The activity of CobB1 protein deacetylase contributes to nucleoid compaction in Streptomyces venezuelae spores by increasing HupS affinity for DNA

Abstract Streptomyces are soil bacteria with complex life cycle. During sporulation Streptomyces linear chromosomes become highly compacted so that the genetic material fits within limited spore volume. The key players in this process are nucleoid-associated proteins (NAPs). Among them, HU (heat unstable) proteins are the most abundant NAPs in the cell and the most conserved in bacteria. HupS, one of the two HU homologues encoded by the Streptomyces genome, is the best-studied spore-associated NAP. In contrast to other HU homologues, HupS contains a long, C-terminal domain that is extremely rich in lysine repeats (LR domain) similar to eukaryotic histone H2B and mycobacterial HupB protein. Here, we have investigated, whether lysine residues in HupS are posttranslationally modified by reversible lysine acetylation. We have confirmed that Streptomyces venezuelae HupS is acetylated in vivo. We showed that HupS binding to DNA in vitro is controlled by the acetylation. Moreover, we identified that CobB1, one of two Sir2 homologues in Streptomyces, controls HupS acetylation levels in vivo. We demonstrate that the elimination of CobB1 increases HupS mobility, reduces chromosome compaction in spores, and affects spores maturation. Thus, our studies indicate that HupS acetylation affects its function by diminishing DNA binding and disturbing chromosome organization.


Introduction
All living organisms face the requirement of compacting their genetic material inside the cell or nucleus while concomitantly providing access for DNA transactions such as chromosome replication and partitioning, transcription, or DNA repair.Bacteria condense their chromosomes approximately 1000-fold into a 'nucleoid' structure suspended within the cytoplasm (1)(2)(3)(4).The levels of DNA compaction become even higher in sporulating bacteria, whose chromosomes are up to 3-fold more compacted within spores than during vegetative growth ( 5 ).Among several factors contributing to chromo-some packaging, nucleoid-associated proteins (NAPs) are one of the key players (6)(7)(8).NAPs are a highly heterogeneous group of small, basic proteins that bind DNA without (or with very low) sequence preferences and promote DNA wrapping, bending or bridging ( 4 ,8-10 ).Due to functional similarities, NAPs are often referred to as histone-like proteins, although their homology with eukaryotic histones is low.The most conserved and abundant NAPs in bacteria are HU ( h eat u nstable ) homologues ( 11 ,12 ).In general, the cellular repertoire of NAPs often differs among bacterial species.Moreover, the levels of NAPs can change notably in response to altered growth conditions or environmental signals, making NAPs the major remodellers of bacterial nucleoids ( 13 ).
Streptomyces, as soil-dwelling bacteria, are particularly exposed to environmental stressors.In response to stress conditions, Streptomyces produce secondary metabolites, many of which have valuable antibacterial and antifungal activities.Environmental stress, particularly nutrient limitation, also induces sporulation, which is the final stage of their fungal-like development.Streptomyces development starts with spore germination, followed by the growth of branched, multigenomic vegetative hyphae.Sporulation is initiated by the formation of unbranched aerial (sporogenic) hyphae.The life cycle ends with the transformation of multigenomic aerial hyphae into long chains of unigenomic exospores ( 14 ).Streptomyces differentiation is tightly controlled by the bld and whi regulatory cascades that synchronize the particular stages of hyphal growth with chromosome segregation and compaction ( 5 ,15-17 ).The structure of Streptomyces linear chromosomes changes dramatically during their complex life cycle, from uncondensed in vegetative hyphae to highly compacted within spores ( 18 ).The gradual rearrangements of their chromosome are executed by the concerted action of NAPs and condensins.Although extensively studied, the role of NAPs during Streptomyces sporulation is complex and still not fully understood.Most of the NAPs studied thus far (i.e.DpsA, DdbA, sIHF, HupS) enhance chromosome compaction in spores, and their removal leads to an increase in the nucleoid volume; however, the elimination of others (DpsB, DpsC) shows the opposite effect ( 5 ).
Among several Streptomyces NAPs that are activated during sporogenic development, the HupS protein is the best studied.HupS is one of two (apart from the HupA protein) HU homologues encoded in Streptomyces genomes ( 19 ).The HupS protein is composed of two domains.Its N-terminal domain shows homology to the Streptomyces HupA protein and Esc heric hia coli HU homologue.On the other hand, the Cterminal domain of HupS resembles the eukaryotic histones or mycobacterial HupB homologue (also referred to as Hlp), as it contains multiple lysine repeats (LR domain) ( 20 ,21 ).Although both HupS and HupB proteins do not show sequence specificity in DNA binding, they exhibit nonuniform distribution along the chromosome.Whereas in Streptomyces venezuelae HupS binding increases towards the chromosomal termini ( ter ), the enrichment of Mycobacterium smegmatis HupB binding sites within the circular mycobacterial chromosome shows an opposite pattern, with elevated binding in proximity to the oriC region ( 18 ,22 ).The deletion of the hupS gene in Streptomyces coelicolor and S. venezuelae increased nucleoid volume in spores ( 18 ,19 ).In S. coelicolor , HupS elimination was also shown to lower spore resistance to heat and UV radiation ( 19 ).In mycobacteria, deletion of the hupB gene also increased sensitivity to environmental stresses, including cold shock, UV radiation, exposure to isoniazid, and host persistence (23)(24)(25).Interestingly, the C-terminal truncation of HupB resulted in abolished DNA affinity and led to nucleoid decompaction in M. smegmatis, thus revealing a similar phenotype to the hupB null mutant ( 22 ,26 ).The identification of LR domains only in DNA binding proteins reinforces their significance for protein −DNA interactions ( 27 ).Markedly, the LR domain of HupB has been shown recently to be a target for posttranslational modifications (PTMs), including reversible multiple lysine methylation and acetylation ( 28 ,29 ).M. tuberculosis strains lacking acetyl-and methyl-transferases that modify the LR domain of HupB showed increased susceptibility to antibiotics and lowered host persistence, characteristic of hupB-deficient strains (29)(30)(31).
Reversible lysine acetylation (RLA) allows immediate upor downregulation of protein activities in response to internal or environmental stimuli, bypassing energy-consuming de novo protein synthesis or degradation ( 32 ).Ubiquitous in bacteria, lysine acetylation can occur spontaneously in the cytoplasm or can be catalysed by acetyltransferases (Pat), which are homologues of the yeast Gcn5 histone N-acetyltransferase (GNAT) ( 33 ).Both lysine acetylation pathways depend on the intracellular availability of acetyl phosphate (AcP) or acetylcoenzyme A (acetyl-CoA).AcP and acetyl-CoA are two highenergy metabolites serving as cytoplasmic donors of acetyl groups, directly linking primary metabolism with the global regulation of protein activities.Lysine acetylation can be reversed only enzymatically, requiring the activity of protein deacetylases.Bacterial deacetylates belong to two groups, depending on the NAD + requirements.CobB deacetylases (also referred to as SrtA) are homologues of the yeast Sir2 protein and require NAD + as a cofactor.On the other hand, AcuC homologues coordinate Zn(II) to catalyse the transacetylation reaction in an NAD + -independent manner ( 32 , 34 , 35 ).
In S. coelicolor , two Sir2 homologues have been identified: CobB1 (SCO0452) and CobB2 (SCO6464).Whereas CobB1 has lysine deacetylase activity in vitro , CobB2 has been reported recently to have a preference for succinate groups attached to the side chain of a lysine residue and was suggested to play a role as a cellular protein desuccinylase with poor deacetylation activity ( 36 ,37 ).Interestingly, the identification of 601 and 667 acetylated proteins in S. coelicolor and Streptomyces roseosporus , respectively, indicates that reversible lysine acetylation may be a widespread mechanism controlling protein activity in these bacteria and suggests the presence of many potential protein substrates for CobB1 protein deacetylase ( 37 ,38 ).
The presence of the LR domain, which is extremely rich in lysine residues, in HupS suggests that its posttranscriptional regulation is similar to that detected for mycobacterial HupB.Therefore, we aimed to test whether HupS is modified by lysine acetylation and whether such a modification could affect the interaction of HupS with DNA.Having confirmed that HupS is the target of acetylation, by combining in vivo and in vitro studies, we showed that increased HupS acetylation lowers its DNA affinity, leading to nucleoid decompaction in S. venezuelae spores.Moreover, we confirmed that lysineacetylated HupS is a substrate for the CobB1 protein deacetylase, whose levels increase during S. venezuelae sporogenic differentiation.Thus, we postulate that CobB1 serves as a regulator of HupS-DNA binding in S. venezuelae, promoting HupSdependent chromosome organization during sporulation.In summary, our studies expand the scope of knowledge on the role of PTMs in the regulation of chromosome organization in sporulating bacteria.

