All plasmids used in this study can be found in Supplementary Table 1 with key sequences provided in Supplementary Tables 2-5. All plasmids were constructed with either Gibson assembly57, Golden Gate assembly58, or inverse PCR (iPCR). DNA manipulations were performed in E. coli strain NEB Turbo. Enzymes were obtained from New England Biolabs.
Strains and growth media
E. coli strains were grown in Luria Bertani (LB) broth, containing 50 μg/ml spectinomycin or 50 μg/mL apramycin as needed. 2,6-diaminopimelic acid was added to LB at a final concentration of 0.1 mM for culturing E. coli strain WM602959. S. venezuelae ATCC 10712 was cultured in complete supplement medium (CSM) unless otherwise indicated. To prepare CSM, 30 g of tryptic soy broth, 1.2 g of Yeast Extract, and 1 g of MgSO4 were added to 1 L of water and autoclaved; filter-sterilized D-(+)-glucose and D-(+)-maltose were then added at a final concentration of 28 mM and 12 mM, respectively. Conjugation involving WM6029 and S. venezuelae was conducted on AS-1 medium. AS-1 was prepared by adding 5 g of soluble starch, 2.5 g of NaCl, 1 g of yeast extract and 18 g of agar to ddH2O to a final volume of 1 L and then autoclaved. A filter-sterilized solution of alanine, arginine and asparagine was added to a final concentration of 0.02 % w/v of each amino acid. Finally, an autoclaved Na2SO4 solution was added to a final concentration of 1 %. To prepare the MSM minimal medium for fermentation experiments, MgSO4 (0.04%, w/v), MOPS (0.377%, w/v), salt solution (0.9%, v/v), trace mineral solution (0.45%, v/v), and 0.2% w/v FeSO4·7H2O solution (0.45%, v/v) were added to ddH2O and the pH was adjusted to 7.5. The salt solution was made by addition of NaCl (1%, w/v) and CaCl2 (1%, w/v) to ddH2O. The trace mineral solution was made by addition of ZnSO4·7H2O (0.088%, w/v), CuSO4·5H2O (0.0039%, w/v), MnSO4·4H2O (0.00061%, w/v), H3BO3 (0.00057%, w/v), and (NH4)Mo7O24·4H2O (0.00037%, w/v) to ddH2O.
Reporter strains construction
Constructs containing different promoter-mCherry combinations were assembled as described above using plasmid JBEI16292, harboring the ΦC31 integration system, as a backbone. These reporter plasmids were subsequently conjugated (see below) and integrated into S. venezuelae ATCC 10712, thus yielding the reporter strains.
E. coli WM6029 cells were transformed with the plasmids to be conjugated. Colonies were picked into liquid LB media containing the appropriate antibiotic and 0.1 mM DAP, and grown overnight at 37 °C. At the same time, S. venezuelae mycelia were grown overnight in CSM. Liquid cultures were then pelleted and the medium was removed. Each WM6029 sample was resuspended in 500 μL of fresh LB with no antibiotics, while S. venezuelae pellets were resuspended in 2 mL of fresh CSM. WM6029 and S. venezuelae were then mixed at a 1:1 ratio, and each co-culture was spotted on AS-1 plates supplemented with 0.1 mM DAP. After incubating for 16-20 hours at 30 °C, the plates were flooded with solutions containing 500 μg of nalidixic acid and 1 mg of the appropriate antibiotic. Plates were then stored at 30 °C until the appearance of exconjugant colonies (3 to 6 days). Exconjugant colonies were streaked on fresh ISP-2 plates supplemented with either apramycin or spectinomycin at a final concentration of 50 μg/mL.
S. venezuelae exconjugants were picked from ISP-2 plates and used to inoculate 5 mL of CSM supplemented with antibiotics as needed (apramycin or spectinomycin, final concentration 50 μg/mL). These seed cultures were grown for 2-3 days at 30 °C, then diluted to an optical density at 600 nm (OD600) of 0.01 in fresh CSM supplemented with antibiotics, and grown for 24 hours. After 24 hours, 25 μL of each culture were transferred to 75 μL of fresh CSM media supplemented with antibiotics as needed inside a 96 well microplate (Costar). OD600 (OD) and fluorescence (FL) (excitation 587 nm and emission 610 nm) were measured in a microplate reader (Tecan Infinite M1000 Pro). When performing CRISPRi and CRISPRa experiments, a S. venezuelae strain harboring a previously integrated mCherry reporter was used as recipient for conjugation. After conjugation, fluorescence measurement experiments were carried out as described above.
Fluorescence data analysis
In each fluorescence measurement experiment, OD600 and FL values for each sample were corrected by subtracting the mean OD and FL values of a media blank. The ratio of FL to OD600 (FL/OD) was then calculated. Data are reported as mean FL/OD values for each condition, and error bars represent standard deviation (s.d.).
