Spatial Rearrangement of the Streptomyces Venezuelae Linear Chromosome

bioRxiv preprint doi: https://doi.org/10.1101/2020.12.09.403915; this version posted December 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1 Spatial rearrangement of the Streptomyces venezuelae linear chromosome

2 during sporogenic development 3 4 5 Szafran MJ. 1,*, Małecki T.1, Strzałka A.1, Pawlikiewicz K.1, Duława J.1, Zarek A.1, Kois- 6 Ostrowska A.1., Findlay K.2, Le TBK.2, Jakimowicz D. 1,* 7

8 1 Faculty of Biotechnology, University of Wrocław, 50-383 Wrocław, Poland

9 2 The John Innes Centre, Norwich Research Park, NR4 7UH Norwich, The United Kingdom

10 *Corresponding authors: [email protected]; [email protected]

12 ABSTRACT

13 Depending on the species, bacteria organize their chromosomes with either spatially

14 separated or closely juxtaposed replichores. However, in contrast to eukaryotes,

15 significant changes in bacterial chromosome conformation during the cell cycle have not

16 been demonstrated to date. Streptomyces are unique among bacteria due to their linear

17 chromosomes and complex life cycle. These bacteria develop multigenomic hyphae that

18 differentiate into chains of unigenomic exospores. Only during sporulation- associated cell

19 division, chromosomes are segregated and compacted. In this study, we show that at entry

20 to sporulation, arms of S. venezuelae chromosomes are spatially separated, but they are

21 closely aligned within the core region during sporogenic cell division. Arm juxtaposition is

22 imposed by the segregation protein ParB and condensin SMC. Moreover, we disclose that

23 the chromosomal terminal regions are organized into domains by the Streptomyces-

24 specific protein - HupS. Thus, we demonstrate chromosomal rearrangement from open to

25 close conformation during Streptomyces life cycle.

27 INTRODUCTION

28 Chromosomal organization differs notably among the domains of life, and even

29 among bacteria, chromosomal arrangement is not uniform. While linear eukaryotic

30 chromosomes undergo dramatic condensation before their segregation, the compaction of

31 usually circular bacterial chromosomes remains relatively constant throughout the cell

32 cycle1–3. To date, chromosome conformation capture methods, applied in only a few model

33 species, have revealed that bacterial chromosomes adopt one of the two distinct

34 conformation patterns. Either both replichores of circular chromosomes align closely (as in

35 Bacillus subtilis4,5, Caulobacter crescentus6, Corynebacterium glutamicum7, Mycoplasma

36 pneumoniae8, and chromosome II of Vibrio cholerae9) or tend to adopt an “open”

37 conformation, which lacks close contacts between both chromosomal arms (e.g., in

38 Escherichia coli10 and V. cholerae chromosome I9). The cytological studies showed that

39 chromosome organisation may alter depending on the growth phase or developmental

40 stage1,11,12,14. While the most prominent change in bacterial chromosome organization was

41 observed during extensive DNA compaction that occurs during sporulation of Bacillus spp13

42 and Streptomyces spps, the details of chromosome arrangement during this process have

43 not been elucidated.

44 The folding of bacterial genomes requires a strictly controlled interplay between

45 several cellular factors that encompass: macromolecular crowding, transcription and the

46 activities of dedicated DNA-organizing proteins - topoisomerases, condensins and nucleoid-

47 associated proteins (NAPs)10,15,16. By nonspecific DNA binding, NAPs induce its bending,

48 bridging or wrapping to govern the local organization and global compaction of bacterial

49 chromosomes1,10,16. Since the subcellular repertoire of NAPs is dependent on growth

50 conditions, changes in their supply account for the fluctuation of chromosome compaction

51 in response to environmental clues. While, the repertoire of NAPs also varies between

52 bacterial species, some of them, such as HU homologues, are widespread among

53 bacteria14,17,18.

54 In contrast to NAPs, condensins are the most conserved DNA-organizing proteins

55 identified in all kingdoms of life. In bacteria, the homologues of condensin are complexes of

56 SMCs or its E. coli functional analogue MukB, associated with their accessory proteins (ScpAB

57 or MukEF, respectively) 19,20,21. According to a recently proposed model, the primary activity

58 of SMC is the extrusion of large DNA loops22,23. SMC complexes were shown to move along

59 bacterial chromosomes from the centrally positioned origin of replication (oriC), leading to

60 gradual chromosome compaction and parallel alignment of both replichores5,6,24,25.

61 Importantly, although the alignment of two chromosomal arms was not detected in E. coli,

62 the MukBEF complex is still responsible for the long-range DNA interactions within

63 chromosomal domains10,26.

64 In B. subtilis, C. glutamicum, and C. crescentus, the loading of SMC in the vicinity of

65 the oriC was shown to be dependent on the segregation protein ParB4,27,28. Binding to a

66 number of parS sequences, ParB forms a large nucleoprotein complex (segrosome), that

67 interact with the partner protein ParA, ATPase non-specifically bound to nucleoid 29,30. ParB

68 interaction with ParA triggers its ATP hydrolysis and nucleoid release, generating a

69 concentration gradient that drives segrosome separation31–33. Segrosome formation was

70 shown to be a prerequisite for chromosomal arm alignment in the model Gram-positive

71 bacteria with circular chromosomes5,7,28. Importantly, there have been no reports on the 3D

72 organization of bacterial linear chromosomes to date.

73 The example bacteria with linear chromosomes are Streptomyces - Gram-positive

74 mycelial soil actinobacteria, which are highly valued producers of numerous antibiotics and

75 other clinically important metabolites. Notably, the production of these secondary

76 metabolites is closely correlated with Streptomyces unique and complex life cycle, which

77 encompasses vegetative growth and sporulation, as well as exploratory growth in some

78 species14,34,35. During the vegetative stage, Streptomyces grow as branched hyphae

79 composed of elongated cell compartments, each containing multiple copies of nonseparated

80 and largely uncondensed chromosomes14,36. Under stress conditions sporogenic hyphae

81 develop, and their rapid elongation is accompanied by extensive replication of

82 chromosomes37. When hyphae stop extending, the multiple chromosomes undergo massive

83 compaction and segregation to unigenomic spores14. These processes are accompanied by

84 multiple synchronized cell divisions initiated by the FtsZ protein, which assembles into a

85 ladder of regularly spaced Z-rings38,39. Chromosome segregation is driven by ParA and ParB

86 proteins, whereas nucleoid compaction is induced by concerted action of SMC and

87 sporulation-specific NAPs that include HupS - one of the two Streptomyces HU homologues,

88 structurally unique to actinobacteria due to the presence of histone-like domain. The

89 elimination of any of these proteins results in DNA decompaction in spores and affects

90 resistance to environmental factors40–44. Thus, while in unicellular bacteria chromosome

91 segregation occurs during ongoing replication, during Streptomyces sporulation, intensive

92 chromosome replication is temporarily separated from their compaction and segregation,

93 with the two latter processes occurring only during sporogenic cell division.

94 While the Streptomyces chromosomes are among the largest in bacterial kingdom,

95 the knowledge concerning their organization is scarce. Limited cytological evidence suggests

96 that during vegetative growth, both ends of the linear chromosome colocalize, and the oriC

97 region of the apical chromosome is positioned at the tip-proximal edge of the nucleoid45,46.

98 The visualization of nucleoid and ParB-marked oriC regions in sporogenic hyphae showed

99 that oriCs are positioned centrally within pre-spore compartments, but did not elucidate

100 their architecture during sporulation39,46.

101 In this study, using the chromosome conformation capture method (Hi-C), we

102 investigated S. venezuelae chromosome organization during differentiation. The application

103 of the Hi-C technique demonstrated the dramatic rearrangement of chromosome structure

104 associated with sporogenic development. Moreover, by combining the Hi-C method with

105 chromatin immunoprecipitation and fluorescence microscopy techniques, we showed the

106 contribution of ParB, SMC and HupS (the sporulation-dedicated HU homologue) to global

107 nucleoid compaction.

108

109

110 MATERIALS and METHODS

111 Bacterial strains and plasmids

112 The S. venezuelae and E. coli strains utilized in the study are listed in Table S1 and S2

113 respectively, and the plasmids are listed in Table S3. DNA manipulations were performed by

114 standard protocols47. Culture conditions, antibiotic concentrations, transformation and

115 conjugation methods followed the commonly employed procedures for E. coli47 and

116 Streptomyces48. A detailed description of strain construction is presented in the

117 Supplementary Information.

118 S. venezuelae cultures and growth rate analysis

119 To obtain the standard growth curves comparing the growth rate of S. venezuelae

120 strains, a Bioscreen C instrument (Growth Curves US) was used. Cultures (in triplicate) in

121 liquid MYM medium (300 µl) were set up by inoculation with 6 x 103 colony-forming units

122 (CFUs) of S. venezuelae spores. The cultures grew for 48 h at 30°C under the “medium”

123 speed and shaking amplitude settings, and their growth was monitored by optical density

124 measurement (OD600) every 20 min.

125 The Hi-C cultures were set up as follows: 5 ml of MYM medium was inoculated with

126 108 CFU, and the cultures were incubated for 6 to 26 h at 30°C with shaking (180 rpm). For

127 the “5 ml culture” growth curve depiction, the optical density (OD600) of the culture diluted

128 (1:10) with MYM medium was measured in 1-h.