Bacterial strains and plasmids
The Esc heric hia coli and S. v enezuelae strains used in this study are listed in the Supplementary Materials and Methods ( Supplementary Tables S1 and S2 ).The oligonucleotides and plasmids are listed in the Supplementary Materials and Methods ( Supplementary Tables S3 and S4 ).The culture conditions, antibiotic concentrations, and transformation and conjugation methods followed the general procedures for E. coli ( 39 ) and Streptomyces ( 40 ).All DNA manipulations were performed according to standard procedures ( 39 ) or the manufacturer's protocols.The DNA-modifying enzymes, restriction enzymes, and DNA polymerases were purchased from New England Biolabs (US) or Thermo Fisher Scientific (USA).The oligonucleotides were synthesized by Genomed S.A. (Poland) or Sigma −Aldrich (US).The details of plasmid and strain constructions as well as the protein purification protocols are described in the Supplementary Materials and Methods.

Growth analyses
S. venezuelae strains were cultivated on MYM (solid or liquid), SFM, or minimal (MM) medium not supplemented with mannitol.The media and antibiotics were purchased from Carl Roth (Germany), Gibco (USA), Becton Dickinson (USA), and A&A Biotechnology (Poland).To obtain the S. venezuelae growth curves, a Bioscreen C instrument (Growth Curves, USA) was used.S. venezuelae cell cultures were set up in sterile honey-comb microplates (Growth Curves, US) in 300 μl liquid MYM medium inoculated with 0.0001 U of spores (1 U is an optical density (OD 600 ) of the spore stock solution equal to 1 when measured in miliQ water).Each strain was cultured in three independent experimental repeats for 48 hours at 30 • C under the 'medium' speed and 'normal' shaking amplitude settings.The growth of S. venezuelae was monitored by optical density (OD 600 ) measurements in 20-min intervals.The data were collected using BioScreener 3.0.0software and the growth curves were plotted using Microsoft Excel.
For S. venezuelae growth and development analyses on solid plates, the spore suspensions in 10% glycerol were diluted with sterile water to reach a spore concentration of 0.001 U / μl.Then, 5 μl of spore dilution was plated onto MYM, SFM or MM, and incubated at 30 • C for up to 72 h (growth analysis) or 120 h (spore pigmentation).
To assess the spore germination efficiency, the frozen spore aliquots were thawed on ice, the spores were counted under microscope using Thom's chamber, and subsequently diluted (10 6 -10 8 times) with chilled 10% glycerol.Next, 50 μl of each spore dilution was spread on MYM plate and incubated at 30 • C for 48 hours.The experiment was performed as 3-4 replicates.The obtained colonies were counted and the average CFU ( c olony f orming u nits ) was calculated.The germination efficiency (%) is the average number of CFU shown as a fraction of the total number of spores streaked on the plate.

His-HupS acetylation and deacetylation in vitro
For the in vitro nonenzymatic acetylation reaction, 1.5 ml of a reaction mixture containing 2.25 mg of purified His-HupS, 100 mM lithium potassium acetyl phosphate (Sigma −Aldrich, US), 10 mM MgCl 2 , 50 mM Tris-HCl pH 8.0 and 300 mM NaCl was incubated at 37 • C for 1 h.After incubation, the volume of the reaction mixture was adjusted to 2.5 ml with buffer C (50 mM Tris-HCl pH 8.0, 300 mM NaCl).Acetyl phosphate (AcP) was then removed using filtration on a PD-10 column equilibrated with buffer C. In parallel, a control, nonacetylated His-HupS was prepared following the same protocol but omitting the acetyl phosphate in the reaction mixture.In vitro, nonenzymatic His-HupS acetylation was confirmed using 1 μg of acetylated or nonacetylated protein applied directly on a nitrocellulose membrane (dot-blot assay).Protein detection was subsequently performed using the Western blotting protocol described earlier using mouse monoclonal anti-HisTag antibody for loading control and rabbit polyclonal anti-AcK antibody for acetyllysine detection.
The in vitro deacetylation was performed at 37 • C for 6 h in a 20 μl reaction mixture containing 25 mM Na 2 HPO 4 pH 8.0, 150 mM NaCl, 1 mM NAD + , 0.5 mM DTT, 3 μg of purified His-CobB1 protein and 180 ng of acetylated His-HupS protein.The reaction was stopped with 4 μl of 6X SB buffer (375 mM Tris-HCl pH 6.8, 60% glycerol, 12% SDS, 600 mM DTT, 0.06% bromophenol blue) followed by incubation of the sample at 95 • C for 5 minutes.After the deacetylation reaction, 10 μl of the sample was analysed in western blotting with anti-His antibodies to detect His-CobB1 and His-HupS proteins, and 10 μl of the sample was analysed with anti-AcK antibodies to detect lysine acetylation.

Biolayer interferometry
Biolayer interferometry (BLI) was used for three different biotinylated DNA fragments, including 5 biotinylated doublestranded DNA (one-end biotinylated dsDNA, 300 bp), 5 and 3 biotinylated double-stranded DNA (two-ends biotinylated dsDNA, 300 bp) and 5 biotinylated single-stranded DNA (one-end biotinylated ssDNA).dsDNA fragments were PCRamplified using one or two 5 -biotinylated oligonucleotides, BLI_ parS 300_FW and BLI_ parS300 _RV (marked with '[Btn]' in Supplementary Table S3 ), and pUC19A7 plasmid as a template (constructed by ligation of S. coelicol or parAB operon 500 bp upstream region amplified using H24 parA SmaRV and parA pset oligonucleotides with pUC19 plasmid) and purified from the agarose gel.5 -Biotinylated single-stranded DNA (ssDNA, 117 nt; Btn-parS -oligo) was synthesized by Sigma −Aldrich (US), diluted and directly loaded as described below.The affinity measurements were performed with an Octet R2 Protein Analysis System (Sartorius, Germany) equipped with Octet High Precision Streptavidin 2.0 SAX2 Biosensors (Sartorius, Germany).In the first step, the streptavidin-coated biosensor was hydrated for 15 minutes on the bench in association buffer (100 mM Tris-HCl pH 8.0, 200 mM NaCl, 1 mM MgCl 2 , 0.005% Tween-20, 0.1 mg / ml BSA).Then, the sensors were mounted onto the Octet R2 Protein Analysis System and calibrated for 60 seconds in association buffer, followed by loading 150 ng of the biotinylated DNA fragments diluted in association buffer for 5 minutes.The association step (450 s) was then established by dipping the biosensors in association buffer containing a broad range of His-HupS concentrations (0-100 nM) in a final volume of 200 μl dispensed in a 96-microwell black plate (Greiner Bio-One, Austria).The His-HupS protein dissociation was subsequently monitored for 180 s in protein-free association buffer.The analyses were performed in cycles of His-HupS association and dissociation.Between each cycle, the biosensor tips were washed for 60 s with association buffer, regenerated for 5 minutes with regeneration buffer (100 mM Tris-HCl pH 8.0, 200 mM NaCl, 1 mM MgCl 2 , 0.005% Tween-20, 0.1% SDS), and equilibrated for 5 min with association buffer.All BLI experiments were performed at 30 • C. The data were collected and analysed using Octet Data Analysis software version 11.1.The dissociation constants ( K D ) were calculated by fitting the data to the log-logistic model using R Studio software.

Pulldown analysis
For the HupS-FLAG pulldown assay, 0.05 U of spores of the S. venezuelae wild-type and hupS -FLAG strains (TM015 strain ( 18 )) were used to inoculate 50 ml liquid MYM medium cultures grown at 30 • C with shaking (180 rpm) for 13, 17 and 21 h.At each time point, the optical density of the cultures was measured.The cells were collected by centrifugation (5000 g, 10 min, 4 • C), washed twice with wash buffer (50 mM Tris-HCl pH 8.0, 250 mM NaCl), and resuspended in 10 ml of lysis buffer (50 mM Tris-HCl pH 8.0, 250 mM NaCl) supplemented with Pierce™ Protease Inhibitor Tablets (Thermo Fisher Scientific, US).The cells were disrupted by sonication, and the obtained cell lysates were clarified by centrifugation (10 000 g, 15 min, 4 • C) followed by the filtering of the supernatant through a 0.45 μm pore filter.The clarified cell lysate was supplemented with Tween-20 (Carl Roth, Germany) to a final concentration of 0.05% and incubated overnight with constant tube rotation at 4 • C with 50 μl of magnetic beads coated with anti-FLAG® BioM2 antibody (Sigma Aldrich, USA).The next day, the magnetic beads were collected and washed three times with 1 ml of wash buffer.To elute bound protein, the magnetic beads were incubated in SDS-elution buffer (62 mM Tris-HCl pH 6.8, 2% SDS 10% glycerol, 0.01% bromophenol blue) at 95 • C for 5 min (for Western blotting).The eluted proteins were analysed by Western blotting using mouse monoclonal anti-FLAG antibodies (for HupS-FLAG detection) or rabbit polyclonal anti-AcK antibodies followed by incubation with HRP-conjugated antimouse IgG or anti-rabbit IgG secondary antibodies, respectively, according to the standard western blotting protocol described earlier.

