Strains and growth conditions
Strains used in this study are listed in Supplementary Table 1. Solid MS (Mannitol Soy flour) medium46 was used for collection of Streptomyces spores and for conjugation experiments, while MYM medium47 was used for obtaining Kitasatospora spores. To compare colony sizes and observe the release of cell wall-deficient (CWD) cells under hyperosmotic stress, solid LPMA medium8 was used. TSBS46 medium was used to grow Streptomyces in liquid medium without hyperosmotic stress. For quantification of the number of CWD cells, as well as measurement of hyphal diameters and membrane fluidity, liquid L-phase broth (LPB) was used8. Briefly, 106 CFU ml− 1 spores were inoculated in 20 ml LPB in 50 ml flasks without coil while shaking at 100 rpm ml− 1. All Streptomyces and Kitasatospora strains were grown at 30°C. For hyphal branching detection, spores were inoculated onto cellophane membranes overlaying LPMA plates, which were then incubated at 30°C for 16 h prior to analysis. Lysozyme sensitive assay were performed essentially as described19.
E. coli DH5α48 was used for cloning and β-lactamase experiments. E. coli BL21 (DE3)49 and BTH10150 were used in protein expression and bacterial two-hybrid assays, respectively. For conjugation, E. coli ET12567 harboring pUZ8002 was used. All E. coli strains were grown in LB medium at 37°C with appropriate antibiotics, if needed.
Plasmid construction and transformation
Plasmids and primers used in this study are listed in Supplementary Table 2 and Supplementary Table 3, respectively. For complementation of the ∆stlP mutant, the coding sequence of stlP (SCO2834) was amplified from genomic DNA of S. coelicolor with primers stlP-F/stlP-R1. The PCR product was ligated as an NdeI/BamHI fragment into plasmid hpXZ219 harboring the constitutive gapAp prompter of SCO1947, yielding plasmid pXZ15.
For localization of StlP and DivlVA, the mCherry reporter was fused to the C-terminus of these two proteins. Coding sequences for stlP and divlVA were amplified from genomic DNA of S. coelicolor with primers stlP-F/stlP-R2 and 2077-F/2077-R, respectively. The gene encoding for mCherry was amplified from plasmid pGWS79127 using primers mCh-F/mCh-R. After digestions with NdeI/HindIII (for stlP and divlVA) and HindIII/XbaI (for mCherry), combinations of stlP/mCherry and divlVA/mCherry were ligated with plasmid hpXZ2 that was cut with NdeI/XbaI, generating plasmid pXZ16 and pXZ17, respectively.
To constitutively express the HflK/C-like protein BOQ63_030050 in Kitasatospora viridifaciens and the S. coelicolor stlP mutant, the coding sequence of BOQ63_030050 was amplified from genomic DNA of K. viridifaciens DSM40239 with primers 0050-F/0050-R. The PCR product was ligated as an NdeI/XbaI fragment into plasmid hpXZ2 harboring the constitutive gapAp prompter, yielding plasmid pXZ41.
To express the stomatin-encoding domain of StlP (StlPSD) in E. coli, nucleotides 610–1107 of stlP were amplified from S. coelicolor genomic DNA with primers 2834sto-F/2834sto-R and ligated into pET28a plasmid as a NdeI and HindIII fragment, yielding plasmid pXZ18. Consequently, StlPSD was expressed carrying a N-terminal Histidine-tag (6xHis).
All plasmids were introduced into E. coli and Streptomyces via heat-shock transformation51 and conjugation46, respectively.
Inactivation of stlP in Streptomyces
For inactivation of stlP in S. coelicolor and S. lividans, we used cosmid StE20 carrying the Tn5062 transposon inserted after nucleotide position 892 relative to the start site of stlP (kindly provided by Prof. Paul Dyson, see Supplementary Table 2). This cosmid was introduced into S. coelicolor via conjugation using ET12567/pUZ800246, after which exconjugants were screened as described52. Colonies that were kanamycin-sensitive and apramycin-resistant were selected and used for further analysis. Mutants carrying the expected phenotype were verified by sequencing.
Growth curve generation
For preparing germinated spores, spores of different strains were resuspended in double-strength germination medium46 at a final concentration of 106 CFU ml− 1 and incubated at 30 ⁰C for 6 ~ 8 h while shaking at 200 rpm min− 1. For generating the biomass growth curve, the RoboLector L-4-BL-II equipped with a parallelized shaken cultivation device was used53. Briefly, the germinated spores were centrifuged and resuspended in either TSBS or LPB media before being distributed (1 ml per well) into a 48-well FlowerPlates (Basesweiler Germany). The temperature and humidity were set at 30 ⁰C and 85%, respectively. The biomass was collected automatically and measured in arbitrary units. All measurements were performed in triplicate.