129 Preparation of Hi-C libraries and data analysis

130 For preparation of S. venezuelae Hi-C contact maps, 5 ml cultures were established as

131 described above. The cultures were incubated for 13 to 25 h at 30°C with shaking (180 rpm).

132 After this step, the cultures were cross-linked with 1% formaldehyde for 7 min and 30 s and

133 blocked with 0.2 M glycine for 15 min. One millilitre of the cross-linked culture was

134 centrifuged (1 min, 5000 rpm at room temperature). The mycelium pellet was washed twice

135 with 1.2 ml of TE buffer and resuspended in 0.6 ml of TE buffer. Two vials, each containing

136 25 µl of the resuspended mycelium, were independently processed according to the

137 modified protocol of Le et al.49 as below. Briefly, the cells were lysed using 0.5 µl of

138 ReadyLyse solution (Invitrogen) at 37°C for 30 min followed by the addition of 1.25 µl of 5%

139 SDS solution and incubation at room temperature for 15 min. After cell lysis, the reaction

140 was supplemented with 5 µl of 10% Triton X-100 (Sigma Aldrich), 5 µl of NEB3 buffer (New

141 England Biolabs) and 10.75 µl of DNase-free water (Invitrogen) and subsequently incubated

142 for 15 min at room temperature. The chromosomal DNA was digested with 2.25 µl of BglII

143 restriction enzyme (New England Biolabs; 50000 U/ml) for 2 h and 30 min at 37°C followed

144 by the addition of 0.25 µl of BglII and incubation for another 30 min at 37°C. The subsequent

145 steps of Hi-C library preparation directly followed the protocol described by Le et al.49with

146 the number of amplification cycles being increased to 18. The amplified libraries were

147 purified from a 1% agarose gel, and their concentrations were measured using a Qubit

148 dsDNA HS Assay Kit (Thermo Fisher Scientific) in OptiPlate-96 White (Perkin Elmer) with 490-

149 nm and 520-nm excitation and emission wavelengths, respectively. The agarose-purified Hi-C

150 libraries were diluted to a 10 nM concentration, pooled together, and sequenced using

151 Illumina NextSeq500 (2x75 bp; paired reads), as offered by the Fasteris SA sequencing

152 facility.

153 Complete data processing was performed using the Galaxy HiCExplorer web server50.

154 The paired reads were first trimmed using Trimmomatic (version 0.36.5) and subsequently

155 filtered using PRINSEQ (version 0.20.4) to remove reads with more than 90% GC content.

156 The pre-processed reads were mapped independently to the S. venezuelae chromosome

157 using Bowtie2 (version 2.3.4.3; with the -sensitive -local presets). The Hi-C matrix was

158 generated using hicBuildMatrix (version 2.1.4.0) with a bin size of 10000 bp. In the next step,

159 three neighbouring bins were merged using hicMergeMatrixBins (version 3.3.1.0) and

160 corrected using the hicCorrectMatrix tool (version 3.3.1.0) based on Imakaev’s iterative

161 correction algorithm51 (with the presets as follows: number of iterations: 500; skip diagonal

162 counts: false; remove bins of low coverage: -1.5; remove bins of high coverage: 5.0). The

163 corrected Hi-C matrixes were normalized and compared using hiCCompareMatrixes (version

164 3.4.3.0, presets as below: log2ratio), yielding differential Hi-C contact maps. The differential

165 maps were plotted using hicPlotMatrix (version 3.4.3.0; with masked bin removal and

166 seismic pallets of colours). To prepare the Hi-C contact map, the corrected matrixes were

167 normalized in the 0 to 1 range using hicNormalize (version 3.4.3.0; settings: set values below

168 the threshold to zero: value 0) and subsequently plotted using hicPlotMatrix (version 3.4.3.0;

169 with the removal of the masked bins), yielding 30-kb-resolution Hi-C heatmaps.

170 Chromatin immunoprecipitation combined with next-generation sequencing (Chip-seq)

171 For chromatin immunoprecipitation, S. venezuelae cultures (in triplicate) were grown

172 in 50 ml liquid MYM medium supplemented with TES at 30°C with shaking (ChIP-seq

173 cultures). Cultures were inoculated with 7x108 CFU. After 14 h of growth, the cultures were

174 cross-linked with 1% formaldehyde for 30 min and blocked with 125 mM glycine. Next, the

175 cultures were washed twice with PBS buffer, and the pellet obtained from half of the culture

176 volume was resuspended in 750 l of lysis buffer (10 mM Tris-HCl, pH 8.0, 50 mM NaCl, 14

177 mg/ml lysozyme, protease inhibitor (Pierce)) and incubated at 37°C for 1 h. Next, 200 μl of

178 zircon beads (0.1 mm, BioSpec products) were added to the samples, which were further

179 disrupted using a FastPrep-24 Classic Instrument (MP Biomedicals; 2 x 45 s cycles at 6 m/s

180 speed with 5-min breaks, during which the lysates were incubated on ice). Next, 750 l IP

181 buffer (50 mM Tris-HCl pH 8.0, 250 mM NaCl, 0.8% Triton X-100, protease inhibitor (Pierce))

182 was added, and the samples were sonicated to shear DNA into fragments ranging from 300

183 to 500 bp (verified by agarose gel electrophoresis). The samples were centrifuged, 25 l of

184 the supernatant was stored to be used as control samples (“input”), and the remaining

185 supernatant was used for chromatin immunoprecipitation and mixed with magnetic beads

186 (60 l) coated with anti-FLAG M2 antibody (Merck). Immunoprecipitation was performed

187 overnight at 4°C. Next, the magnetic beads were washed twice with IP buffer, once with IP2

188 buffer (50 mM Tris-HCl pH 8.0, 500 mM NaCl, 0.8% Triton X-100, protease inhibitors

189 (Pierce)), and once with TE buffer (Tris-HCl, pH 7.6, 10 mM EDTA 10 mM). DNA was released

190 by overnight incubation of beads resuspended in IP elution buffer (50 mM Tris-HCl pH 7.6,

191 10 mM EDTA, 1% SDS) at 65°C. The IP elution buffer was also added to the “input” samples, 8

192 which were treated further as immunoprecipitated samples. Next, the samples were

193 centrifuged, and proteinase K (Roche) was added to the supernatants to a final

194 concentration of 100 µg/ml followed by incubation for 90 min at 55°C. DNA was extracted

195 with phenol and chloroform and subsequently precipitated overnight with ethanol. The

196 precipitated DNA was dissolved in nuclease-free water (10 µl). The concentration of the DNA

197 was quantified using a Qubit dsDNA HS Assay Kit (Thermo Fisher Scientific).

198 DNA sequencing was performed by Fasteris SA (Switzerland) using the Illumina ChIP-

199 Seq TruSeq protocol, which included quality control, library preparation and sequencing

200 from both ends (2x75 bp) of amplified fragments. The bioinformatic analysis was performed

201 using the R packages edgeR, normr, rGADEM, and csaw, as well as the MACS2 and MEME

202 programs. HupS-FLAG binding was analysed using an earlier published protocol52. The

203 mapping of ChIP-seq data was performed using the Bowtie2 tool (version 2.3.5.153,54). The

204 successfully mapped reads were subsequently sorted using samtools (version 1.10)55. The

205 total number of mapped reads was above 106 on average. Regions differentially bound by

206 HupS-FLAG were identified using the R packages csaw and edgeR56–59, which also normalizes

207 the data, as described earlier by Lun and Smyth52. The mapped reads were counted within a

208 69-bp-long sliding window with a 23-bp slide. Peaks were filtered using the local method

209 from the edgeR package, which compares the number of reads in the region to the

210 background computed in the surrounding window of 2000 bp. Only regions with log2 fold

211 change above 2.0 were utilized for further analysis. Regions that were less than 100 bp apart

212 were merged, and the combined p-value for each was calculated. Only regions with false

213 discovery rate (FDR) values below the 0.05 threshold were considered to be differentially

214 bound. The identified regions were further confirmed by analysis with the MASC2 program

215 using the broad option60. The R packages rGADEM and MEME Suite were used to search for

216 probable HupS binding sites61. SMC-FLAG bound regions were analysed using the R package

217 normr using Input ChIP-seq data as a control. Reads were counted in 250-bp-long regions,

218 and a region was considered to be bound by SMC-FLAG if FDR value compared to Input was

219 below the 0.05 threshold, and if it was not found to be significant in a control strain not

220 producing SMC-FLAG protein62.

221 Quantification of the oriC/ter ratio

222 To estimate the oriC/ter ratio, chromosomal DNA was extracted from S. venezuelae 5

223 ml cultures (as for Hi-C) growing for 13-26 h or from 50 ml culture (as for ChIP-seq) growing

224 for 8 to 16 h. One millilitre of each culture was centrifuged (1 min at 5000 rpm and at room

225 temperature). The supernatant was discarded, and the mycelium was used for the

226 subsequent isolation of chromosomal DNA with application a Genomic Mini AX Streptomyces

227 kit (A&A Biotechnology) according to the manufacturer’s protocol. The purified DNA was

228 dissolved in 50 µl of DNase-free water (Invitrogen). The DNA concentration was measured at

229 260 nm and subsequently diluted to a final concentration of 1 ng/ml.