Intact protein LC −MS
The acetylated and nonacetylated His-HupS samples were concentrated and initially desalted by overnight acetone precipitation at 4 • C. The next day, the protein pellet was washed three times with ice-cold acetone and subsequently resuspended in a 0.1% formic acid solution.The total His-HupS mass analysis was carried out with the use of the M-Class Acquity UPLC system coupled to the Synapt XS high-resolution mass spectrometer (Waters Corporation, US) equipped with the Zspray LockSpray II ion source utilizing a 125 μm OD steel emitter.Approximately 7.5 and 75 pmol of the protein were injected for the nonacetylated and acetylated His-HupS samples, respectively.On the system, the samples were further desalted and chromatofocused on a nanoEase M / Z BEH C4 300 Å, 5 μm, 300 μm × 50 mm trap column, while a nanoEase M / Z BEH300 C4 300 Å, 1.7 μm, 300 μm × 100 mm analytical column was employed for chromatographic separation with an eight-minute 10% to 80% acetonitrile linear gradient at a 10 μl / minute flow rate.The columns' compartment temperature was kept at 80 • C. MS data acquisition was collected at 1 scan / s through a 300-3000 m / z range in positive polarity and TOF resolution mode.Source conditions were as follows: capillary voltage: 3 kV, sampling cone: 40 V, source offset: 15 V, source temperature: 150 • C, desolvation gas temperature: 550 • C, desolvation gas flow: 800 l / h, cone gas flow: 30 l / h, nebulizer gas pressure: 5.5 bar.Glufibrinopeptide B solution was acquired in the reference function, and the correction was applied post-acquisition.The raw data were processed and analysed using MassLynx V4.2 software.The protein peak from each analysis was integrated, and the combined spectra were background subtracted and then deconvoluted using the MaxEnt1 algorithm.

Bottom-up LC-MS
For the proteome-scale analysis, the wild-type (WT) and the hupS (AKO200) strains grew for 21 h in liquid MYM medium as described earlier for the pulldown experiment.The mycelium was collected and washed twice with 50 mM Tris-HCl buffer pH 7.2.The cell lysate was prepared in 50 mM Tris-HCl buffer pH 7.2, 8 M urea, 0.2% sodium deoxycholate.10 μg of total protein was denatured for 30 min at 50 • C with an addition of 5 mM DTT. Urea concentration was then lowered to 2 M via dilution with 50 mM Tris-HCl buffer pH 7.2, 5 mM DTT, 0.2% sodium deoxycholate solution, and 100 ng of trypsin (EMS0006, Sigma-Aldrich, USA) were added for an overnight digestion at 30 • C. The next day the sample was acidified by the addition of formic acid to a final concentration of 0.5%, incubated for 30 min at 37 • C, and then centrifuged for 15 min at 15 000 g.The supernatant was further cleaned up using the STAGE tip ( 42 ), from which the peptides were eluted using 80% acetonitrile (ACN), and 0.1% formic acid solution.The organic solvent was then removed using the SpeedVac, and the peptide pellet was resuspended in 0.1% formic acid.
For the HupS-FLAG analysis, the wild-type (WT) and hupSflag complementation (TM015) strains grew for 21 h in a liquid MYM medium, and HupS-FLAG purification followed the procedure described earlier for the pulldown experiment.The eluted proteins were resolved in SDS-PAGE electrophoresis and CBB-stained.The sections corresponding to the molecular weight of HupS were then excised, and further destained for 30 min at 37 • C in 40% ACN, 200 mM ammonium bicarbonate (ABC).The solution was then aspirated, the gel piece crushed with a pipette tip, and dehydrated by another round of incubation in fresh 40% ACN, 200 mM ABC followed by solution aspiration, and its complete removal using a Speed-Vac.Dried gel pieces were rehydrated for 45 min at 4 • C in 10 μl of 9% ACN, 50 mM ABC, and 5 mM DTT containing 100 ng of trypsin (EMS0006, Sigma-Aldrich, USA).Trypsinloaded gel samples were then further hydrated via an addition of 30 μl of 9% ACN, 50 mM ABC and 5 mM DTT and then left for overnight digestion at 30 • C. The next day the samples were briefly centrifuged, and the peptide-containing supernatant was transferred into a fresh tube.An additional peptide extraction step was carried out by incubation of the sedimented gel pieces for 30 min at 37 • C in 30 μl of 9% ACN, 50 mM ABC, and 5 mM DTT.The recovery solutions were pooled, volatiles were removed by SpeedVac, and the obtained pellet was reconstituted in 40 μl of 0.1% formic acid.
Analyses were carried out with the use of the M-Class Acquity UPLC system coupled to the Synapt XS high-resolution mass spectrometer as described above.The amount corresponding to 800 ng of total input protein was loaded for the lysate samples, and 1 μl of a 4 × dilution of the final sample was used for the pull-downs.Chromatofocusing was carried out with the nanoEase M / Z Symmetry C18 100 Å, 5 μm, 180 μm x 20 mm trap column while the nanoEase M / Z HSS T3 C18 100 Å, 1.8 μm, 75 μm x 150 mm analytical column was employed for the chromatographic separation with 90 min (lysates) or 60 min (pulldown samples) 5% to 35% acetonitrile linear gradient at a 300 nl / min flow rate.The columns' compartment temperature was kept at 40 • C. Data were collected in ultra-definition ion mobility DIA (UDMS E ) at 0.6 scan / s through a 50-2000 m / z range in positive polarity and TOF resolution mode.An ion's drift-time-adjusted collision energy ramp, specifically determined for the instrument, was applied on the transfer cell ( 43 ).Source conditions were as follows; capillary voltage: 2.5-2.6 kV, sampling cone: 40 V, source offset: 15 V, source temperature: 100 • C, cone gas flow: 40 l / h, purge gas flow: 0.6 l / h, NanoFlow gas pressure: 0.5-1.9bar.A (Glu1)-Fibrinopeptide B solution was acquired in the reference function, and the correction was applied postacquisition.Before processing, lysates' raw data were noise reduced with an ion count threshold of 10.
Data processing was performed using the Protein Lynx Global Server (PLGS) v3.0.3 software.For the deconvolution, the low and high energy ion threshold counts were set to 135 and 30, respectively, and the chromatographic and TOF peak detection parameters were set to Apex3D engine's autovalues.Obtained deconvoluted spectral data were merged from duplicates (cell lysate) or triplicates (pulldown), and searched via the Ion Accounting algorithm against either a S. venezuelae protein sequence databank (UP000006854) to which the trypsin sequence accession was appended (cell lysate) or an alike databank in which the wild-type HupS accession (F2R5C9) was substituted with the recombinant HupS-FLAG accession (pulldown).The set search parameters were as follows; peptide mass tolerance: 25 ppm; fragment mass tolerance: 25 ppm; min.fragment ion matches per peptide: 1 (lysates), 3 (pull-downs); min.fragment ion matches per protein: 3 (cell lysate), 7 (pulldown); min.peptide matches per protein: 1; max.protein mass: 1 MDa; digest reagent: Trypsin (cut after K and / or R, but not before P); Allowed missed cleavages: 10; fixed modification: none Variable modification: Oxidation of M (+15.99491Da, applies to M sidechain); Enriched variable modification (in order to enforce searching in pass one): Acetylation of K (+42.01056Da, applies to K sidechain), FDR: 1%.
The obtained data were phenotype-compliant with no HupS hit in the ΔhupS strain sample (with 1024 protein groups identified total), as well as no HupS hit in the wildtype anti-FLAG pull down (6 protein groups total).Consistently HupS was identified in the wild-type strain sample (819 protein groups total), and the TM015 anti-FLAG pull down (12 protein groups total).The former two samples were therefore used as false positive controls in confirming HupS acetylated peptides.HupS accession peptide tables from both samples were manually analyzed and peptides with an acetyllysine (AcK) modification were validated based on generally known characteristics of AcK bearing peptides, such as a non-possibility of trypsin cleavage at the C-terminus of an acetylated lysine and a general upshift of retention time between an acetylated and a non-acetylated peptide of same kind ( 44 ,45 ).Peptides not complying were assumed as false positives.Further validation was based on the quality assessment of the MS2 spectral evidence and the peptide's absence in the unmatched spectral data from corresponding control samples.Peptides with poor MS2 matches (ions within noise level, abnormal precursor / fragment abundance ratio, etc.), as well as peptides, whose ions were found in controls (based on MH+, RT, Drift and intensity comparison), were rejected.In the pulldown sample an expected higher count of MS2 fragment matches was obtained (due to HupS enrichment) enabling precise identification down to a specific HupS lysine residue.Based on the pulldown results, assumptions were made for peptides identified in the lysate sample if the MS2 evidence was not unambiguous in terms of AcK residue assignment.