Bacterial 2-hybrid analysis
To assess interactions of StlP with other tip-localizing proteins, the bacterial hybrid assay was used50. Therefore, lpmP (SCO2833), stlP (SCO2834), sco2835, cslA (SCO2836), glxA (SCO2837), cslZ (SCO2838), divlVA (SCO2077), filP (SCO5396) and scy (SCO5397) were amplified using primers Th2833-F/Th2833-R, Th2834-F/Th2834-R, Th2835-F/Th2835-R, Th2836-F/Th2836-R, Th2837-F/Th2837-R, Th2838-F/Th2838-R, divlVA-F/divlVA-R, filP-F/filP-R and scy-F/scy-R, respectively. All amplified DNA fragments were cloned into the pKT25 and pUT18C plasmids using EcoRI and XbaI, yielding plasmids pXZ19 (pUT18C + lpmP), pXZ20 (pKT25 + stlP), pXZ21 (pUT18C + stlP), pXZ22 (pKT25 + SCO2835), pXZ23 (pUT18C + SCO2835), pXZ24 (pKT25 + cslA), pXZ25 (pUT18C + cslA), pXZ26 (pKT25 + glxA), pXZ27 (pUT18C + glxA), pXZ28 (pKT25 + cslZ), pXZ29 (pKT25 + divlVA), pXZ30 (pUT18C + divlVA), pXZ31 (pKT25 + filP), pXZ32 (pUT18C + filP), pXZ33 (pKT25 + scy) and pXZ34 (pUT18C + scy) (see Supplementary Table 2).
E. coli BT101 carrying combinations of these constructs were used in bacterial 2-hybrid experiments to evaluate protein interactions as described54.
Topology determination of StlP
To study the transmembrane topology of StlP, the β-lactamase-encoding gene blaM without its signal sequence was fused to the 3’ end of stlP, as described previously27. Only if BlaM is secreted, cells will be resistant to ampicillin. To this end, a blaM variant lacking the region encoding the signal sequence for secretion (blaMNS) was amplified from plasmid pHJL40155 with primers blaM-F/blaM-R. In parallel, full length blaM (including the region for the signal sequence, hereinafter referred to as blaMFL) was amplified from the same plasmid using the blaMFL-F/blaM-R primers. The stlP gene was amplified from S. coelicolor genomic DNA using primers stlP-F/stlP-R3. Subsequently, the amplified products were cut using restriction enzymes NdeI-HindIII (stlP), HindIII-EcoRI (blaMNS) and NdeI-EcoRI (blaMFL), and digested combinations of stlP/blaMNS, and blaMFL were separately ligated into pXZ219 that was cut with NdeI-EcoRI, yielding pXZ35 and pXZ36 (see Supplementary Table 2).
E. coli DH5α harboring plasmid pSET152, pXZ35 and pXZ36 were used to assess the membrane topology of StlP, which were performed as described previously27. Full-length BlaM (BlaMFL) expressed from pXZ36 served as a control for β-lactamase activity in the medium.
Quantification of CWD cells
Culturing and filtration of CWD cells of Streptomyces and Kitasatospora strains was essentially performed as described8, with the exception that S. coelicolor strains were grown for 16 h, while K. viridifaciens strains were grown for 40 h. Quantification of CWD cells was performed with a Bright-Line™ Hemocytometer (Merck), as described56. Briefly, 10 µl of filtered culture supernatant were loaded into the counting chamber, after which cells were quantified under a Zeiss Axio microscope equipped with an Axiocam 105 camera. CWD cell numbers were counted manually, and the density was calculated. For each strain, the measurements were performed in triplicate.