230 qPCR was performed using Power Up SYBR Green Master Mix (Applied Biosystems)

231 with 2 ng of chromosomal DNA serving as a template for the reaction and oligonucleotides

232 complemented to the oriC region (oriC) (gyr2_Fd, gyr2_Rv) or to the termination region (ter)

233 (arg3_Fd, arg3_Rv)(Table S4). The changes in the oriC/ter ratio were calculated using the

234 comparative ΔΔCt method, with the ter region being set as the endogenous control. The

235 ori/ter ratio was estimated as 1 in the culture (5 ml) growing for 26 h, corresponding to the

236 appearance of spore chains.

237 Light Microscopy

238 The analyses of spore sizes were performed using brightfield microscopy. Spores

239 from 72-h plate cultures (MYM medium without antibiotics) were transferred to microscopy

240 slides. The samples were observed using a Zeiss Microscope (Axio Imager M1), and images

241 were analysed using Fiji software. The lengths of 300 spores were measured for each strain.

242 For the snapshot analysis of S. venezuelae development in liquid, 5 µl from the 5 ml

243 culture (as for Hi-C) growing for 18 to 26 h or from the 50-ml culture (as for ChIP-Seq)

244 growing for 8 to16 h was spread on a coverslip and fixed by washing twice with absolute

245 methanol. The nucleoids were subsequently stained for 5 min at room temperature with 7-

246 amino-actinomycin D (1 mg/ml 7-AAD in DMSO, Thermo Fisher Scientific) diluted 1:400 in

247 PBS buffer. In the next step, the mycelium was washed twice with PBS buffer, and the

248 coverslip was mounted using 50% glycerol in PBS. Microscopic observation of FtsZ-YPET and

249 DNA stained with 7-AAD was performed using a Zeiss Microscope (Axio Imager M1). The

250 images were analysed with dedicated software (Axio Vision Rel software). To calculate the

251 nucleoid separation, the distance between the two neighbouring DNA-free areas was

252 measured for 250 nucleoids. The boxplots showing the length of nucleoid-covered areas

253 were constructed using the Plotly tool (plotly.com).

254 Time-lapse fluorescence microscopy was performed using a previously described

255 protocol38 with modification of the applied pressure (3 psi), increased dilution of the loaded

256 spores, and application of repeated spore loading. The observation was performed using a

257 DeltaVision Elite Ultra High-Resolution microscope equipped with a camera and the CellAsic

258 Onix microfluidic system using B04A plates (ONIX). The images were taken every 10 min for

259 approximately 20 h using DV Elite CoolSnap and 100X, 1.46 NA, DIC lens for specimens

260 exposed for 50 ms at 5% transmission in the DIC channel and for 80 ms at 50% transmission

261 in the YFP and mCherry channels. The time-lapse movies were analysed using ImageJ Fiji

262 software. Nucleoid area was measured using HupA-mCherry as the marker at the time of

263 disassembly of Z-rings (visualized using FtsZ-Ypet fusion), which was approximately 120 min

264 after their appearance and hyphal growth cessation. Analysis of the HupA-mCherry-labelled

265 chromosome images was based on determining the area and Feret diameter (the maximal

266 distance between two points) of nucleoid fluorescence using the Versatile Wand Tool with 8-

267 connected mode and manually adjusted value tolerance. The distances between Z-rings

268 were measured 60 min after hyphal growth cessation for approximately 300 Z-rings. The

269 fluorescence intensity of the Z-rings was collected along and averaged across the hyphae

270 with a line selection tool whose width was adjusted to the width of the measured hyphae.

271 The fluorescence intensity of each Z-ring in 25-30 hyphae of each strain was subsequently

272 determined by analysing the data with RStudio using a custom shiny application findpeaks

273 based on the Peaks R package. Tracking of Z-rings in the time-lapse images was performed

274 using the nearest neighbour algorithm. Only Z-rings observed for more than 90 min were

275 used in the analysis. The data from each time point were combined to show how the

276 standard deviation of fluorescence intensity changed during hyphal development.

277 Representative hyphae of each strain were selected, the FtsZ-YPet fluorescence intensity

278 and Z-ring position were determined using the script described above, and its variability was

279 visualized in the form of a kymograph with a custom script based on the ggplot2 package.

280 Scanning Electron Microscopy

281 Streptomyces samples were mounted on an aluminium stub using Tissue TekR OCT

282 (optimal cutting temperature) compound (BDH Laboratory Supplies, Poole, England). The

283 stub was then immediately plunged into liquid nitrogen slush at approximately -210°C to

284 cryo-preserve the material. The sample was transferred onto the cryo-stage of an ALTO 2500

285 cryo-transfer system (Gatan, Oxford, UK) attached to an FEI Nova NanoSEM 450 (FEI,

286 Eindhoven, The Netherlands). Sublimation of surface frost was performed at -95oC for

287 approximately three minutes before sputter coating the sample with platinum for 150

288 seconds at 10 mA, at colder than -110°C. After sputter-coating, the sample was moved onto

289 the cryo-stage in the main chamber of the microscope, held at -125°C. The sample was

290 imaged at 3 kV and digital TIFF files were stored.

291

292 RESULTS

293 Hi-C reveals an alignment chromosomal arms extending to terminal domains during S. 294 venezuelae sporogenic cell division 295 296 During Streptomyces sporulation, their multigenomic hyphae undergo considerable

297 morphological changes turning into a chains of spores through multiple cell divisions. This

298 event is linked to inhibition of DNA replication followed by synchronized compaction and

299 segregation of a dozen chromosomes38,39,64–66. To explore whether the changes in

300 chromosome compaction during sporulation are reflected in its global organization, we

301 applied the chromosome conformation capture technique (Hi-C), which enabled us to

302 establish a map of DNA contact frequency6,67,68.

303 First, we determined the time points of critical sporulation events for S. venezuelae,

304 namely, the reduction of hyphal growth and cessation of DNA replication, chromosome

305 compaction and separation, formation of the FtsZ ladder, and appearance of spore chains

306 (Fig. 1A and S1). After 13-14 h of growth (under 5 ml culture conditions), we observed

307 vegetative growth inhibition. The slowed hyphal growth correlated with the reduction in the

308 oriC/ter ratio corresponding to decreased DNA replication (Fig. 1A). These events indicated

309 culture entry into a developmental stage that we named the pre-sporulation phase. During

310 this phase, which lasted 5 h, the DNA replication rate transiently increased (at the 19th-21st

311 hour). This increased replication rate was observed during a slight increase in the optical

312 density of the culture, which could be explained by the rapid growth of sporogenic hyphae

313 (Fig. 1A). The subsequent drop in the replication rate was accompanied by an increase in the

314 FtsZ-YPet level (at the 21st-23rd hour), assembly of FtsZ-YPet into ladders of Z-rings (detected

315 at the 22nd hour), and initiation of chromosome compaction (Figs. 1A and S1). At this time

316 point, the spore chains entered the maturation phase, during which chromosome

317 condensation and separation increased, while the FtsZ-YPet level rapidly decreased and the

318 hyphae fragmented. This stage ended when mature spores could be detected (at the 25th-

319 26th hour) (Fig. 1A and Fig. S1). Thus, we established that 22nd hour of S. venezuelae culture

320 is the time point when multigenomic hyphae undergo synchronized cell division, replication

321 of chromosomes ceases, while their condensation increases, which minimizes

322 interchromosomal contacts. Based on these findings, we set out to establish the

323 chromosome conformation at this time point.

324 Hi-C mapping of the chromosomal contacts in S. venezuelae sporogenic hyphae

325 undergoing cell division indicated the presence of two diagonal axes crossing

326 perpendicularly in the proximity of the centrally located oriC region (Fig. 1B). The primary

327 diagonal corresponds to the contact frequency between neighbouring chromosomal loci. We

328 distinguished (based on PCA1/PCA2 analysis) three distinct chromosome regions - the core

329 region that encompasses 4.4 Mbp (position: 1.9 Mbp to 6.3 Mbp) around the centrally

330 positioned oriC and two well-organized domains flanking the chromosome core, which we

331 named the left and right terminal domains (LTD and RTD, respectively), with each measuring

332 approximately 2 Mbp in size (Fig. 1B and S2A). Interestingly, the boundaries of LTD and,

333 especially, RTD overlap the chromosomal regions with significantly lower transcriptional

334 activity in comparison to the chromosome core region (Fig. S2B), suggesting a correlation

335 between chromosome organization and transcription. The secondary diagonal, that results

336 from the interarm interactions, was not complete during cell divisions and was limited in

337 each direction to approximately 2.0-2.2 Mbp from the oriC region (Fig. 1B). Additionally, the

338 Hi-C contact map showed signals at the ends of the secondary diagonal, suggesting the

339 existence of the spatial juxtaposition of terminal ends of the linear S. venezuelae

340 chromosome (Fig. 1B).

341 In summary, our results demonstrated that at the time of S. venezuelae synchronized

342 cell division, the linear chromosomes are arranged with both arms closely aligned within the

343 core region, while the terminal regions form distinct domains.

344

345 Arm juxtaposition progresses during sporulation and requires segregation protein activity

346 Having established chromosomal contact maps during S. venezuelae sporogenic cell

347 division, we set out to characterize the changes in chromosomal arrangement during

348 sporogenic development. To this end, we obtained normalized Hi-C contact maps for the

349 wild type strain at the entry into the pre-sporulation phase when vegetative growth and

350 DNA replication decrease (at the 13th hour) and during the phase when the chromosomes

351 are not fully compacted (at the 15th and 17th hours), as well as during at the final stage of

352 spore maturation, which is when chromosome compaction reaches a maximum (Fig. 1A and

353 S1).

354 At the entry into the pre-sporulation phase (at the 13th hour), only the primary

355 diagonal was clearly visible, whereas the secondary diagonal was barely detectable (Fig. 1C).