RNA isolation and RT −qPCR
For RT −qPCR analyses, the S. venezuelae wild-type strain was inoculated with 0.05 U of spores and cultivated in 5 ml liquid MYM medium for 8-25 h as described earlier.RNA from mycelia was isolated using the GeneJET RNA isolation kit (Thermo Fisher Scientific, USA), according to the manufacturer's protocol.However, the lysozyme concentration in the suspension buffer was increased to 10 mg / ml to improve the yield of mycelium lysis.The isolated RNA was digested with TURBO DNase I (Invitrogen, United States) and subsequently purified and concentrated using a GeneJET RNA Cleanup kit (Thermo Fisher Scientific, USA).500 ng of total DNA-free RNA were used for cDNA synthesis using the Maxima First Strand cDNA synthesis kit (Thermo Fisher Scientific, USA).The original manufacturer's protocol was modified for GC-rich S. venezuel ae transcripts by increasing the temperature of the first-strand synthesis to 65 • C and extending the synthesis time to 30 min.Next, the cDNA sample was diluted to 100 μl with miliQ water, and 2 μl was directly used for quantitative qPCR.qPCR was performed using PowerUP SYBR Green Master Mix (Applied Biosystems, USA).The relative levels of the cobB1 transcript were compared using the comparative Ct method and the hrdB transcripts as the endogenous control.The oligonucleotides are listed in Supplementary Table S3 .Transcripts per million (TPM) were obtained from the raw RNA-Seq data available in Array Express (EMBL-EBI) database under accession numbers E-MTAB-13607 ( 46) and E-MTAB-13911 ( 47 ).The complete description of the RNA-Seq protocols including data acquisition and sample collection is attached to the deposited data.

Fluorescence microscopy
For the snapshot analysis of nucleoid areas, spores of S. venezuelae strains were diluted in sterile miliQ water and cultured for 22-24 h on SFM agar plates with sterile coverslips inserted at a 45 • angle.Next, the mycelia on the coverslips were fixed by washing twice with absolute methanol.The nucleoid and cell wall were stained for 30 min at room temperature with 7-amino-actinomycin D (1 mg / ml 7-AAD in DMSO, Invitrogen, USA) diluted 1:300 and wheat germ agglutinin conjugated with Alexa-Fluor-350 (1 mg / ml WGA-AF 350 in DMSO, Invitrogen, USA) diluted 1:200 in filtered PBS buffer.In the next step, the mycelium was washed four times with filtered PBS buffer, and the coverslips were mounted on microscopic glass slides using 50% glycerol solution in filtered PBS buffer.The mycelium was imaged using a Leica DM6 fluorescence microscope equipped with a 100 ×/ 1.4 oil objective in differential interference contrast (DIC), DAPI and mCherry channels.The nucleoid area was analysed in aerial hyphae with visible septa staining (prespores) with custom Fiji software protocols.The boxplots showing the distribution of nucleoid areas and statistical analysis were obtained using an EDA tool (the code is available at https://github.com/astrzalka/EDA ).The Wilcoxon test was used to determine the statistical significance of the differences between populations.

Single-particle tracking
For single-particle tracking (SPT), S. venezuelae hupS-halotag strains (AZ01 and JD14) were cultured for 23 h in 1 ml of liquid MYM medium supplemented with 5 nM HaloTag® ligand TMRDirect™ (Promega, USA).Immediately before imaging, mycelia were washed three times with filtered PBS buffer and spread onto agar pads (1.1% low gelling temperature; Sigma-Aldrich, USA) poured into Gene Frames (chamber 1.0 × 1.0 cm; Thermo Fisher Scientific, USA) and covered with an 18 × 18 mm coverslip (High Precision Microscope Cover Glasses, Carl Roth, Germany).Before sample preparation, microscope slides and coverslips were prepared by overnight incubation in 1 M KOH solution and subsequent drying with nitrogen flow.SPT was carried out at 30 • C using a Zeiss Elyra 7 inverted microscope equipped with an Andor iXon DU 897 electron multiplying charge-coupled device (EMCCD) camera and an alpha Plan-Apochromat 100 ×/ 1.46 Oil DIC M27 Elyra objective in combination with laser lines HR DPSS 561-500 200 mW 561 nm.The Z-axis was stabilized via the 'definite focus' system.The samples were prebleached, and the images were recorded using a 561 nm laser (80% laser power, EMCCD gain of 100) with an exposure time of 20 ms in HILO illumination mode (TIRF Angle 47.26 • ).For each position, 10000 frames were collected.SPT was performed using Fiji, TrackMate and SMTracker software ( 48 ,49 ).Cell segmentation was performed using Fiji and Oufti software ( 50 ).First, for each strain, the frequency of jump distances (JD) between consecutive frames was plotted.Then, the Squared Displacement (SQD) or Gaussian Mixture Model (GMM) analyses using SMTracker software were performed independently to distinguish between mobile and immobile HupS-HaloTag molecules.Whereas the SQD method uses the cumulative probability of square displacements, the GMM method uses the displacement of molecules between consecutive frames to fit two Gaussian models representing DNA-bound and free proteins.Both methods allowed independently for the calculation of the apparent diffusion coefficients and the relative fractions of two diffusive states of the HupS-HaloTag protein.To calculate the average residence time [ms] for molecules bound to DNA the stationary localization analysis (SLA) with SMTracker software was used to determine the number of events when the molecule remained within a specified radius and to calculate the average residence time.Due to cell confinement and motion blurring, D* is an apparent diffusion coefficient.All experiments were carried out as two independent biological replicates and combined subsequently for GMM or SQD analyses as described above.

HupS protein is modified posttranslationally by lysine acetylation
The S. venezuelae HupS protein (234 aa; 23.8 kDa), encoded by the vnz_25 950 gene ( hupS ), is composed of two domains.The N-terminal domain shows homology to HU proteins, whereas the 144 amino acid C-terminal fragment (90-234 aa) resembles the histone H2B tail and contains 43 lysine residues, and 38 of them are accumulated as KK doublets (Figure 1 A and Supplementary Figure S1 A) ( 18 ).Earlier studies have shown that in S. coelicolor, HupS levels increase during sporogenic development due to the transcriptional upregulation of the hupS gene ( 19 ).To confirm whether the same hupS transcriptional upregulation occurs during the development of S. venezuelae , we measured the levels of the hupS transcript during differentiation of the wild-type strain, and we monitored HupS-FLAG and HupS-HaloTag protein levels using strains expressing hupS-flag or hupS-halotag genes, (TM015 and AZ01 respectively).Both genes encoding fusion proteins were delivered in trans under the control of the native hupS promoter ( Supplementary Table S2 ).Our previous studies ( 18 ), also reinforced here (Figure 5 C and E), showed that HupS C-terminal fusion with the FLAG-tag complemented the phenotype of S. venezuelae hupS deletion strain, confirming the functionality of HupS-FLAG recombinant protein.
Analyses of hupS transcript levels between 9th and 25th hour of growth confirmed that, the hupS gene is strongly activated during S .venezuelae differentiation, reaching the maximum transcript levels prior to sporulation (up to 6-fold in relation to the vegetative phase) and gradually decreasing during spore maturation (Figure 1 B).The increase of HupS-FLAG protein levels detected with anti-FLAG antibodies correlated with hupS gene upregulation during S. venezuelae sporogenic development, although the protein levels remained constant throughout most of the spore maturation phase (with the exception of 25 h of growth when we detected a rapid decrease in HupS-FLAG levels; Figure 1 B).This observation corroborates the microscopy observation of the HupS-HaloTag signal during development.In S. venezuelae vegetative hyphae, HupS-HaloTag fluorescence was relatively low but increased strongly in the aerial mycelium and remained elevated during spore maturation (Figure 1 C).Our analyses confirmed the previous reports ( 18 ,19 ) of the HupS upregulation during sporogenic development and its localisation in Streptomyces spores (Figures 1 B and C).
Next, we investigated whether HupS may be regulated not only by the transcriptional activation of the hupS gene (Figure 1 B) but also posttranslationally by reversible lysine acetylation, as found previously for its M. tuberculosis homologue (HupB) ( 30 ) or the lysine-rich tail of eukaryotic histone H2B ( 51 ).First, we assessed global protein acetylation (acetylome) in lysates obtained from S. venezuelae wild-type and hupS null mutant strain cultures.Western blotting with anti-AcK polyclonal antibodies showed that the total signal from acetylated proteins increased in the wild-type S. venezuelae strain during development progression.Moreover, one of the acetylated proteins (with a molecular mass slightly below 35 kDa) was present in the wild-type strain but not in the acetylome of the hupS deletion strain, while other acetylated proteins were still detectable in both strains at comparable levels.Thus, we speculated that this protein may correspond to HupS (Figure 1 D).
To confirm HupS acetylation in vivo, we used the S. venezuelae hupS-flag strain to purify the recombinant HupS-FLA G protein (with anti-FLA G antibodies immobilized on magnetic beads) from the cell lysates collected at three developmental stages: vegetative growth (VM), aerial mycelium formation (AM) and the spore maturation phase (SP).Subsequently, the purified HupS-FLAG protein was detected with anti-FLAG and anti-AcK antibodies to detect the HupS-FLAG in the eluent and HupS-FLAG lysine acetylation, respectively.The lysine-acetylated protein was detected at all three developmental stages, suggesting that HupS-FLAG is constantly acetylated throughout the S. venezuelae life cycle (Figure 1 E).
We also attempted to identify HupS acetylation sites in vivo .For that, we used LC-MS analyses based on untagged HupS protein (global proteome analysis of the WT type strain) or purified HupS-FLAG (pulled down from TM015 strain lysate).We identified two acetylated lysine residues positioned within the N-terminal HU domain (K51 and K85) and four acetylated lysine residues located within the LR domain (K104, K119, K193 or K194).Two of them (K85 and K119) were detected in both analyses (Figure 1 A and Supplementary Figure S2 A) confirming HupS acetylation in vivo .Unfortunately, the low peptide coverage within the C-terminal domain strongly limited the detection of other potential acetylation sites within the LR domain ( Supplementary Figure S2 A).However, the western blotting (with anti-FLAG antibodies) identified the polypeptides, which co-purified with the intact HupS-FLAG.Their molecular weights, at least 10 kDa lower than the intact protein, suggested partial degradation of the N-terminal fragment of the HupS-FLAG protein.These peptides were also detected using anti-AcK antibodies indicating lysine acetylation within the LR domain ( Supplementary Figure S2 B).Thus, even though LC-MS studies delivered limited information concerning acetylation of the LR domain, our analyses reinforced by the Western blot confirmed the lysine acetylation within either the N-and C-terminal domains of HupS.
In summary, we confirmed that in S. venezuelae, hupS is regulated at the transcriptional level, and its expression increases during sporogenic development.Moreover, we provided the first evidence that HupS-FLAG, similar to its mycobacterial homologue HupB, is acetylated on lysine residues in vivo.