Protein expression and purification
To purify the stomatin-encoding domain of StlP (StlPSD), E. coli BL21 (DE3) cells harboring plasmid pXZ17 were cultured at 37°C to an OD600 of 0.8 in LB medium containing 50 mg ml− 1 kanamycin. Then, 0.5 mM isopropyl β-D-thiogalactopyranoside was added to induce protein expression, after which cells were grown at 30°C for 18 h. The induced cells were subsequently lysed by sonication in binding buffer (50 mM Tris–HCl, 200 mM NaCl, pH 8.0), and after centrifugation, the lysate was loaded on a Ni2+-chelating column equilibrated with binding buffer. Ten column volumes of washing buffer (50 mM Tris–HCl, 200 mM NaCl, 0.1 mM imidazole, pH 8.0) and 5 mL of elution buffer (50 mM Tris–HCl, 200 mM NaCl, 10 mM imidazole, pH 8.0) were used to wash and elute StlPSD, respectively. Finally, the protein was purified by gel filtration using a Hiload 16/600 Superdex 200 pg column (GE Healthcare) equilibrated with buffer (50 mM Tris–HCl, 100 mM NaCl, pH 8.0). Sample fractions were analyzed on a 12.5% SDS-PAGE gel. Fractions were stored directly at -80°C or first concentrated to 1 mg ml− 1 with the 3 kDa molecular weight cutoff concentrator (Millipore) and then stored.
Chemical cross-linking
The cross-linking of StlPSD with glutaraldehyde (GA, 50% in H2O, Sigma) was performed as described44 with the following modifications. Briefly, 1 mg ml− 1 of StlPSD was treated with 0.02, 0.05 or 0.1% GA for 10 or 60 min at room temperature in 50 µl buffer (50 mM Tris–HCl, 100 mM NaCl, pH 8.0). Reactions were quenched with 0.2 M glycine-NaOH (pH 9.5) for 5 min at room temperature. Detection of cross-linked protein was performed by loading the samples on a 12.5% SDS-PAGE gel.
Microscopy
To visualize the emergence of CWD cells, spores of the wild-type strain and the ∆stlP mutant were pre-germinated in double strength germination medium46. Then 10 µl of germlings were used for live-imaging, for which an ibiTreat 35 mm low imaging dish (ibidi) and a LPMA-pad was used as before8. Live imaging was carried out using a Zeiss LSM900 Airyscan 2 microscope. If necessary, Z-Stack acquisitions were used. For visualization of membrane and nucleic acids, 0.05 mg ml− 1 FM5-95 and 0.5 µM SYTO-9 (Sigma) were used, respectively.
The detection of nascent peptidoglycan was done by using BODIPY-FL vancomycin (Sigma), essentially as described57. Briefly, after growing Streptomyces strains in LPB medium for 16 h, mycelia were collected (3300 rpm min− 1, 10 min) and resuspended in 200 µl of fresh LPB medium containing 1 µg ml− 1 BODIPY FL vancomycin and 1 µg ml− 1 vancomycin. After 10 min incubation at 30°C, the mycelia were washed 3 times with PBS. 5 µl of the washed mycelia was used for microscopy analysis using a Zeiss LSM900 Airyscan 2 microscope.
For visualizing fluid membrane microdomains, mycelia were stained with DilC1233,34. Briefly, DilC12 was dissolved in DMSO and 10 mg ml− 1 stock was prepared. For sample preparation, spores of each strain were inoculated in LPB medium at a final concentration of 106 CFU ml− 1. After 16h of growth, mycelia were collected by centrifuge (3300 rpm min− 1, 10 min) and resuspended in prewarmed fresh LPB medium supplemented with 100 µg ml− 1 DilC12, followed by growth for 3 h at 30°C while shaking at 100 rpm min− 1. Then, mycelia were collected, washed 3 times with prewarmed LPB supplemented with 1% DMSO, and resuspended in the same wash buffer. DilC12 signal was detected via a Cy3 filter (535 nm excitation and 590 nm emission) using a Zeiss LSM900 Airyscan 2 microscope, wherein the cultivation chamber had been set 30°C to avoid temperature changes during imaging.
For measurement of membrane fluidity, samples were prepared essentially as described32. Briefly, Laurdan (6-Dodecanoyl-N, N-dymethyl2-naphthylamine, Sigma) was dissolved in dimethylformamide (DMF) and a 10 mM stock was prepared. For preparation of mycelia for Laurdan staining, 20 ml 16 h-old mycelia were collected (3300 rpm min− 1, 10 min) and resuspended in 1 ml 30°C pre-warmed LPB medium containing 1 mM Laurdan. For sample preparation of the stlP mutant, the culture was filtrated through a 100 µm cut-off filter (Falcon Cell Strainer 100 µm Nylon) to remove the majority of the CWD cells. The filtered mycelium was resuspended in 1 ml 30°C pre-warmed LPB medium and used for Laurdan staining. After 10 min incubation in the dark at 30°C, stained mycelia were collected again and washed twice in 30°C pre-warmed PBS buffer supplemented with 20% sucrose and 1% DMF, and finally resuspended in 200 µl pre-warmed PBS buffer supplemented with 20% sucrose. Fluorescent intensities were measured at 435 and 490 nm, following excitation at 350 nm using a Zeiss LSM900 Airyscan 2 microscope, wherein the cultivation chamber had been set 30°C to avoid temperature changes during imaging. To calculate the membrane fluidity at hyphal tips, the tip region was cropped as a square (1 µm x 1 µm) from the image and the corresponding generalized polarization (GP) value was determined as described58.
