. In addition to the pamamycin cluster, also other genes of secondary metabolism were found significantly upregulated ( Figure 5). The expression changes covered different gene functions: Type I polyketide structures (candicidin biosynthesis), polyketide sugar unit biosynthesis, siderophores (2,3dihydroxybenzoate synthesis), sesquiterpenoid and triterpernoid biosynthesis (germacradienol/geosmin synthase), terpenoid backbone synthesis, and phenazine biosynthesis (Table S3).

| Microparticles modulate supporting pathways in central carbon metabolism
As shown, cells preferably produced higher mass pamamycin derivatives in the presence of microparticles. Generally, different derivatives originate from the incorporation of different precursor metabolites, that is, CoA-esters of different type (Figure 1). We, therefore, hypothesized that the microparticles could have impacted genes encoding enzymes for CoA-ester synthesis and interconversion. Indeed, talc supply affected the expression of CoA-ester related genes (Table 2; Figures S4 and S5). As example, talc led to a strong downregulation (log 2 -fold change = 6) of the α-subunit of the acetyl/propionyl-CoA carboxylase (XNRR2_4211), responsible for the formation of methylmalonyl-CoA T A B L E 1 Impact of talc microparticles on the expression of morphology associated genes in pamamycin producing Streptomyces albus J1074/R2 Note: Samples were taken from a control and a talc supplied culture (10 g/L) in SGG medium during growth (5 hr) and production (21 hr). The values correspond to log 2 -fold expression changes, considering the control culture during growth (5 hr) as reference. The listed genes represent previously discovered morphology-associated genes in S. albus (SALB) and genes, identified by BLAST search as homologs to morphology-associated genes in S. coelicolor (SCO), S. griseus (SGR), and S. venezuelae (SVE), indicated by the percentage of homology The identification of homologs was supported by the fact that Streptomycetes share many genetic elements of morphology control (van Dissel et al., 2014), including SsgA like proteins (Traag & van Wezel, 2008) and Factor C (Birkó et al., 1999). In addition to genes identified in S. albus before (Zaburannyi, Rabyk, Ostash, Fedorenko, & Luzhetskyy, 2014), further candidates could be inferred from previous studies on morphological development in other Streptomyces spp., including S. coelicolor (Chakraburtty & Bibb, 1997;Hesketh et al., 2007;van Wezel & McDowall, 2011), S. griseus (Saito et al., 2006;McDowall, 2011), andS. lividans (van Wezel et al., 2006) 3.8 | Microparticles affect intracellular CoA-ester pools during pamamycin production in S. albus J1074/R2 Due to the significant transcriptomic changes around the pamamycin cluster and the genes of its precursor metabolism, it appeared interesting to assess the availability of the product precursors during the production process. We focused our analysis on 11 CoA-esters that are either directly incorporated into pamamycins, that is, succinyl-CoA, malonyl-CoA, methylmalonyl-CoA, and ethylmalonyl-CoA (Figure 1), or are connected to these building blocks. For this purpose, we acquired CoA-ester levels in a talc supplied (10 g/L) and a control culture during the major phase of production (20 hr). Interestingly, the microparticles strongly affected the CoA-ester pools (Figure 7).
F I G U R E 6 Hierarchical cluster analysis of the expression of the pamamycin biosynthetic pathway genes in Streptomyces albus J1074/R2. Samples were taken from a control and a talc supplied culture (10 g/L) in SGG medium during exponential growth (5 hr) and production phase (21 hr). The expression level of the control culture during growth (5 hr) was used as a reference. The cluster comprises the genes pamA, 3-oxoacyl-synthase 2; pamB, 3-oxoadipate CoA-transferase subunit A; pamC, acyl carrier protein; pamD, 3-oxoacyl-synthase 3 protein 1; pamE, 3-oxoacyl-synthase 3; pamF, 3-oxoacyl-synthase 2; pamG, 3-oxoacyl-synthase 3; pamH, aminohydrolase; pamJ, 3-oxoacyl-synthase 2; pamK, 3-oxoacyl-synthase 3; pamL, putative sulfoacetate-CoA ligase; pamM, 3-oxoacyl-reductase FabG; pamN, putative oxidoreductase; pamO, 3-oxoacyl-reductase FabG; pamR1, response regulator protein VraR; pamR2, tetracycline repressor protein class E; pamS, carnityl-CoA dehydratase; pamW, antiseptic resistance protein; pamX, L-lysine-8-amino-7-oxononanoate aminotransferase; pamY, cypemycin methyltransferase (Rebets et al., 2015) [Color figure can be viewed at wileyonlinelibrary.com] 4 | DISCUSSION 4.1 | Microparticle-enhanced production supports future exploration of pamamycins as lead molecules for novel antituberculosis drugs As shown in this study, talc microparticles increased the production of pamamycins in recombinant S. albus J1074/R2 almost threefold to 50 mg/L, the highest titer observed so far for these polyketides (Figure 3). In this regard, the strongly improved pamamycin production by addition of talc is supposed to facilitate future exploration of this important polyketide. As shown, only small amounts of talc were required to achieve the stimulating effect (Figures 2 and 3), so that a use of this cheap material for pamamycin production appears feasible also from a cost perspective. Remarkably, the addition of talc selectively triggered the formation of larger variants (Figure 3a), that is, Pam 635, Pam 649, and Pam 663. Due to the fact that the different derivatives apparently differ in biological activity (Lefèvre et al., 2004;Natsume, 2016), a microparticle-based process might help to selectively enrich heavier pamamycin variants in the total product spectrum with potentially other activities. Note: Samples were taken from a control and a talc supplied culture (10 g/L) in SGG medium during growth (5 hr) and production (21 hr). The values correspond to log 2 -fold expression changes, considering the control during growth (5 hr) as reference. The annotation was taken from the genome (entry T02545 in KEGG) of S. albus (Zaburannyi et al., 2014). XNRR2_2306 (Factor C protein), stimulating sporulation in submerged culture (Birkó et al., 1999;van Wezel & McDowall, 2011); (e) hydrophobic coat proteins such as chaplins and rodlins, which form the surface rodlet layer on spores (Claessen et al., 2004); and (f) XNRR2_3527 (sigma factor BldN) controlling their expression in S. venezuelae (Bibb et al., 2012) and likely in S. coelicolor (McCormick & Flärdh, 2012; Figure 8; Table 1). There seems no doubt that this morphological development during the culture triggered the pamamycin formation (van Wezel & McDowall, 2011). It was interesting to note that the addition of talc influenced the utilization of nutrients, including phosphate and glycerol ( Figure S8) and furthermore affected the amount of biomass formed (Figure 2). We cannot provide a clear conclusion on the underlying effects at this point but would like to notice that the understanding of these effects remains an important question.