Lysine acetylation affects the DNA affinity of HupS in vitro
Having confirmed HupS acetylation in vivo , we set out to determine the impact of this HupS modification on its DNA binding.To study the influence of lysine acetylation on HupS-DNA interactions, we overproduced His-HupS recombinant protein (26.1 kDa) in E. coli , and purified it using ion exchange chromatography and salt gradient fractionation HupS-HaloTag (red) at subsequent S. venezuelae developmental stages (in the hupS-halotag strain, AZ01).The cell wall was visualized using wheat germ agglutinin (WGA) conjugated with Alexa Fluor 350 (blue).The hyphal morphology was visualized using phase contrast (PC).Examples of aerial mycelia and chains of prespores are marked with white and y ello w arro ws, respectiv ely.Scale bar: 1 μm.( D ) Western blot detection of protein lysine acetylation in the cell lysates of the wild-type S. venezuelae strain cultured in liquid MYM.As the control, the lysate obtained from the hupS deletion strain (in 9 hour of growth, ΔhupS strain, AKO200) was used.The putative band corresponding to the HupS protein is marked with a black arrow.The loading control (CBB-stained acrylamide gel) is shown below.The 25, 35 and 70 kDa bands of the protein molecular wemultight ladder are marked.( E ) The pulldown of HupS-FLAG protein from S. venezuelae cell lysates obtained at different developmental stages (v egetativ e (VG) and aerial (AM) m y celium, or spore maturation (SP)) of S. venezuelae ( hupS-flag strain, TM015) growing for 13, 17 and 21 h in liquid MYM medium.The optical density of the culture is indicated for each growth phase.The HupS-FLAG protein levels were detected using anti-FLAG antibodies ( αFLAG).Lysine acetylation was detected using anti-AcK antibodies ( αAcK).The co-eluted heavy chain of IgG antibodies detected by anti-mouse IgG served as a loading control.
(Figure 2 A).Taking into account the high positive charge of the nonacetylated HupS protein (theoretical pI = 11.14), which is lowered when the protein is acetylated, this simplified, one-step procedure delivered nonacetylated His-HupS (or acetylated at a low level).Although the purified recombinant protein migrated slower (slightly below 35.0 kDa) in SDS-PAGE gel than suggested by its molecular weight prediction (26.1 kDa), anti-HisTag antibodies detected His-tagged HupS protein in the collected fraction III (Figure 2 A), while the anti-AcK antibodies and LC −MS analysis did not detect its acetylation (Figures 2 B and C).In the next step, we acetylated the His-HupS protein in vitro (Figure 2 B) in the presence of 100 mM acetyl phosphate (AcP) as an acetyl group donor ( 52 ,53 ).The acetylation of the His-HupS protein (+AcP) compared to the nonacetylated protein (-AcP) was verified using an anti-AcK antibodies (Figure 2 B) and LC −MS analysis of the nontrypsinized His-HupS protein (Figure 2 C).The acetylation in vitro delivered the multiacetylated His-HupS protein carrying from 12 to up to 21 modified lysine residues per a single His-HupS molecule with a dominant fraction of the proteins modified from 13 to 17 times per single His-HupS molecule ( Supplementary Figure S2 C).
To study the impact of multiple acetylation of His-HupS protein on its DNA binding, we used biolayer interferometry (BLI) and one-end biotinylated double-stranded DNA fragments (300-kb dsDNA).The comparative DNA binding analysis of two His-HupS variants, acetylated (+AcP) and nonacetylated (-AcP), showed that the nonacetylated His-HupS protein associated more rapidly with DNA than its acetylated variant at all tested protein concentrations (20-60 nM) (Figure 2 D).The subsequent quantification of the dissociation constants ( K d ) for the acetylated and nonacetylated proteins showed a decreased DNA affinity of lysineacetylated His-HupS ( K d = 28.6 ± 3.2 nM) in comparison to its nonacetylated variant ( K d = 17.4 ± 1.3 nM) (Figure 2 E).When dsDNA fragments were immobilized using both ends (two-ends immobilized dsDNA), His-HupS affinity to DNA was higher than when the DNA was immobilized using only one end.Nevertheless, acetylated His-HupS had an affinity lower than that of the acetylated protein ( K d = 11.3 ± 1.0 nM and K d = 20.0 ± 2.6 nM for nonacetylated and acetylated variants, respectively) (Figure 2 E).Analysis of a single-stranded (117 nt), one-end immobilized oligonucleotide showed that His-HupS was also able to bind single-stranded DNA (ssDNA) with comparable affinity as for double-stranded, two-end immobilized DNA ( K d = 10.6 ± 0.9 nM), and the affinity of His-HupS to this substrate also decreased as a result of His-HupS lysine acetylation ( K d = 17.6 ± 2.6 nM) (Figure 2 E).In addition to BLI analyses, we also investigated His-HupS affinity for supercoiled and linearized plasmid DNA using an electrophoretic mobility shift assay (EMSA), showing that the nucleoprotein complex formed more efficiently in the presence of supercoiled plasmid DNA in comparison to the same amount of the linearized plasmid ( Supplementary Figure S3 ).These observations suggested that although HupS has no nucleotide sequence specificity ( 18 ), its DNA affinity might depend on the DNA substrate with a preference for ssDNA or supercoiled DNA.
Earlier studies on HU homologues in E. coli showed that HU α and HU β may form dimers ( 54 ).It could not be excluded that HupS acetylation disturbs HupS-DNA complex formation by affecting HupS dimerization.Thus, next, we tested whether the lower DNA affinity might be correlated with putative disturbances in protein −protein interactions of the acetylated His-HupS variants.Glutaraldehyde crosslinking followed by SDS −PAGE analysis showed that both the acetylated (AcP+) and nonacetylated (AcP-) His-HupS variants formed high-molecular weight complexes with similar efficiency (Figure 2 F).Thus, this observation suggested that the decreased DNA affinity of the acetylated HupS did not result from disturbances in protein −protein interactions but rather was the effect of lowered affinity of HupS monomers and / or dimers for DNA due to lysine acetylation.
In summary, our in vitro studies showed that the His-HupS protein bound single-and double-stranded DNA and that multiple lysine acetylation decreased HupS affinity for all tested DNA substrates but did not affect HupS dimerization.Moreover, the in vitro studies showed the preference of the HupS protein for single-stranded or supercoiled DNA.