For measurement of hyphal diameters, 16 h-old mycelia were collected (3300 rpm min− 1, 10 min) and resuspended in fresh LPB medium containing 0.05 mg ml− 1 FM5-95. The distance between the stained membranes was used to measure the hyphal diameters by averaging the diameter at 3 distinct spots in the hypha.
The cellulose-like glycan at hyphal tips was visualized by calcofluor white (Sigma) staining and quantified as previous described19.
For analysis of hyphal branching patterns, mycelium was grown from single spore and imaged using a Zeiss Axio microscope equipped with an Axiocam 105 camera as described previously19. The distance from the tip to the proximal branch point was measured. A proximal branch was defined as having a length of 1–4 µm as previous described59.
For measurement of colony sizes, strain were grown on LPMA medium in petri dishes with a 9 cm diameter aiming for ± 100 colonies per plate. After growing them for 5 ~ 7 days at 30°C, plates were scanned with Epson Perfection V37 scanner and colony size was measured subsequently.
Mycelial Live/dead staining was performed using the LIVE/DEAD BacLight™ Bacterial viability kit (L7012; ThermoFisher) following the manufacturer’s instructions. Briefly, spores of the wild-type strain and the stlP mutant were streaked on LPMA medium. After growth for 16h, the excised ager pieces were inverted and positioned atop 10 µl mixture of SYTO-9 and propidium iodide (PI) nucleic acid stains from the kit with final concentration at 5 µM. Images were taken using a Zeiss LSM900 Airyscan 2 microscope after 10 mins incubation. The viability was calculated by dividing the integrated grey intensity of the fluorescence in the green channel by the integrated intensity of the fluorescence in the red channel. All measurement and images processing were executed with ImageJ software (version 2.0.0/1.53c/Java 1.8.0_172/64-bit).
Sacculus isolation and Cryo-electron tomography
Isolation of sacculi of S. coelicolor M145 and the stlP mutant was essentially performed as described60, except that 16h-old liquid cultures were used and the step of removing teichoic acids was neglected.
Sample preparation for cryo-electron tomography (cryo-ET) was performed as described60. Briefly, after adding the colloidal gold beads, sacculi solutions were vitrificated and applied on the EM grids. Grids were examined using a 120 kV Talos TEM (FEI/ThermoFisher) and cryo-ET data were collected using a Titan Krios instrument (ThermoFisher Scientific). The measurement of cell wall thickness was performed as described60.
Bioinformatic analysis
Protein domains and protein structures were predicted by InterPro (https://www.ebi.ac.uk/interpro/) and AlphaFold 2.061. The prediction of protein membrane topology was performed by TMHMM (Version 2.0) (https://services.healthtech.dtu.dk/service.php?TMHMM-2.0). Alignment of protein structures was done by PyMOL software (Version 2.5). Amino acids sequence alignment was done by ESPript 3.0 (https://espript.ibcp.fr/ESPript/cgi-bin/ESPript.cgi).
To phylogenetically compare StlP with other SPFH proteins, MEGA 7 was used. Amino acid sequences of all SPFH proteins were downloaded from the UniPort database. For phylogenetic analysis of the distribution of StlP, the amino acid sequence of StlP was used to run Position-Specific Iterative (PSI)-BLAST to find homologs in the dataset of 15405 RefSeq representative bacteria and archaea. The homologs with a bitscore > 130 were chosen and subsequently each hit was subjected to membrane topology prediction using TMHMM server (Version 2.0). Only hits with an identical membrane topology were considered valid StlP homologs. The phylogenetic tree was annotated using iTOL (https://itol.embl.de/).
Statistical analysis
For statistical analyses, GraphPad Prism software (version 8.0.2) was used. Significance was determined using student’s t-test.