356 Additionally, the juxtaposition of chromosomal termini was also not, or only slightly,

357 noticeable at this stage of growth. During the pre-sporulation phase, the signals of the

358 secondary diagonal were enhanced and extended at distances of up to 1 Mbp from the oriC

359 region at the 15th and 17th hours of growth (Fig. 1C). Finally, in the mature spores (at the 25th

360 hour), we observed almost complete arm alignment, similar to that detected at 22nd hour

361 (Fig. 1C). This result suggests that although nucleoid compaction was still underway until the

362 end of the spores maturation phase (Fig. S2B), the global chromosome organization did not

363 change critically after that established during cell division (Fig. 1C).

364 In model bacteria whose chromosomal arms are juxtaposed, their close positioning is

365 imposed by ParB-dependent SMC loading within the oriC region4,7,28. During Streptomyces

366 sporogenic cell division, at the time of observed chromosomal arm juxtaposition, ParB was

367 shown to assemble into regularly spaced complexes that position oriC regions along the

368 hyphae69. At the same time, ParA accumulates in sporogenic hyphae and facilitates the

369 efficient formation and segregation of ParB complexes39,70. The elimination of S. venezuelae

370 segregation proteins disturbed chromosome segregation39 but did not affect the growth rate

371 (Fig. S5A). To verify whether segregation proteins play a role in the arrangement of S.

372 venezuelae chromosomes, we obtained Hi-C maps of chromosomal contacts in sporogenic

373 hyphae of parA or parB mutants.

374 The Hi-C map generated for the parB mutant showed complete disappearance of the

375 secondary diagonal axis, reinforcing the role of the ParB complex in inducing interarm

376 contacts (Fig. 1D). The same lack of secondary diagonal was observed for the parB mutant at

377 each time of S. venezuelae sporulation (Fig. S3B). The influence of the parA deletion on arm

378 alignment was also detectable, although it varied among the samples analysed (Fig. 1C and

379 S3C). The strength of the signals at the diagonal axis was either significantly or only slightly

380 lowered in the parA mutant, as indicated by the differential Hi-C map in comparison to the

381 wild type strain (Fig. 1C and S3C). While the effect of parA or parB deletions was

382 predominantly manifested by the partial or complete disappearance of the interarm

383 contacts, respectively, their influence on the local chromosome structure, including LTD and

384 RTD folding, was marginal (Fig. 1C and S3).

385 In summary, these results indicated that chromosomal arm alignment progressed

386 during pre-sporogenic and sporogenic hyphal development (Fig. 1C, inset) and chromosomal

387 rearrangement was dependent on ParB.

388

389 SMC- and HupS-induced chromosome compaction permits regularly spaced cell division

390 In numerous bacteria, the role of ParB in chromosome organization is associated with

391 SMC loading which, on the other hand, promotes global chromosome compaction. Prior S.

392 coelicolor studies showed that elimination of either SMC- or Streptomyces-specific

393 sporulation-associated NAP-HupS affected nucleoid compaction in spores40,42,71. We

394 predicted that SMC would be responsible for ParB-induced chromosomal arm alignment, but

395 since ParB elimination did not influence short-range chromosome interactions, we

396 hypothesized that other DNA organizing factors, such as HupS, may also contribute to

397 chromosome compaction in sporogenic hyphae. Therefore, in this study, we analysed

398 sporulation and chromosome organization in hupS and smc mutant strains of S. venezuelae.

399 The disruption of either the hupS or smc gene in S. venezuelae did not significantly

400 affect the growth rate or differentiation of these bacteria; however, the double hupS smc

401 mutant growth rate was slower than that of single mutants (Fig. S4A, inset and S4B).

402 Notably, the elimination of hupS or, to a lesser extent, smc increased spore size, and the

403 double hupS smc deletion led to more pronounced disturbances in spore size than single

404 deletion (Fig. 2A and 2B). The hupS or smc deletion phenotype, manifested by spore length,

405 was restored by complementation of deletion strains with in trans-delivered hupS-FLAG or

406 smc-FLAG genes, respectively (Fig. 2B).

407 Next, to visualize nucleoid compaction and segregation during cell division, we

408 observed sporogenic hyphal development of smc and/or hupS mutant strains labelled using

409 HupA-mCherry and FtsZ-YPet (Fig. 2C and S4C, Supplementary Movies 1-4). None of the

410 mutant strains formed anucleate spores, but all of them showed significantly increased

411 nucleoid areas, and this phenotype was the most pronounced in the double hupS smc

412 mutant, where notably enlarged spores could be observed (Fig. 2A, C and D, yellow

413 arrowhead). The increased nucleoid area was accompanied by increased distances between

414 Z-rings (Fig. S4D). The fraction of elongated (longer than 1.3 m) pre-spore compartments

415 was the highest in the double hupS smc mutant and in the smc mutant (Fig. S4D).

416 Additionally, the intensity of Z-rings in single smc or the double mutant was more varied

417 along the hyphae than in the wild type strain, indicating their lowered stability (Fig. 2E and

418 F). The disturbed positioning and stability of Z-rings explains the presence of elongated

419 spores in mutant strains.

420 In summary, both SMC and HupS contribute to chromosome compaction in spores.

421 The decreased chromosome compaction in the mutant strains correlated with decreased Z-

422 ring stability and aberrant positioning which, in all studied mutants, resulted in increased

423 distance between septa and elongated spores.

424

425 SMC and HupS collaborate during sporogenic chromosome rearrangement

426 Having confirmed that both SMC and HupS mediate global chromosome compaction

427 during S. venezuelae sporulation, we investigated how both proteins contribute to

428 chromosomal contacts during sporulation-associated chromosome rearrangement.

429 The Hi-C contact map obtained for the smc mutant at sporogenic cell division (at the

430 22nd hour) showed the complete disappearance of the secondary diagonal (similar to the

431 map obtained for the parB mutant), indicating abolished chromosomal arm interactions (Fig.

432 1B and 3A). On the other hand, the elimination of smc resulted in a slight increase in short-

433 range DNA contacts along the primary diagonal, whereas in the parB mutant, the short-

434 range DNA contacts were not affected (Fig. 1B, 3A and S5).

435 Next, to confirm that ParB promotes the interaction of SMC with DNA, we analysed

436 FLAG-SMC binding in the wild type background and in the parB mutant. The specific

437 developmental time points of ChIP-seq experiment cultures (50-ml culture conditions) were

438 identified, corresponding to the earlier described time points used for the Hi-C experiments

439 (Fig. 1A and S6A). At the 12th hour of “ChIP-seq culture”, DNA replication decreased

440 (corresponding to the 15th-17th hour of “Hi-C culture”), while at the 14th hour of “ChIP-seq

441 culture”, DNA compaction and the formation of regularly spaced Z-rings were clearly

442 detected (Fig. S6B) (corresponding to the 22nd hour in “Hi-C culture” conditions), indicating

443 sporogenic cell division and chromosome segregation. Notably, the level of FLAG-SMC

444 remind constant during sporogenic development (Fig. S9C).

445 The analysis of DNA fragments bound by FLAG-SMC at the time of sporogenic cell

446 division did not identify any specific binding sites for FLAG-SMC, suggesting a lack of

447 sequence preference. However, the quantification of the detected SMC binding sites along

448 the chromosome showed that their frequency increased up to twofold in the chromosomal

449 core region (Fig. 3B and S7A). The increased frequency of SMC binding overlapped with the

450 secondary diagonal identified in the Hi-C experiments (Fig. 3A, 3B and S7B). SMC binding

451 was severely diminished in the parB mutant (Fig. 3B), demonstrating that the ParB complex

452 is a prerequisite for SMC loading.

453 The Hi-C contact map generated for the hupS mutant at the time sporogenic cell

454 division (at the 22nd hour) revealed the presence of both diagonals. However, the identified

455 local signals within both diagonals were more dispersed than in the wild type, and diagonals

456 were visibly broadened. Moreover, along the primary diagonal, a less clear distinction

457 between the core region and terminal domains could be observed. The effect of hupS

458 deletion, though affecting whole chromosome compaction, was most prominent within

459 terminal domains, especially within RTD (Fig. 3A).

460 To confirm HupS involvement in the organization of terminal domains and to test the

461 cooperation between HupS and SMC, we analysed HupS-FLAG binding in the complemented

462 hupS and double hupS smc deletion strains at the time of cell division. Notably, in contrast to

463 SMC, HupS-FLAG showed low sequence binding preference, and we identified 307 regions

464 differentially bound by HupS-FLAG with a mean width of 221 bp (Fig. 3C, D and S8A). The

465 analysis of the number of HupS-FLAG binding sites exhibited a significant increase in the

466 terminal domains, especially the right terminal domain, but HupS binding was also detected

467 in the core region of the chromosome (Fig. 3C). The enhanced HupS binding at chromosomal

468 termini is consistent with the lowered number of chromosomal contacts in RTD observed in

469 the Hi-C maps obtained for the hupS mutant.