CobB1 affects HupS acetylation levels in vitro and in vivo
The S. venezuelae genome encodes two putative Sir2like protein deacetylases-CobB1 ( vnz_03 080 ) and CobB2 ( vnz_31 025 ).Our transcriptomic analyses ( 46 ), confirmed subsequently by RT −qPCR, showed that cobB1 transcript levels increased during S. venezuelae differentiation, reaching up to 2-fold elevated levels during aerial mycelium formation or sporogenic differentiation.On the other hand, the cobB2 gene was poorly transcribed during both vegetative and aerial growth (Figure 3 A).Taking into account the levels of both potential deacetylases in relation to the increase in protein acetylation levels during S. venezuelae development (Figure 1 D), we set out to test whether HupS could be the target for CobB1.
First, we assessed the effect of cobB1 gene deletion on HupS acetylation in vivo .To this end, we constructed cobB1 and cobB2 deletion strains and compared the acetylome (using anti-AcK antibodies) in these strains to that of the wild-type S. venezuelae and hupS deletion strains.Only cobB1 deletion mildly affected the level of most protein acetylation, with the exception of the HupS protein.The signal corresponding to HupS (observed in the wild-type strain but not detected in the hupS deletion background) was up to 2-fold elevated in the cobB1 deletion background (Figure 3 B and Supplementary Figure S4 A) while the hupS transcript levels remained constant.Moreover, we observed that cobB1 deletion did not induce cobB2 gene transcription and, on the other hand, cobB2 deletion did not affect HupS acetylation levels in contrast to the cobB1 or double cobB1 / cobB2 deletion strains ( Supplementary Figures S4 B and C), thus indicating the preferential involvement of CobB1 deacetylase in the regulation of the HupS acetylation state in vivo .To gain more insight into the impact of CobB1 on HupS, we also constructed and analysed an S. venezuelae strain expressing an additional copy of the cobB1 gene under the control of the constitutive ermE promoter ( his -cobB1 ↑ strain).Analyses of the acetylome in the cobB1 overexpressing strain ( his -cobB1 ↑ ) showed decreased lysine acetylation signals of several proteins, including the HupS protein, indicating a more pronounced effect of cobB1 overexpression than its deletion ( Supplementary Figure S5 ).Since the S. venezuelae acetylome was only partially affected by cobB1 deletion or overexpression, this result suggests that the CobB1 deacetylase may show some specificity towards acetylated substrates, and among a limited number of protein substrates, acetylated HupS can be recognized and processed preferentially by CobB1.To confirm the activity of the CobB1 protein as a HupS deacetylase in vitro , we overproduced His-CobB1 (35.0 kDa) recombinant protein in E. coli and purified it ( Supplementary Figure S6 A).Subsequently, we used the in vitro acetylated His-HupS protein as a substrate for His-CobB1 deacetylase .We detected decrease in the relative His-HupS acetylation signal (up to 40%) in the presence of purified His-CobB1 and NAD + as the essential cofactor as compared to the control sample in the absence of His-CobB1 (Figure 3 C and Supplementary Figure S6 B).
In summary, we showed that the levels of cobB1 transcripts increase during sporogenic S. venezuelae development.We confirmed that the lack of CobB1 deacetylase increases HupS acetylation while its overproduction decreases HupS acetyla- tion in vivo.Moreover, we showed that S. venezuelae CobB1 has NAD + -dependent deacetylase activity in vitro , recognizing acetylated HupS as a substrate.

The mobility of HupS molecules in vivo is influenced by CobB1 activity
Based on our above-described observations, we hypothesized that CobB1 activity may affect HupS affinity to DNA in S. venezuelae hyphae and spores.To test our hypothesis, we used S. venezuelae strains producing HupS-HaloTag protein in the wild-type and ΔcobB1 background and analysed the mobility of protein molecules by high-resolution microscopy and a single particle tracking (SPT) approach (Figure 4 A and Supplementary Figure S7 ).The earlier described applications of SPT for DNA binding proteins showed that DNA-bound proteins had low mobility, while cytoplasmic, nonbound protein molecules were highly mobile (55)(56)(57)(58).
HupS-HaloTag SPT studies showed that in the wild-type strain, the fraction of immobile, presumably DNA-bound protein increased from 43.6% mycelium to 64.9% in spores (measured in 23th hour of growth) (Figure 4 B).Moreover, in the wild-type background, the average HupS-HaloTag residence time increased from 0.175 to 0.236 s in mycelium and spores, respectively (Figure 4 B).These observations suggest that during sporogenic development, more HupS-HaloTag molecules are associated with DNA or higher DNA compaction restricts HupS-HaloTag mobility.The cobB1 deletion resulted in a slight increase in the immobile, DNA-bound fraction of HupS-HaloTag in hyphae; however, the average residence time on DNA was similar (0.170 s) in comparison to the wild-type background (0.175 s).The effect of cobB1 deletion was more prominent in spores and led to an increased fraction of mobile HupS-HaloTag proteins, from 35.1% in the wild-type strain to 41.8% in the ΔcobB1 strain.Moreover, the HupS-HaloTag residence time decreased from 0.236 s to 0.212 s during spore maturation in the absence of CobB1 (Figure 4 B).
Our results indicate that HupS binding to DNA increases during S. venezuelae sporogenic development.Moreover, in the absence of CobB1, when HupS acetylation is enhanced, the fraction of immobile, DNA-bound protein molecules decreases, confirming that acetylation modulates HupS-DNA binding.

Alterations in CobB1 levels affect S. venezuelae growth and decrease nucleoid compaction in spores
Considering the role of CobB1 in HupS deacetylation, the influence of HupS lysine modifications on its DNA binding, and the significance of HupS binding for DNA compaction, we expected to observe changes in nucleoid area in the cobB1 deletion strain, similar to those described earlier in the hupS deficiency strain.Moreover, since hupS deletion resulted in decreased spore viability, we expected that cobB1 deletion may also affect spore germination ( 18 ,19 ).
The analysis of the influence of cobB1 deletion on the growth rate showed mild growth inhibition in comparison to the wild-type strain, detectable mostly during the stationary phase corresponding with sporogenic differentiation of S. venez uelae (Figure 5 A).The analysis of the ΔcobB1 strain growth on solid medium (rich MYM and sporulationaccelerating SFM medium) suggested aberrant colony pigmentation, similar to those observed for hupS deletion strain (Figure 5 B), and corroborating the phenotype of the hupS deletion strain in S. coelicolor ( 19 ).Surprisingly, the colony pigmentation was not fully restored by the complementation of the hupS deletion strain with HupS-FLAG protein (Figure 5 B).On the other hand, the effect of CobB1 overproduction ( his -cobB1 ↑ ) was more pronounced.An increase in CobB1 levels lowered the growth rate in both liquid and solid medium (MYM) (Figures 5 A and Supplementary Figure S8 ) but did not influence differentiation on minimal medium (MM) ( Supplementary Figure S8 ) or colony pigmentation (Figure 5 B).
The observed slowed culture growth and aberrant pigmentation of colonies of the ΔcobB1 strain suggested the putative contribution of CobB1 deacetylase to spore development.To assess spore viability, we quantified the percentage of germinating spores (germination efficiency) for ΔcobB1, cobB1 overexpression, and ΔhupS compared to wild-type and two complemented strains ( ΔcobB1 + his-cobB1 or ΔhupS + hupS-flag ).The germination efficiency for the ΔcobB1 strain (14% ± 2.2) was up to 5.4-fold less than that of the wild-type strain (76% ± 19.7) and comparable to that of the hupS deletion strain (29% ±12.5).When the cobB1 gene under the control of the constitutive ermE promoter was delivered in trans in the cobB1 deletion background ( ΔcobB1 + his -cobB1, complementation strain), germination efficiency (67% ± 12.7) was comparable to that of the wildtype strain (Figure 5 C) or hupS deletion strain complemented with hupS -FLAG under its native promoter (65% ± 13.4).On the other hand, the germination efficiency quantified for the his -cobB1 ↑ strain (98% ± 18.4) did not show any significant differences in comparison to the wild-type strain.
The observed similar impact of cobB1 or hupS deletions on spore germination efficiency and the fact that HupS contributes to chromosome compaction prompted us to analyse nucleoid organisation in the absence of CobB1.We previously observed that hupS deletion significantly disturbed chromosome compaction, increasing the nucleoid area ( 18 ).Here, we analysed microscopically the late sporogenic hyphae of wildtype and ΔcobB1 strains by measuring the nucleoid area.The microscopy observations did not indicate chromosome segregation defects in ΔcobB1 strain compared to wild type strain (Figure 5 D) corroborating the lack of segregation defects in hupS deletion strain observed previously in S. venezuelae ( 18 ).Moreover, and also similar to hupS strain, cobB1 gene deletion showed increased nucleoid areas and their higher variation compared to the wild-type strain (Figures 5 D and E).However, the effect of cobB1 deletion was weaker than that of hupS deletion, suggesting that cobB1 deletion led to only mild nucleoid decompaction.The nucleoid area was restored by complementing the cobB1 deletion with in trans-delivered his -cobB1 gene on the integrative plasmid ( ΔcobB1 + his -cobB1 strain).This observation suggested that in the absence of CobB1 deacetylase, the overacetylated HupS had a lowered capacity for nucleoid compaction.Surprisingly, CobB1 overexpression ( his -cobB1 ↑ strain) did not enhance detectable chromosome compaction (Figure 5 E).
In summary, the analyses of cobB1 deletion and overexpression in S. venezuelae showed that the altered CobB1 levels affect sporogenic development and spores maturation.Increased cobB1 expression lead to growth retardation and influenced differentiation in rich medium.On the other hand, the absence of CobB1, which lead to increased acetylation of HupS, lowered nucleoid compaction in spores and affected spore maturation and their efficiency of germination similarly as hupS deletion.