| Microparticles accelerate the morphogenesis of S. albus toward sporulation-oriented mycelial development and cell division
Notably, the morphogenesis program was massively upregulated in the presence of the talc particles ( Figure 8; Table 1). The activation was already visible during the initial growth phase and was even stronger during the later production phase. It was observed for practically all genes, which were identified as part of the morphology control cascade. We therefore conclude that the microparticles significantly speeded up the aging of the S. albus culture and accelerated the shift to second mycelium formation and submerged sporulation response. It appears highly likely that the enhanced pamamycin formation, including a strong overexpression of the pam cluster itself, was a consequence of the accelerated morphogenesis program, induced by the microparticles. It is well known that natural product formation is linked to morphological differentiation in Streptomycetes (Chater, 1984), so that its perturbation can be efficiently exploited to influence antibiotic production in S. coelicolor, S. lividans, and other species (Chakraburtty & Bibb, 1997;Hesketh et al., 2007;van Wezel & McDowall, 2011). Admittedly, the data did not allow to identify the specific link between morphology development, pamamycin cluster expression, and altered expression of its two regulators ( Figure 6). More work will be needed in the future to resolve this in greater detail. S. albus possesses a range of different (mainly uncharacterized) ECF sigma factors and regulators, of which the majority was found affected by talc and gives a flavor of the complexity to be explored (Table S1).
4.4 | Talc microparticles orchestrate the regulatory and metabolic network of S. albus to a highly efficient program for pamamycin production We now mapped the obtained transcriptomic and metabolomic data on the carbon core network of S. albus to obtain a systemsview on pamamycin production and supporting pathways ( Figure 9). First, specific adjustments were observed in central carbon metabolism. Genes, related to conversion of glycerol were generally upregulated during the major production phase, linked to the on-set of glycerol consumption at this point ( Figure S8).
This substrate shift could also explain the increased expression of the EMP pathway genes alongside the downregulation of the oxidative PP pathway at the level of 6-phosphogluconate dehydrogenase. The effects were slightly more pronounced in the presence of talc. Second, most genes encoding for enzymes of CoA-ester metabolism were upregulated by the microparticles as F I G U R E 7 Relative changes of intracellular CoA-ester pools in pamamycin producing Streptomyces albus J1074/R2 by the addition of talc microparticles (10 g/L) to the culture during the major production phase (21 hr). The intracellular availability of 11 CoA-esters, directly and indirectly linked to pamamycin assembly was assessed during the production phase (21 hr We conclude that this was the consequence of an increased availability of the larger building block methylmalonyl-CoA, together with a reduction in the malonyl-CoA pool. It would be interesting to further explore this link in other natural producers, which obviously differ in the spectrum of pamamycin homologues (Natsume, Yasui, Kondo, & Marumo, 1991;Natsume et al., 1995;Kozone, Chikamoto, Abe, & Natsume, 1999). Metabolic engineering of CoA-ester supply appears promising to streamline pamamycin production towards selective derivatives (Lu, Zhang, Jiang, & Bai, 2016).
F I G U R E 8 Impact of talc microparticles on the morphogenesis of pamamycin producing Streptomyces albus J1074/R2 in submerged culture. The data reflect differential gene expression of morphology-associated genes during the pamamycin production phase (21) in the presence of 10 g/L talc (ellipse) and the control process without particles (rectangle). The gene expression during the growth phase (5 hr) of the control is used as a reference. The numbers denote log 2 -fold expression change. Further details on the displayed components of the morphology cascade are given in  (Antecka, Bizukojc, & Ledakowicz, 2016). In this regard, our insights appear of general value for further exploration and industrialization of microparticle-enhanced processes. The use of microparticles, furthermore promises to support strain engineering by suggesting novel genetic targets, given the rich response on the genomic level observed in this study. Altogether, it appears fair to state that microparticle-enhanced production is advancing into a broadly applicable strategy to tailor Streptomyces, Amycolatopsis, and other related filamentous Actinomycetes for natural product formation (e.g. rifamycin, ivermectin, etc.) and should be further explored to support their discovery, development, and industrialization (Barton et al., 2018).
During the work, the streamlined S. albus "clean" chassis turned out to grow fast and exhibit high and stable pamamycin biosynthesis, suggesting further use (Myronovskyi et al., 2018). "Cluster-free" chassis strains appear particularly promising for selective production in the future, given the activation of native clusters observed here ( Figure 5).

DATA AVAILABILITY STATEMENT
The raw and processed RNA sequencing data of this article are available as MIAME-compliant datasets in Gene Expression Omnibus under the accession number GSE155008. The authors declare that all other data supporting the findings of this study are available within the article and its supplementary information file.