470 Further analysis of ChIP-seq data performed using two independent methods

471 (rGADEM and MEME) identified the motif recognized by HupS-FLAG (Fig. S8B). Although the

472 motif was observed to be uniformly widespread along the chromosome, and was not

473 required for HupS binding, the analysis of probability showed its enrichment in HupS binding

474 sites versus random regions and its preferred occurrence in the middle of the identified

475 HupS-bound region (confirmed with CentriMo) (Fig. S8B-E). Additionally, in vitro studies

476 showed that immobilized GST-HupS bound digested cosmid DNA with no sequence

477 specificity (Fig. S8F). Markedly, ChIP-seq analysis of HupS-FLAG binding in the absence of

478 SMC (hupS smc deletion complemented with hupS-FLAG) showed a lower number of reads

479 for all HupS-FLAG binding sites in the absence of SMC, indicating that SMC enhanced the

480 HupS-DNA interaction (Fig. 3E). This observation could not be explained by the low level of

481 HupS-FLAG in the smc deletion background (Fig. S9A). We also confirmed earlier

482 observations that the HupS-FLAG level during the pre-sporulation phase was lower than that

483 observed during sporogenic cell division42, which corresponded with the marginal effect of

484 hupS deletion on chromosome organization at the early pre-sporulation phase, as indicated

485 by the Hi-C contact map for the hupS mutant at this time point (Fig. S9B).

486 Since ChIP-seq analysis indicated that the cooperation between HupS and SMC

487 binding and double hupS smc deletion most severely affected chromosome compaction, we

488 investigated how double deletion affects chromosomal arrangement. The Hi-C contact map

489 generated for the double hupS smc mutant at the time of sporogenic cell division (at the 22nd

490 hour) showed the combination of effects observed independently for hupS and smc single

491 mutations, namely, the disruption of arm alignment and the destabilization of local DNA

492 contacts (Fig. 3A). Thus, while long-range interarm interactions are executed by SMC, short-

493 range compaction is governed by HupS, and the effect of the elimination of both proteins is

494 additive.

495 In summary, the Hi-C and ChIP-seq analyses of smc and hupS mutants showed their

496 contribution to global chromosome compaction. While SMC binds in the chromosomal core

497 and is responsible for the juxtaposition of arms, HupS binds in terminal domains and

498 organizes them. The cooperation between both proteins is manifested by SMC enhancing

499 HupS binding to DNA. Elimination of both proteins disturbs global chromosome organization.

500

501 DISCUSSION

502 In this study, we provide a complex picture of the dynamic rearrangement of linear

503 bacterial chromosome. The Hi-C mapping of the ~8.2 Mbp S. venezuelae chromosome at the

504 time of sporogenic cell division revealed arm juxtaposition, which extends approximately 4

505 Mbp around the oriC region. The interarm contacts are diminished at the independently

506 folded 1.5 and 2 Mbp C-terminal domains LTD and RTD, respectively. Further analysis of

507 chromosomal contacts shows spatial proximity of both chromosomal termini, which is

508 consistent with the findings of previous cytological studies45. Folding of the Streptomyces

509 linear chromosome into core and terminal domains was also independently demonstrated

510 for the S. ambofaciens (see Lioy et al., submitted back to back).

511 Our Hi-C data disclosed profound rearrangement of the chromosomal conformation

512 that occurs during S. venezuelae sporulation. At the end of vegetative growth, at the

513 entrance to the sporulation phase contacts between the chromosomal arms are scarce.

514 After this time point, the chromosomal arms become increasingly aligned, reaching the

515 maximum at the time of initiation of sporogenic cell division (Z-ring formation) which is also

516 the time point when chromosome segregation is initiated39. Thus, during Streptomyces

517 differentiation, the changes in hyphal morphology, gene expression, chromosome

518 replication and mode of cell division14,34 are accompanied by the rearrangement of

519 chromosomes from their extended conformation during vegetative growth to almost fully

520 folded and compacted structures.

521 Our data showed that in S. venezuelae, both ParB and SMC are essential for

522 chromosomal arm juxtaposition, reinforcing the ParB-dependent SMC loading onto the

523 chromosome. Importantly, in S. venezuelae, SMC binding is enriched within the core region

524 of the chromosome, reflecting the extension of interarm contacts. SMC was previously

525 shown to be involved in chromosome compaction during S. coelicolor sporulation40,71.

526 Interestingly, parB deletion was also observed to decrease chromosome condensation at the

527 time of cell division39. This observation may be explained by the lack of SMC loading in

528 absence of ParB, which was also reported for the other model bacteria4,6,7,9. The increase in

529 long-range chromosomal contacts induced by SMC in the chromosomes of S. venezuelae and

530 other bacteria corroborates proposed model of SMC action which zips DNA regions when

531 translocating from the loading site4.

532 While the increase of arm alignment during sporulation cannot be explained by

533 elevation of SMC level during sporulation, the parAB genes were earlier shown to be

534 transcriptionally induced in Streptomyces sporogenic hyphae65,75. ParB binds to oriC regions

535 of all chromosomes throughout the whole life cycle but complexes formed during vegetative

536 growth and at the early stages of sporogenic development are not as massive as regularly

537 spaced segrosomes that accompany sporogenic-associated cell division70. Since interarm

538 contacts occur during cell division, characteristic spatial architecture of sporogenic

539 segrosomes is implied to be a prerequisite for SMC loading. The significance of the ParB

540 complex architecture for SMC loading has been established by studies of ParB mutants with

541 impaired spreading or bridging in B. subtilis and in C. glutamicum7,27. Therefore, our

542 observations suggest that during Streptomyces differentiation ParB complexes spatially

543 rearrange to increase their capability of SMC loading and initiating chromosomal arm

544 juxtaposition.

545 What factors could induce the reorganization of ParB complexes during Streptomyces

546 sporulation? The recently shown ParB ability to bind and hydrolyse CTP that affects

547 formation of segrosomes and appears to be conserved among ParB homologs76–78 while not

548 yet confirmed in Streptomyces may be involved in the rearrangement of ParB complexes in

549 their sporogenic hyphae. On the other hand, the formation of ParB complexes was also

550 shown to be associated with sporulation-dependent deacetylation of ParB in S. coelicolor82.

551 The factor contributing to segrosomes rearrangement may be the increased transcriptional

552 activity of the parAB operon. ParA accumulates in sporogenic cells, promoting segrosome

553 formation and their regular distribution39,70. In other bacteria (B. subtilis, V. cholerae, and M.

554 smegmatis) the elimination of ParA also decreased ParB binding79–81. Indeed, our Hi-C

555 mapping showed that in the absence of ParA, the interarm interactions are somewhat

556 diminished, reinforcing the role played by ParA in executing ParB architecture. Finally, since

557 transcriptional activity was proven to have a considerable impact on chromosomal arm

558 alignment5,83 (see also Lioy et al., submitted back to back), sporulation-associated changes in

559 transcription within the chromosomal core may account for the development of interarm

560 contacts.

561 We established that HupS, unique actinobacterial sporulation-specific HU

562 homologue42, although promotes contacts all over the chromosome, is particularly involved

563 in the organization of the terminal domains. HupS exhibited binding sequence preference in

564 vivo, but no sequence specificity when binding linear DNA fragments. Thus, we infer that

565 specific DNA structural features may account for enhanced HupS binding in particular

566 regions. Lowered HupS binding in the absence of SMC could be explained by preferential

567 binding by HupS to structures such as loops, generated by SMC. This notion corroborates

568 earlier suggestions that HU homologues could stabilize DNA plectonemic loops, increasing

569 short-range contacts6. E. coli studies also suggested cooperation between condensin MukB

570 and HU homologues in maintaining long-range chromosomal contacts10. In fact, in E. coli,

571 condensin activity was suggested to be increased by DNA structures induced by HU

572 protein10. Previous analysis of S. coelicolor showed that HupS deletion resulted in

573 chromosome decompaction in sporogenic hyphae42. The Hi-C matrixes established that both

574 proteins, SMC and HupS contribute to chromosome compaction during sporulation, and the

575 double hupS smc deletion has the strongest effect on chromosomal conformation, affecting

576 both arm juxtaposition and short-range chromosomal contacts.

577 Although elimination of SMC and HupS disturbed the S. venezuelae chromosome

578 structure, the phenotype of the double mutant was still rather mild and manifested

579 predominantly in increased nucleoid volume and formation of elongated spores. While the

580 mild phenotype of the smc mutant was also described in C. glutamicum, in most bacteria

581 (e.g., B. subtilis. C. crescentus, and S. aureus), smc deletion resulted in either chromosome

582 decompaction and/or segregation defects and often led to the formation of elongated

583 cells7,84,85. Interestingly, condensins were suggested to be particularly critical for

584 chromosome segregation in bacteria lacking the complete parABS system (such as E. coli)86–

585 88. The observed in S. venezuelae smc mutants aberrations of spore size correlate with

586 irregular placement and lowered stability of Z-rings, suggesting that decreased chromosome

587 compaction affects Z-ring formation and/or stability.

588 In summary, we established maps of chromosomal contacts for linear bacterial

589 chromosomes. The S. venezuelae chromosome conformation undergoes extensive

590 rearrangement during sporulation. Similar chromosomal rearrangement was observed

591 during the transition from the vegetative to stationary phase of growth in S. ambofaciens

592 (Lioy et al. submitted back to back). DNA organizing proteins HupS and SMC contribute to

593 chromosome compaction, and they also impact cell division. Although HU and SMC roles

594 have been established for numerous bacteria, in Streptomyces, their function is seemingly

595 adjusted to meet the requirements of compacting the large linear chromosomes. Due to the

596 unique Streptomyces differentiation, in which chromosome replication and segregation are

597 spatiotemporally separated, the transition between the particular stages of life cycle

598 requires the chromosomal rearrangement similarly profound as rearrangement of

599 chromosomes during eukaryotic cell cycle.