Discussion
While in bacteria, lysine acetylation was initially found to control acetyl-CoA synthetase (Acs) in Salmonella enterica ( 33 ,59 ), further studies showed a much broader impact of RLAs on cellular processes, including virulence ( 30 ), carbon and nitrogen metabolism ( 60 ,61 ), biofilm formation or motility ( 62 ), and stress response ( 63 ).Surprisingly, although many acetylome studies clearly indicated the high abundance of acetylated proteins in bacteria ( 64 ), including Streptomyces , the involvement of RLAs in the regulation of protein −DNA interactions and chromosome organisation has been scarcely described.However, the recent discovery of histone-like HU homologues being posttranslationally modified revealed novel molecular mechanisms that control bacterial chromosome organization and dynamics ( 28 , 30 , 53 , 65-67 ).
Here, we confirmed that the S. venezuelae HupS protein is subjected to lysine acetylation in vivo, corroborating the previously reported modifications of other HU homologues.Intriguingly, earlier LC −MS analyses of S. griseus and S. coelicolor acetylomes did not detect HupS acetylation in vivo ( 37 ,68 ).Although in our studies we confirmed with high accuracy acetylation of 2 (confirmed in two LC-MS experiments) up to 5 (confirmed at least in one experiment) lysine residues, 4 of them were positioned within, or close to, the N-terminal HU domain.Moreover, the complementary Western blotting analyses indicated acetylation within the LR domain ( Supplementary Figure S2 B).The lack of HupS in Streptomyces acetylomes determined by LC-MS in several studies, including ours, could be explained by the applied procedures in which the proteins were subjected to prolonged trypsin di-gestion ( 68 ).Due to the high abundance of lysine residues in HupS, extensive trypsin digestion could deliver very short peptides with low sequence uniqueness, leading to the discrimination of proteins that are rich in lysine repeats ( 69 ).However, in the acetylome of S. griseus investigated using anti-AcK antibodies, a protein with a molecular mass slightly below 30 kDa was detected to be extensively acetylated ( 68 ).Although not commented on by the authors in the manuscript, their findings resemble our acetylome studies on S. venezuelae (Figure 1 ).
In our studies, we restored the high HupS acetylation state in vitro by using AcP to achieve multiple lysine acetylation, covering also the lysine residues located in the LR domain.Due to LC-MS limitations, we were not able to quantify the number of in vivo acetylated lysine residues in HupS.However, based on the complementary Western blotting analysis, we hypothesize that HupS is also a multi-lysine-acetylated protein, similar to the M. tuberculosis HupB homologue, in which the number of identified acetylated lysine residues varies from 8 residues for endogenous MtHupB ( 28 ) up to 31 if MtHupB was acetylated in vitro by a GNAT-like acetyltransferase named Eis ( 30 ) .Overproduction of the Eis protein in M. smegmatis led to HupB hyperacetylation in vivo ( 28 ).Although Eis can serve as a potent HupB acetylase (with low and / or unknown sequence specificity), nonspecific HupB acetylation in vivo by AcCoA or AcP cannot be excluded.However, the cellular mechanism of HupS acetylation is not known.Since Streptomyces genomes encode up to 72 putative acetyltransferases ( 70 ), the identification of any HupSdedicated acetyltransferase is highly challenging.
Given that lysine-rich LR domains are identified predominantly in DNA-binding proteins ( 27 ) and taking into account that histone tail acetylation affects their interaction with DNA ( 28 ), we expected that the identified HupS acetylation would affect DNA affinity.According to our prediction, the acetylated recombinant His-HupS showed an approximately 2-fold lower affinity for single-and double-stranded DNA in comparison to the nonacetylated protein.Interestingly, acetylation of K86 residue was shown to modify E. coli HU β binding to different DNA substrates, promoting its binding to long DNA fragments and decreasing the affinity for nicked or gapped DNA ( 66 ).Although we detected that HupS shows a preference for supercoiled or single-stranded DNA, corroborating earlier observations in E. coli ( 71 ), the decreased DNA affinity resulting from HupS nonspecific acetylation was comparable for all tested DNA substrates.Moreover, we noticed that His-HupS acetylation did not influence homodimer formation, which could explain the decreased protein −DNA interactions observed for E. coli HU β ( 66 ).
Since the major role of HupS is DNA compaction in spores, HupS levels increase strongly prior to sporulation ( 18 ,19 ), but the exact role of HupS acetylation during Streptomyces development remains unclear.Surprisingly, we found that HupS is constantly acetylated during S. venezuelae development or that acetylation levels even slightly increase during spore maturation when hyphal growth is arrested and following the gradual increase in protein acetylation in S. venezuelae .This observation can be explained by the observation that in general, protein acetylation is strongly induced during sporogenic development in S. griseus ( 68 ) and S. venezuelae (this study).The source of protein acetylation during sporulation is unclear .However , recent studies have shown that AcP is able to acetylate S. coelicolor proteins nonspecifically in vitro ( 72 ) and that AcP levels increase during the stationary phase, which in E. coli leads to the accumulation of acetylated proteins while growth and cell division are arrested ( 73 ).Most bacterial species generate AcP by transferring an acetyl group from AcCoA into inorganic phosphate (Pta-dependent pathway) or by acetate fermentation (AckA-dependent pathway), and both pta and ackA genes are present in Streptomyces genomes.Thus, constant HupS acetylation might be a natural consequence of LR domain exposure to AcP.Lysine susceptibility to nonenzymatic AcP-dependent acetylation is strongly correlated with lysine exposure to the solvent and / or the presence of particular amino acid residues.The proximity of specific amino acids (Thr , Ser , Asn, Gln) promotes AcP coordination ( 74 ).Thus, the long and intrinsically disordered C-terminal domain of Streptomyces HupS or mycobacterial HupB, in which KK doublets are separated by several serine or threonine residues, could serve as a preferable target for AcP-dependent multiple, nonenzymatic lysine acetylation in vivo.Given the increase CobB1 levels during sporulation, we speculate that its major role is presumably preventing nonspecific overacetylation of HupS.Our studies showed that CobB1 is responsible for HupS deacetylation in vitro promoting HupS-DNA interactions.Our observation that acetylated HupS binds DNA with lower affinity corroborates the lowered fraction of DNA-bound protein molecules in S. venezuelae spores as detected in the cobB1 deletion strain.However, since the CobB1 deacetylase activity in vitro was low (also shown by other groups ( 36 ,37 )), we hypothesize that CobB1 activity in vivo could be stimulated by unknown cellular factors including PTMs and / or interactions with other macromolecules or cofactors.
The observation that cobB1 deletion has no effect on S. venezuelae vegetative growth in all tested media, including rich and minimal media, solid and liquid, prompted us to hypothesize that CobB1 activity is particularly important when Streptomyces growth is arrested and the colony differentiates into spores.As assumed, the cobB1 deletion strain had aberrant colony pigmentation similar to those observed in the hupS deletion strain, suggesting the involvement of HupS acetylation / deacetylation state controlled by CobB1 in spore maturation.Moreover, based on the observation that HupS-FLAG does not fully complement hupS deletion phenotype, we speculate that either expression of hupS-flag from different chromosomal locus , or FLAG-tag may have impact on HupS activity in vivo i.e. by affecting protein (including CobB1) accessibility to the LR domain of HupS-FLAG.The particular importance of CobB1 during sporogenic development is also supported by the observation that in S. coelicolor , the CobB1 homologue specifically deacetylates ParB, a segregation protein that assists chromosome distribution to prespores ( 36 ).In S. coelicolor , the lack of cobB1 gene leads to chromosome segregation defects (22.5%) comparable with the parB deletion (17,4%) ( 53 , 75 , 76 ).However, the lysine residue acetylated in S. coelicolor ParB ( 53 ) is not conserved in S. venezuelae , in which it is substituted with positively charged, but acetylation insensitive arginine.These observations suggest, that the role of CobB1 during sporogenic development is not conserved and may differ between particular Streptomyces species.
It should be noted, when CobB1 was overproduced (as a second copy under the constitutive ermE promoter), the impact on vegetative growth was stronger and less clear.We observed strong growth and development inhibition of the CobB1-overproducing strain, but only in a rich medium.This suggests that the overproduction of CobB1 protein can have a pleiotropic effect on cell functioning, which is highly dependent on nutrient availability .Intriguingly , the observation that CobB1 overproduction does not affect the number of germinating spores but decreases the growth rate on rich medium leads us to speculate that CobB1 activity may be also required during germination or, at least, that increased deacetylase activity does not disturb the process.On the other hand, during vegetative growth, when the cobB1 gene is typically downregulated, cobB1 overexpression may lead to the deregulation of cell metabolism.In S. coelicolor , among 601 proteins identified as acetylated, 36.5% are involved in essential cell metabolism, including glycolysis, the tricarboxylic acid (TCA) cycle, and energy metabolism.Thus, the deregulation of the key metabolic pathways, normally controlled by the RLAs, in the presence of high deacetylase activity of CobB1 could explain the growth inhibition of the cobB1 mutant.
In summary, our studies showed that the HupS protein, a close HU homologue, is acetylated in S. venezuelae , decreasing its capacity for DNA interactions in vitro and in vivo .Since we detected that protein acetylation increases during S. venezuelae sporogenic differentiation, HupS overacetylation could negatively affect chromosome compaction in spores, thus requiring CobB1 deacetylase activity.