Growth curve experiments
S. venezuelae colonies were grown at 30 °C with 250 rpm shaking in 5 mL CSM supplemented with apramycin or spectinomycin as needed. Upon reaching high cell density (~2 days), cells were diluted in the same medium to an OD600 of 0.08 directly in a Costar 96 well microplate (final volume 100 μL). Cells were then grown in a microplate reader (Tecan Spark multimode plate reader) at 30 °C with 90 rpm shaking for 24 hours. OD600 measurements were automatically taken every 10 minutes. For data analysis, OD600 values of the media controls were averaged at each time point, and then subtracted from the individual values of each condition at each time point.
Distance-dependent CRISPRa experiments
To evaluate distance-dependent effects, sgRNAs were designed to target sequences proximal to PAMs positioned on the non-template strand at a distance of 73, 83, 93 bp upstream of the reporter promoter’s TSS. Distances do not include either the PAM or the TSS. The same sgRNAs were used to target sequences upstream of a reporter that contained 5 additional nucleotides to extend the distance between the TSS and each PAM by 5 bp (thus creating 78 and 88 bp binding sites). Fluorescence measurements were performed as described above, and FL/OD for each sample were calculated. As each reporter had slightly different mCherry expression values, data are reported as fold activation. Fold activation was calculated by normalizing FL/OD values of each sample to the mean FL/OD values of the no-CRISPRa control of each reporter. Corresponding FL/OD values used for fold activation calculations are shown in Supplementary figure 4.
Fermentation for jadomycin B production
Fermentations were conducted under previously described conditions60,61. CRISPRi or CRISPRa plasmids carrying appropriate sgRNAs were conjugated into wild-type S. venezuelae, as described above. Exconjugants were then picked to inoculate 100 mL of CSM supplemented with apramycin, and grown to high density (usually 3-5 days) in 250 mL Erlenmeyer flasks at 30 °C. Cultures were then centrifuged, and the pellets washed twice in MSM to remove any trace of CSM. After resuspension in 6 mL of MSM, each sample was diluted in MSM supplemented with apramycin to an OD600 of 0.6 and a final volume of 50 mL. Fermentation cultures were incubated at 30 °C with 250 rpm shaking. After 72 hours, cultures were centrifuged and the pellets were stored at −80 °C for LC-MS analysis.
Extraction and LC-MS analysis
Pellets obtained from fermentations were extracted with an equal volume of acetone. Mixtures were transferred to an Erlenmeyer flask and shaken at room temperature for 30 minutes at 180 rpm. Acetone was then evaporated in a rotovap at 40 °C and 200 mbar. Crude extracts were reconstituted in 5 mL of acetonitrile, and 500 μL of the reconstituted extract was combined with 500 μL of LC-MS grade water. LC-MS was carried out on an Agilent 6470B Triple Quadrupole Mass Spectrometer interface to an Agilent 1290 Infinity II HPLC system through a Jet Spray ESI source. The LC column was an Agilent 2.1 mm ID x 50 mm, 1.8 μm SB-C18 column. Mobile phase A was water with 0.1% formic acid, and mobile phase B was acetonitrile with 0.1% formic acid. LC flow rate was 0.4 ml/min. Initial conditions were 10% B up to 95% B over three minutes. The column was flushed with 95% B from 3-6 minutes and returned to initial conditions from 6-6.5 minutes. The column was equilibrated at initial conditions from 6.5-9.5 min.
Experiments were performed using at least three biological replicates per condition tested unless otherwise indicated. Data for figures 1b, 1c, 1d, and 2b are reported as bars showing mean, with error bars showing standard deviation. Individual samples are also shown as black circles. Jadomycin B production experiments were performed in biological duplicates. Data in figures 3d and 3e show extracted ion chromatograms of one representative biological sample for each condition. All remaining biological replicates for LC-MS experiments are shown in Supplementary figures 5-6.
Article TitleActivating natural product synthesis using CRISPR interference and activation systems in Streptomyces
The rise of antibiotic-resistant bacteria represents a major threat to global health, creating an urgent need to discover new antibiotics. Natural products derived from the genus Streptomyces represent a rich and diverse repertoire of chemical molecules from which new antibiotics are likely to be found. However, a major challenge is that the biosynthetic gene clusters (BGCs) responsible for natural product synthesis are often poorly expressed under laboratory culturing conditions, thus preventing isolation and screening of novel chemicals. To address this, we describe a novel approach to activate silent BGCs through rewiring endogenous regulation using synthetic gene regulators based upon CRISPR-Cas. First, we create CRISPR interference (CRISPRi) and CRISPR activation (CRISPRa) systems that allow for highly programmable and effective gene repression and activation in Streptomyces. We then harness these tools to activate a silent BGC through perturbing its endogenous regulatory network. Together, this work advances the synthetic regulatory toolbox for Streptomyces and facilitates the programmable activation of silent BGCs for novel chemical discovery.