600

601 DATA DEPOSITION

602 Raw data are available at ArrayExpress (EMBL-EBI) with accession numbers: E-MTAB-9810

603 (Hi-C data), E-MTAB-9821 (ChIP-Seq data, HupS-FLAG immunoprecipitation), and E-MTAB-

604 9822 (ChIP-Seq data, SMC-FLAG immunoprecipitation).

605

606 ACKNOWLEDGEMENTS

607 We are grateful to Susan Schlimpert for sharing pSS170 and pSS172hupA-mcherry. We thank

608 the Bioimaging facility of the John Innes Centre, supported by a core capability grant from

609 BBSRC, for use of their resources. T.B.K.L. acknowledges the Royal Society University

610 Research Fellowship (UF140053). This work was funded by the Polish National Science

611 Centre: HARMONIA grant 2016/22/NZ1/00122 (to M.J.S.), OPUS grant

612 2018/31/B/NZ1/00614 (to D.J.).

613

614

615 LITERATURE 616 1. Badrinarayanan, A., Le, T. B. K. & Laub, M. T. Bacterial chromosome organization and 617 segregation. Annu. Rev. Cell Dev. Biol. 31, 171–199 (2015). 618 2. Wang, X., Montero Llopis, P. & Rudner, D. Z. Organization and segregation of bacterial 619 chromosomes. Nat. Rev. Genet. 14, 191–203 (2013). 620 3. Wang, X. & Rudner, D. Z. Spatial organization of bacterial chromosomes. Curr. Opin. 621 Microbiol. 22, 66–72 (2014). 622 4. Wang, X. et al. Condensin promotes the juxtaposition of DNA flanking its loading site in 623 Bacillus subtilis. Genes Dev. 29, 1661–1675 (2015). 624 5. Wang, X., Brandão, H. B., Le, T. B. K., Laub, M. T. & Rudner, D. Z. Bacillus subtilis SMC 625 complexes juxtapose chromosome arms as they travel from origin to terminus. Science 626 355, 524–527 (2017). 627 6. Le, T. B. K., Imakaev, M. V., Mirny, L. A. & Laub, M. T. High-resolution mapping of the 628 spatial organization of a bacterial chromosome. Science 342, 731–734 (2013). 629 7. Böhm, K. et al. Chromosome organization by a conserved condensin-ParB system in the 630 actinobacterium Corynebacterium glutamicum. Nat. Commun. 11, 1485 (2020). 631 8. Trussart, M. et al. Defined chromosome structure in the genome-reduced bacterium 632 Mycoplasma pneumoniae. Nat. Commun. 8, 14665 (2017).

633 9. Val, M.-E. et al. A checkpoint control orchestrates the replication of the two 634 chromosomes of Vibrio cholerae. Sci. Adv. 2, e1501914 (2016). 635 10. Lioy, V. S. et al. Multiscale Structuring of the E. coli Chromosome by Nucleoid-Associated 636 and Condensin Proteins. Cell 172, 771-783.e18 (2018). 637 11. Remesh, S. G. et al. Nucleoid remodeling during environmental adaptation is regulated 638 by HU-dependent DNA bundling. Nat. Commun. 11, 2905 (2020). 639 12. Wang, X., Montero Llopis, P. & Rudner, D. Z. Bacillus subtilis chromosome organization 640 oscillates between two distinct patterns. Proc. Natl. Acad. Sci. U. S. A. 111, 12877–12882 641 (2014). 642 13. Errington, J. Septation and chromosome segregation during sporulation in Bacillus 643 subtilis. Curr. Opin. Microbiol. 4, 660–666 (2001). 644 14. Szafran, M. J., Jakimowicz, D. & Elliot, M. A. Compaction and control - the role of 645 chromosome organizing proteins in Streptomyces. FEMS Microbiol. Rev. (2020) 646 doi:10.1093/femsre/fuaa028. 647 15. Forterre, P. & Gadelle, D. Phylogenomics of DNA topoisomerases: their origin and 648 putative roles in the emergence of modern organisms. Nucleic Acids Res. 37, 679–692 649 (2009). 650 16. Dame, R. T., Rashid, F.-Z. M. & Grainger, D. C. Chromosome organization in bacteria: 651 mechanistic insights into genome structure and function. Nat. Rev. Genet. (2019) 652 doi:10.1038/s41576-019-0185-4. 653 17. Dillon, S. C. & Dorman, C. J. Bacterial nucleoid-associated proteins, nucleoid structure 654 and gene expression. Nat. Rev. Microbiol. 8, 185–195 (2010). 655 18. Wang, W., Li, G.-W., Chen, C., Xie, X. S. & Zhuang, X. Chromosome organization by a 656 nucleoid-associated protein in live bacteria. Science 333, 1445–1449 (2011). 657 19. Cobbe, N. & Heck, M. M. S. The evolution of SMC proteins: phylogenetic analysis and 658 structural implications. Mol. Biol. Evol. 21, 332–347 (2004). 659 20. Gligoris, T. & Löwe, J. Structural Insights into Ring Formation of Cohesin and Related Smc 660 Complexes. Trends Cell Biol. 26, 680–693 (2016). 661 21. Nolivos, S. & Sherratt, D. The bacterial chromosome: architecture and action of bacterial 662 SMC and SMC-like complexes. FEMS Microbiol. Rev. 38, 380–392 (2014). 663 22. Ganji, M. et al. Real-time imaging of DNA loop extrusion by condensin. Science 360, 102– 664 105 (2018). 665 23. Banigan, E. J. & Mirny, L. A. Loop extrusion: theory meets single-molecule experiments. 666 Curr. Opin. Cell Biol. 64, 124–138 (2020). 667 24. Marko, J. F., De Los Rios, P., Barducci, A. & Gruber, S. DNA-segment-capture model for 668 loop extrusion by structural maintenance of chromosome (SMC) protein complexes. 669 Nucleic Acids Res. 47, 6956–6972 (2019). 670 25. Terakawa, T. et al. The condensin complex is a mechanochemical motor that 671 translocates along DNA. Science 358, 672–676 (2017). 672 26. Mäkelä, J. & Sherratt, D. SMC complexes organize the bacterial chromosome by 673 lengthwise compaction. Curr. Genet. (2020) doi:10.1007/s00294-020-01076-w. 674 27. Gruber, S. & Errington, J. Recruitment of condensin to replication origin regions by 675 ParB/SpoOJ promotes chromosome segregation in B. subtilis. Cell 137, 685–696 (2009). 676 28. Sullivan, N. L., Marquis, K. A. & Rudner, D. Z. Recruitment of SMC by ParB-parS organizes 677 the origin region and promotes efficient chromosome segregation. Cell 137, 697–707 678 (2009).

679 29. Kawalek, A., Wawrzyniak, P., Bartosik, A. A. & Jagura-Burdzy, G. Rules and Exceptions: 680 The Role of Chromosomal ParB in DNA Segregation and Other Cellular Processes. 681 Microorganisms 8, (2020). 682 30. Leonard, T. A., Butler, P. J. & Löwe, J. Bacterial chromosome segregation: structure and 683 DNA binding of the Soj dimer--a conserved biological switch. EMBO J. 24, 270–282 684 (2005). 685 31. Surovtsev, I. V. & Jacobs-Wagner, C. Subcellular Organization: A Critical Feature of 686 Bacterial Cell Replication. Cell 172, 1271–1293 (2018). 687 32. Lim, H. C. et al. Evidence for a DNA-relay mechanism in ParABS-mediated chromosome 688 segregation. eLife 3, e02758 (2014). 689 33. Surovtsev, I. V., Campos, M. & Jacobs-Wagner, C. DNA-relay mechanism is sufficient to 690 explain ParA-dependent intracellular transport and patterning of single and multiple 691 cargos. Proc. Natl. Acad. Sci. U. S. A. 113, E7268–E7276 (2016). 692 34. Flardh, K. & Buttner, M. J. Streptomyces morphogenetics: dissecting differentiation in a 693 filamentous bacterium. Nat. Rev. Microbiol. 7, 36–49 (2009). 694 35. Jones, S. E. & Elliot, M. A. ‘Exploring’ the regulation of Streptomyces growth and 695 development. Curr. Opin. Microbiol. 42, 25–30 (2018). 696 36. HOPWOOD, D. A. & GLAUERT, A. M. Observations on the chromatinic bodies of 697 Streptomyces coelicolor. J. Biophys. Biochem. Cytol. 8, 257–265 (1960). 698 37. Ruban-Ośmiałowska, B., Jakimowicz, D., Smulczyk-Krawczyszyn, A., Chater, K. F. & 699 Zakrzewska-Czerwińska, J. Replisome localization in vegetative and aerial hyphae of 700 Streptomyces coelicolor. J. Bacteriol. 188, 7311–7316 (2006). 701 38. Schlimpert, S., Flärdh, K. & Buttner, M. Fluorescence Time-lapse Imaging of the Complete 702 S. venezuelae Life Cycle Using a Microfluidic Device. J. Vis. Exp. JoVE 53863 (2016) 703 doi:10.3791/53863. 704 39. Donczew, M. et al. ParA and ParB coordinate chromosome segregation with cell 705 elongation and division during Streptomyces sporulation. Open Biol. 6, 150263 (2016). 706 40. Kois, A., Swiatek, M., Jakimowicz, D. & Zakrzewska-Czerwińska, J. SMC protein- 707 dependent chromosome condensation during aerial hyphal development in 708 Streptomyces. J. Bacteriol. 191, 310–319 (2009). 709 41. Swiercz, J. P., Nanji, T., Gloyd, M., Guarné, A. & Elliot, M. A. A novel nucleoid-associated 710 protein specific to the actinobacteria. Nucleic Acids Res. 41, 4171–4184 (2013). 711 42. Salerno, P. et al. One of the two genes encoding nucleoid-associated HU proteins in 712 Streptomyces coelicolor is developmentally regulated and specifically involved in spore 713 maturation. J. Bacteriol. 191, 6489–6500 (2009). 714 43. Facey, P. D. et al. Streptomyces coelicolor Dps-like proteins: differential dual roles in 715 response to stress during vegetative growth and in nucleoid condensation during 716 reproductive cell division. Mol. Microbiol. 73, 1186–1202 (2009). 717 44. Kim, H. J., Calcutt, M. J., Schmidt, F. J. & Chater, K. F. Partitioning of the linear 718 chromosome during sporulation of Streptomyces coelicolor A3(2) involves an oriC-linked 719 parAB locus. J. Bacteriol. 182, 1313–1320 (2000). 720 45. Yang, M. C. & Losick, R. Cytological evidence for association of the ends of the linear 721 chromosome in Streptomyces coelicolor. J. Bacteriol. 183, 5180–5186 (2001). 722 46. Kois-Ostrowska, A. et al. Unique Function of the Bacterial Chromosome Segregation 723 Machinery in Apically Growing Streptomyces - Targeting the Chromosome to New 724 Hyphal Tubes and its Anchorage at the Tips. PLoS Genet. 12, e1006488 (2016).