Data availability
The RNA-Seq data extracted in this study (shown in Figure 3 A and Supplementary Figure S4 C) are available in the Array Express database (EMBL-EBI) under accession number E-MTAB-13607 ( 46 ) and E-MTAB-13911 ( 47 ), respectively.The mass spectrometry proteomics data have been deposited to the deposited to the ProteomeXChange Consortium via the PRIDE partner repository ( 77 ).The LC-MS data shown in Figure 1 A and Supplementary Figure S2 A

Figure 1 .
Figure 1.HupS le v els upregulation and its acetylation during S. venezuelae growth and development.( A ) The scheme of S. venezuelae HupS protein organisation.The HU domain (amino acids 1-89) homologous to the E. coli HU protein is marked in yellow, whereas the C-terminal domain (amino acids 90-234) rich in lysine repeats (LR domain) is marked in white.The lysine repeats (KK doublets) are marked in red.The acetylated lysine residues (K*) identified in the HupS-FLAG pulldown experiment (marked in blue) or in the untagged HupS protein during global proteome studies (marked in grey), or in both LC-MS analyses (marked in green), are indicated.The HupS or HupS-FLAG lysine modifications were studied in S. venezuelae culture growing for 21 h in liquid MYM medium.( B ) Analysis of hupS transcript (in wild-type strain, WT) and HupS-FLAG protein le v els (in hupS-flag strain, TM015) during the S. venezuelae life cycle (8-25 h of growth in MYM medium).The particular growth stages, including vegetative growth (VM), aerial mycelium formation (AM), and spore maturation (SP), are marked.The hupS transcripts were quantified in relation to their levels in 8 hour of growth (estimated as 1, red line).HupS-FLAG protein le v els w ere quantified at corresponding time points using anti-FLA G antibodies ( αFLA G).TopA le v els quantified using anti-TopA serum ( αTopA) served as an internal control for the experiment.The loading control (CBB-stained acrylamide gel) is shown below.( C ) Localization of

Figure 2 .
Figure 2. Acetylation of recombinant His-HupS and its impact on protein −DNA interactions in vitro.( A ) His-HupS (26.1 kDa) purification using ion e x c hange c hromatography.His-HupS protein (blac k arro w) w as eluted using a 3-step NaCl gradient (250 mM (I), 50 0 mM (II), and 80 0 mM (III)) and analysed using SDS −PAGE electrophoresis followed by CBB staining or Western blotting using anti-HisTag antibodies.( B ) Left panel: scheme of His-HupS acetylation in vitro in the presence of 100 mM acetyl phosphate (AcP) followed by protein desalting.Right panel: detection of His-HupS using a dot-blot assay with anti-HisTag (for protein detection, αHisTag) or anti-acetyllysine antibodies (for lysine acetylation detection, αAcK).( C ) LC −MS detection of non-trypsin-digested acetylated (+AcP) and nonacetylated (-AcP) recombinant His-HupS protein.( D ) Binding of acetylated (+AcP, orange) and nonacetylated (-AcP, blue) His-HupS (protein concentrations: 20, 40 and 60 nM) to one-end biotinylated double-stranded DNA (dsDNA, 300 kb) measured using biola y er interferometry (BLI).The binding of nonacetylated His-HupS (40 nM) to a DNA-free sensor (grey) served as a control.The HupS-His association and dissociation steps are marked in the scheme.( E ) The analysis of acetylated and nonacetylated His-HupS protein affinity to 3 types of DNA substrates (one-or two-end biotinylated double-stranded DNA, dsDNA, 300 bp, or one-end biotinylated single-stranded DNA, ssDNA, 117 nt) performed in a broad range of His-HupS protein concentrations (0-100 nM).The quantified His-HupS dissociation constants (K d ) calculated for particular DNA substrates are shown above each plot.( F ) Glutaraldehyde crosslinking of acetylated (+AcP) and nonacetylated (-AcP) His-HupS.The migration of His-HupS monomers and dimers was analysed using SDS-PAGE electrophoresis.The 25, 35 and 70 kDa bands of the protein molecular weight ladder are marked on the left.The glutaraldehyde concentrations (%) are indicated.

Figure 3 .
Figure 3. CobB1 protein deacetylase activity in vivo and in vitro .( A ) Left panel: cobB1 and cobB2 levels (TPM, transcripts per million) based on transcriptome analyses (RNA-Seq data).The analysed S. venezuelae developmental stages, v egetativ e m y celium (VM; 10 h of gro wth in liquid MYM) and aerial m y celium / spore maturation (AM / SP; 20 h of growth in liquid MYM), are indicated.Right panel: Relative cobB1 transcript levels quantified in triplicate using RT-qPCR and RNA extracted at different S. venezuelae developmental stages; v egetativ e m y celium (VM), aerial m y celium (AM) and during spores maturation (SP) after 13, 17 and 21 h of growth in liquid MYM, respectively.( B ) Left panel: Western blotting detection of lysine acetylation in S. venezuelae cell lysates obtained from the hupS (AKO200), wild-type (WT) and cobB1 (JD01) strains.The loading control (CBB-stained acrylamide gel) is shown below.The levels of hupS transcript (transcripts per million, TPM) in the corresponding strains are shown below.Right panel: HupS acetylation quantified in triplicate in the ΔcobB1 and wild-type strains.The statistical significance was tested using an unpaired t test ( P value < 0.01, **).( C ) The activity in vitro of recombinant 3 μg His-CobB1 (4.3 μM) protein against 180 ng of multi-lysine-acetylated (13-17 times on a v erage) His-HupS (0.4 μM) after 6 h of incubation at 37 • C. The estimated molar ratio of AcK sites / His-CobB1 sites was from 1.2 to 1.6 on a v erage.T he His-CobB1 (black arrow) and His-HupS (white arrow) levels were detected using Western blotting and anti-His antibodies (serving as a loading control, left panel, αHisTag).His-HupS acetylation was detected using western blotting and anti-AcK antibodies ( αAcK).Right panel: quantification of the His-HupS acetylation signal in the presence of His-CobB1 in triplicate in relation to the control lacking His-CobB1.The statistical significance was tested using an unpaired t test ( P value < 0.05, *).The 25, 35 and 70 kDa bands of the protein molecular weight ladder are marked on the left.

Figure 5 .
Figure 5.The impact of cobB1 gene deletion on S. v enezuelae gro wth, nucleoid compaction, and germination efficiency.( A ) Analy sis of the gro wth rates of S. venezuelae in liquid MYM.( B ) The growth S. venezuelae strains on solid SFM and MYM media after 120 h.The aberrant colony pigmentation is marked with asterisk.( C ) Quantification of the percent of colony forming units (CFU) in relation to total spore number (germination efficiency) of S. venezuelae strains cultivated on solid MYM medium.Each analysis was performed in 3-4 experimental replicates.Individual data points and standard deviations are marked in each bar.The statistical significance tested using an unpaired t test is marked ( P value < 0.01, **).( D ) Visualization of the nucleoid in S. venezuelae prespores.The cell wall was stained using wheat germ agglutinin (WGA) conjugated with AlexaFluor 350 (blue).Nucleoids w ere visualiz ed using 7-amino-actinom y cin D (7-AAD; red).T he bright field, separate fluorescent channel, and o v erla y of fluorescence on brightfield images are shown as indicated.Scale bar: 1 μm.( E ) Box plot analysis of the nucleoid area [ μm 2 ] in prespores of S. venezuelae stains growing for 22-24 on solid SFM medium.The boxplots show a median with the first and third quartile, while the lower and upper 'whiskers' extend to the value no further than 1.5 * QR (interquartile range) from the 'hinge'.The statistical significance was determined with the Wilcoxon test.The significant differences ( P value < 0.001.***) are marked on the diagram.The presented data were collected in two separate experiments and combined.The number of analyzed spores ( n ) is shown below the box plot.All the presented S. venezuelae strains are listed below: wild-type (WT; red), ΔcobB1 (JD01; violet), ΔcobB1 complementation ( ΔcobB1 + his -cobB1, JD11; blue), CobB1 o v erproduction ( his -cobB1 ↑ , JD04; y ello w), ΔhupS (AKO200; green), and ΔhupS complementation ( ΔhupS + hupS -FLAG, TM015; grey).
data base under accession number PXD050776 .The LC-MS data shown in Figure 2 C and Supplementary Figure S2 C were deposited under accession number PXD048203 .