725 47. Sambrook, J. & Russell, D. Molecular cloning: a laboratory manual, 3rd edition. (Cold 726 Spring Harbor Laboratory Press). 727 48. Kieser, T., Bibb, M., Buttner, M., Chater, K. & Hopwood, D. Practica Streptomyces 728 genetics. (John Innes Foundation). 729 49. Le, T. B. K. Chromosome Conformation Capture with Deep Sequencing to Study the Roles 730 of the Structural Maintenance of Chromosomes Complex In Vivo. Methods Mol. Biol. 731 Clifton NJ 2004, 105–118 (2019). 732 50. Wolff, J. et al. Galaxy HiCExplorer: a web server for reproducible Hi-C data analysis, 733 quality control and visualization. Nucleic Acids Res. 46, W11–W16 (2018). 734 51. Imakaev, M. et al. Iterative correction of Hi-C data reveals hallmarks of chromosome 735 organization. Nat. Methods 9, 999–1003 (2012). 736 52. Lun, A. T. L. & Smyth, G. K. From reads to regions: a Bioconductor workflow to detect 737 differential binding in ChIP-seq data. F1000Research 4, 1080 (2015). 738 53. Langmead, B., Wilks, C., Antonescu, V. & Charles, R. Scaling read aligners to hundreds of 739 threads on general-purpose processors. Bioinforma. Oxf. Engl. 35, 421–432 (2019). 740 54. Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat. Methods 741 9, 357–359 (2012). 742 55. Li, H. et al. The Sequence Alignment/Map format and SAMtools. Bioinforma. Oxf. Engl. 743 25, 2078–2079 (2009). 744 56. Lun, A. T. L. & Smyth, G. K. csaw: a Bioconductor package for differential binding analysis 745 of ChIP-seq data using sliding windows. Nucleic Acids Res. 44, e45 (2016). 746 57. Lun, A. T. L. & Smyth, G. K. De novo detection of differentially bound regions for ChIP-seq 747 data using peaks and windows: controlling error rates correctly. Nucleic Acids Res. 42, 748 e95 (2014). 749 58. Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for 750 differential expression analysis of digital gene expression data. Bioinforma. Oxf. Engl. 26, 751 139–140 (2010). 752 59. McCarthy, D. J., Chen, Y. & Smyth, G. K. Differential expression analysis of multifactor 753 RNA-Seq experiments with respect to biological variation. Nucleic Acids Res. 40, 4288– 754 4297 (2012). 755 60. Zhang, Y. et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol. 9, R137 (2008). 756 61. Bailey, T. L. et al. MEME SUITE: tools for motif discovery and searching. Nucleic Acids Res. 757 37, W202-208 (2009). 758 62. Helmuth, J. et al. normR: Regime enrichment calling for ChIP-seq data. bioRxiv 082263 759 (2016) doi:10.1101/082263. 760 63. Bush, M. J., Chandra, G., Al-Bassam, M. M., Findlay, K. C. & Buttner, M. J. BldC Delays 761 Entry into Development To Produce a Sustained Period of Vegetative Growth in 762 Streptomyces venezuelae. mBio 10, (2019). 763 64. Wolanski, M. et al. The level of AdpA directly affects expression of developmental genes 764 in Streptomyces coelicolor. J. Bacteriol. 193, 6358–6365 (2011). 765 65. Bush, M. J., Bibb, M. J., Chandra, G., Findlay, K. C. & Buttner, M. J. Genes required for 766 aerial growth, cell division, and chromosome segregation are targets of WhiA before 767 sporulation in Streptomyces venezuelae. mBio 4, e00684-00613 (2013). 768 66. Willemse, J., Borst, J. W., de Waal, E., Bisseling, T. & van Wezel, G. P. Positive control of 769 cell division: FtsZ is recruited by SsgB during sporulation of Streptomyces. Genes Dev. 25, 770 89–99 (2011).

771 67. van Berkum, N. L. et al. Hi-C: a method to study the three-dimensional architecture of 772 genomes. J. Vis. Exp. JoVE (2010) doi:10.3791/1869. 773 68. Lieberman-Aiden, E. et al. Comprehensive mapping of long-range interactions reveals 774 folding principles of the human genome. Science 326, 289–293 (2009). 775 69. Jakimowicz, D., Gust, B., Zakrzewska-Czerwinska, J. & Chater, K. F. Developmental-stage- 776 specific assembly of ParB complexes in Streptomyces coelicolor hyphae. J. Bacteriol. 187, 777 3572–3580 (2005). 778 70. Jakimowicz, D., Zydek, P., Kois, A., Zakrzewska-Czerwińska, J. & Chater, K. F. Alignment of 779 multiple chromosomes along helical ParA scaffolding in sporulating Streptomyces 780 hyphae. Mol. Microbiol. 65, 625–641 (2007). 781 71. Dedrick, R. M., Wildschutte, H. & McCormick, J. R. Genetic interactions of smc, ftsK, and 782 parB genes in Streptomyces coelicolor and their developmental genome segregation 783 phenotypes. J. Bacteriol. 191, 320–332 (2009). 784 72. Gehrke, E. J. et al. Silencing cryptic specialized metabolism in Streptomyces by the 785 nucleoid-associated protein Lsr2. eLife 8, (2019). 786 73. Jeong, Y. et al. The dynamic transcriptional and translational landscape of the model 787 antibiotic producer Streptomyces coelicolor A3(2). Nat. Commun. 7, 11605 (2016). 788 74. Huang, J., Lih, C. J., Pan, K. H. & Cohen, S. N. Global analysis of growth phase responsive 789 gene expression and regulation of antibiotic biosynthetic pathways in Streptomyces 790 coelicolor using DNA microarrays. Genes Dev. 15, 3183–3192 (2001). 791 75. Jakimowicz, D., Mouz, S., Zakrzewska-Czerwinska, J. & Chater, K. F. Developmental 792 control of a parAB promoter leads to formation of sporulation-associated ParB 793 complexes in Streptomyces coelicolor. J. Bacteriol. 188, 1710–1720 (2006). 794 76. Soh, Y.-M. et al. Self-organization of parS centromeres by the ParB CTP hydrolase. 795 Science 366, 1129–1133 (2019). 796 77. Jalal, A. S., Tran, N. T. & Le, T. B. ParB spreading on DNA requires cytidine triphosphate in 797 vitro. eLife 9, (2020). 798 78. Osorio-Valeriano, M. et al. ParB-type DNA Segregation Proteins Are CTP-Dependent 799 Molecular Switches. Cell 179, 1512-1524.e15 (2019). 800 79. Ginda, K. et al. ParA of Mycobacterium smegmatis co-ordinates chromosome 801 segregation with the cell cycle and interacts with the polar growth determinant DivIVA. 802 Mol. Microbiol. 87, 998–1012 (2013). 803 80. Baek, J. H., Rajagopala, S. V. & Chattoraj, D. K. Chromosome segregation proteins of 804 Vibrio cholerae as transcription regulators. mBio 5, e01061-01014 (2014). 805 81. Breier, A. M. & Grossman, A. D. Whole-genome analysis of the chromosome partitioning 806 and sporulation protein Spo0J (ParB) reveals spreading and origin-distal sites on the 807 Bacillus subtilis chromosome. Mol. Microbiol. 64, 703–718 (2007). 808 82. Li, P., Zhang, H., Zhao, G.-P. & Zhao, W. Deacetylation enhances ParB-DNA interactions 809 affecting chromosome segregation in Streptomyces coelicolor. Nucleic Acids Res. 48, 810 4902–4914 (2020). 811 83. Tran, N. T., Laub, M. T. & Le, T. B. K. SMC Progressively Aligns Chromosomal Arms in 812 Caulobacter crescentus but Is Antagonized by Convergent Transcription. Cell Rep. 20, 813 2057–2071 (2017). 814 84. Yu, W., Herbert, S., Graumann, P. L. & Götz, F. Contribution of SMC (Structural 815 Maintenance of Chromosomes) and SpoIIIE to Chromosome Segregation in 816 Staphylococci. J. Bacteriol. 192, 4067 (2010).

817 85. Nolivos, S. et al. MatP regulates the coordinated action of topoisomerase IV and MukBEF 818 in chromosome segregation. Nat. Commun. 7, 10466 (2016). 819 86. Rybenkov, V. V., Herrera, V., Petrushenko, Z. M. & Zhao, H. MukBEF, a chromosomal 820 organizer. J. Mol. Microbiol. Biotechnol. 24, 371–383 (2014). 821 87. Danilova, O., Reyes-Lamothe, R., Pinskaya, M., Sherratt, D. & Possoz, C. MukB colocalizes 822 with the oriC region and is required for organization of the two Escherichia coli 823 chromosome arms into separate cell halves. Mol. Microbiol. 65, 1485–1492 (2007). 824 88. Niki, H., Jaffé, A., Imamura, R., Ogura, T. & Hiraga, S. The new gene mukB codes for a 177 825 kd protein with coiled-coil domains involved in chromosome partitioning of E. coli. 826 EMBO J. 10, 183–193 (1991). 827 89. Szafran, M. et al. Topoisomerase I (TopA) is recruited to ParB complexes and is required 828 for proper chromosome organization during Streptomyces coelicolor sporulation. J. 829 Bacteriol. 195, 4445–4455 (2013). 830 831 832 FIGURE LEGENDS

833 Figure 1. Spatial organization of the S. venezuelae linear chromosome during sporulation.

834 (A) The S. venezuelae (wild type ftsZ-ypet derivative, MD100) growth curve (in 5 ml culture).

835 The critical time points of sporogenic development are marked with arrows: growth arrest

836 (red), the appearance of FtsZ ladders (green) and the formation of spore chains (blue). The

837 normalized FtsZ-YPet fluorescence [U/µg] and the relative oriC/ter ratio are shown as insets

838 in the plot, with X axes corresponding to the main plot X axis. The oriC/ter ratio at the 25th

839 hour of growth was set as 1.0. The right panel shows the visualization of condensing

840 nucleoids (DNA with staining 7-AAD) and FtsZ-YPet at 22nd hour of growth; scale bar - 2 µm.

841 (B) The normalized Hi-C contact map obtained for the wild type (ftsZ-ypet derivative,

842 MD100) after 22 h of growth (5 ml culture). The dotted pink lines mark the position of the

843 oriC region. The dotted black lines mark the positions of the left (LTD) and right (RTD)

844 terminal domains. (C) Top panel: the normalized Hi-C contact maps obtained for the wild

845 type (ftsZ-ypet derivative, MD100) strain growing for 13, 15, 17, 22 and 25 h (in 5 ml

846 culture). Bottom panel: the differential Hi-C maps in the logarithmic scale (log2) comparing

847 the contact enrichment at 22 h of growth (red) versus other time points (blue). The X- and Y-

848 axes indicate chromosomal coordinates binned in 30 kb. The inset shows the model of

849 chromosome rearrangement in the course of sporulation. (D) Top panel: the normalized Hi-C

850 contact maps generated for parA and parB mutants (ftsZ-ypet derivatives MD011 and

851 MD021, respectively) growing for 22 h (5 ml culture). Bottom panel: the differential Hi-C

852 maps in the logarithmic scale (log2) comparing the contact enrichment in the wild type strain

853 (red) versus the mutant strain (blue). The inset shows the model of chromosome

854 organization in the parA and parB mutants.

855

856 Figure 2. Phenotypic effects of smc, hupS and double smc/hupS deletions. (A) Scanning

857 electron micrographs showing sporulating hyphae and spores of the wild type and hupS, smc

858 and hupS smc double mutant (ftsZ-ypet derivatives, MD100, TM005, TM004 and TM006,

859 respectively). Elongated spores are marked with a yellow arrowhead. Scale bar: 10 µm. (B)

860 Box plot analysis of the spore length distribution in the S. venezuelae wild type (light green),

861 hupS (yellow), smc (blue) and hupS smc (pink) (AK200, TM010, TM003, respectively), as well

862 as in hupS and smc mutants complemented with in trans delivered hupS-flag (orange) and

863 smc-flag genes (dark blue) (TM015 and TM019, respectively) (approximately 300 spores

864 were analysed for each strain). The statistical significance between strains is marked with

865 asterisks. (C) Representative examples of time-lapse images showing visualization of

866 nucleoids (marked with mCherry-HupA fusion) and Z-rings (visualized by FtsZ-YPet fusion) in

867 the wild type background (ftsZ-ypet, hupA-mCherry derivative, TM011) hupS, smc and hupS

868 smc double mutant background (ftsZ-ypet, hupA-mCherry derivative, TM013, TM012, and

869 TM014, respectively). An abnormal hyphal fragment is marked with a yellow arrowhead.

870 Scale bar: 5 µm. (D) Box plot analysis of the nucleoid area (visualized by mCherry-HupA

871 fusion) in the wild type (light green), hupS (yellow), smc (blue) and hupS smc (pink) double

872 mutant (ftsZ-ypet, hupA-mcherry derivative, MD100, TM011, TM013, TM012, TM014,

873 respectively) (E) Kymographs showing the intensity of Z-ring fluorescence along the

874 representative hyphae in the wild type and hupS, smc and hupS smc double mutant (ftsZ-

875 ypet derivatives, MD100, TM005, TM004, TM006, respectively) during maturation (F) The

876 variation (standard deviation) of FtsZ-Ypet (Z-rings) fluorescence during spore maturation of

877 strains calculated for 25-30 hyphae of the wild type (light green), hupS (yellow), smc (blue)

878 and hupS smc (pink) double mutant (ftsZ-ypet derivatives, TM005, TM004, TM006,

879 respectively). Fluorescence was measured starting from 30 min before growth arrest until

880 spore formation. Points represent values for each hyphal, while lines show loess model fit

881 with 95% confidence intervals.

882

883 Figure 3. Influence of SMC and HupS on chromosome organization. (A) Top panels: the

884 normalized Hi-C contact maps obtained for the wild type and smc, hupS and smc hupS

885 double mutants (ftsZ-ypet derivatives, MD100, TM005, TM004, TM006, respectively)

886 growing for 22 h or for 25 h (only in the case of the double hupS smc mutant) (in 5 ml

887 cultures). Bottom panels: the corresponding differential contact maps in the logarithmic

888 scale (log2) comparing the contact enrichment in the wild type strain (red) versus the

889 mutant strains (blue) are shown below each Hi-C contact map. (B) The number of SMC-FLAG

890 binding sites along the chromosome of the wild type (blue) (TM017 strain) or parB mutant

891 background (red) (KP4F4) determined by ChIP-Seq analysis at the time of sporogenic cell

892 division (14th hour of 50 ml culture growth). The number of SMC-FLAG binding sites (250 bp)

893 was determined by normr analysis averaged over 0.5 Mbp sliding window every 1000 bp

894 with a loess model fit. Points below the plot show positions of individual sites coloured

895 according to their FDR (false discovery rate) values (yellow less significant, dark purple more

896 significant). (C) The number of HupS-FLAG binding sites along the chromosome in the wild

897 type background (hupS deletion complemented with hupS-FLAG delivered in trans, TM015)

898 determined by ChIP-Seq analysis at the time of sporogenic cell division (14th hour of 50 ml

899 culture growth). The number of HupS-FLAG binding sites was determined by differential

900 analysis of the HupS-FLAG and wild type strains using edgeR and averaged over 0.5 Mbp

901 sliding window every 1000 bp. Smooth line shows loess model fit. The points below the plot

902 show positions of individual sites coloured according to their FDR values (yellow less

903 significant, dark purple more significant). (D) Heatmaps showing the number of reads for all

904 307 regions bound by HupS-FLAG in the wild type background (TM015), smc mutant

905 background (TM016) and negative control (the wild type strain). The reads were normalized

906 by the glmQL model from the edgeR package. For each region, position 0 is the position with

907 the maximum number of reads. Regions are sorted according to their logFC values. (E)

908 Comparison of HupS-FLAG binding in the wild type (blue) and smc deletion background (red).

909 The lines show mean values of reads (for 69 bp long regions) with 95% confidence intervals

910 for all 307 sites differentially bound by HupS-FLAG protein. For all sites, 1000 bp long

911 fragments were extracted from the chromosome centred around the best position as

912 determined by edgeR analysis.

913

914 Figure 4. Model of spatial rearrangement of the S. venezuelae chromosome during

915 sporogenic development showing the contribution of ParB (green), SMC (orange) and

916 HupS (blue) to chromosomal arm alignment and organization of terminal domains (TLD

917 and RTD). Blue lines indicated intrachromosomal contacts, oriC region is marked with yellow

918 circle.

919

920

35 bioRxiv preprint doi: https://doi.org/10.1101/2020.12.09.403915; this version posted December 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2020.12.09.403915; this version posted December 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2020.12.09.403915; this version posted December 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2020.12.09.403915; this version posted December 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.