An expanded CRISPRi toolbox for tunable control of gene expression in Pseudomonas putida

Abstract

Owing to its wide metabolic versatility and physiological robustness, together with amenability to genetic manipulations and high resistance to stressful conditions, Pseudomonas putida is increasingly becoming the organism of choice for a range of applications in both industrial and environmental applications. However, a range of applied synthetic biology and metabolic engineering approaches are still limited by the lack of specific genetic tools to effectively and efficiently regulate the expression of target genes. Here, we present a single-plasmid CRISPR-interference (CRISPRi) system expressing a nuclease-deficient cas9 gene under the control of the inducible XylS/P_m expression system, along with the option of adopting constitutively expressed guide RNAs (either sgRNA or crRNA and tracrRNA). We showed that the system enables tunable, tightly controlled gene repression (up to 90%) of chromosomally expressed genes encoding fluorescent proteins, either individually or simultaneously. In addition, we demonstrate that this method allows for suppressing the expression of the essential genes pyrF and ftsZ, resulting in significantly low growth rates or morphological changes respectively. This versatile system expands the capabilities of the current CRISPRi toolbox for efficient, targeted and controllable manipulation of gene expression in P. putida.

Figure 1

Overview of the workflow for CRISPRi‐mediated gene knock‐down in Pseudomonas putida. Day 1: Choose a PAM sequence (5′‐NGG‐3′, where N represents any nucleotide) within the non‐template DNA strand sequence of the target gene. For efficient gene repression, we recommend to choose a PAM within the promoter sequence, or, if the promoter is not clearly defined or if it overlaps with a coding sequence, chose the PAM closest to the start ATG codon. Design and order two oligonucleotides for cloning of the spacer (upstream the PAM, following the format 5′‐spacer‐NGG‐3′) in either sgRNA or crRNA. Day 2: Anneal single‐strand DNA oligonucleotides by reverse cooling to form a double‐stranded DNA spacer‐insert array and clone it into the respective derivative of vector pCRi. Transform a suitable cloning E. coli strain and incubate the plates overnight (under streptomycin selection). Day 3: Inoculate three transformants in liquid cultures and grow the cultures overnight. Day 4: Purify three independent plasmids and send them out for sequence verification by DNA sequencing. Transform the target Pseudomonas strain with isolated and sequence‐verified plasmids, and incubate the plates under streptomycin selection. Day 5: Inoculate a fresh culture in the appropriate medium with verified Pseudomonas colonies and the additives needed, and incubate the cultures overnight. Day 6: Perform CRISPRi‐mediated gene downregulation in the presence of 1 mM 3‐methylbenzoate (or other inducer concentrations as needed) to activate the system

Figure 2

Overview of key expression vectors constructed for CRISPRi‐mediated knock‐down of gene expression in Pseudomonas putida. A. The sgRNA‐based, XylS/P_m‐inducible CRISPRi vector pMCRi. Vector pMCRi contains SpdCas9 under the control of the XylS/P_m expression system and a constitutively expressed sgRNA cassette. The sgRNA cassette is composed by the synthetic, constitutive P_EM7 promoter followed by the sgRNA chimera, spanning three domains: a 20‐nt region for target‐specific binding, a 42‐nt hairpin for dCas9 binding (dCas9 handle) and a 40‐nt transcription terminator (Sp Terminator) derived from S. pyogenes. To clone the target spacer, two BsaI recognition sites have been incorporated between the P_EM7 promoter and the sgRNA cassette B. The crRNA‐based CRISPRi vector pGCRi‐R. This vector contains a GC‐rich SpdCas9 gene (the expression of which is placed under control of the XylS/P_m expression system), a constitutively expressed crRNA cassette and the constitutively expressed tracrRNA. The crRNA cassette is formed by an AT‐rich sequence (leader) that contains a promoter driving the transcription of the crRNA (Pul et al., 2010) and the two direct repeats (DRs) with an intervening gene encoding the eforRed chromoprotein from Echinopora forskaliana (expressed from the constitutive BBa_J23100 Anderson promoter and equipped with the BBa_B0034 ribosome binding site). The crRNA cassette is flanked by two BsaI recognition sites to facilitate cloning the target spacer sequence. If the target spacer is successfully assembled by Golden Gate cloning into vector pGCRi, the eforRed reporter gene is split (and inactivated), and the resulting E. coli transformants will appear white instead of reddish when isolated on solid culture medium plates

Figure 3

Assembly of pCRiMs vectors. Double‐stranded sgRNA ultramers with unique target spacers are combined with a PCR‐amplified pMCRi vector for USER cloning reactions, resulting in a suite of vectors tailored for downregulation of several targets (i.e. pCRiMs plasmids)

Figure 4

Schematic representation of dsDNA spacer units and multiplex assembly strategy A. The annealed synthesized oligonucleotides consist of the 30‐nt target spacer sequence flanked with four and five nucleotides (indicated in grey) of the upstream and downstream direct repeats (DRs), respectively, and two BsaI recognition sites (highlighted in blue). The crRNA cassette also contains two BsaI recognition sites between the DRs with an intervening eforRed chromoprotein gene to facilitate the cloning of the intended double‐stranded DNA (dsDNA) spacer unit. B. Scheme of the dsDNA oligonucleotides required for the construction of a double crRNA array via Golden Gate cloning. The double‐stranded (ds) spacer₁‐repeat unit consists of the 30‐nt spacer₁ sequence (yellow) with additional four and six nucleotides of its upstream and downstream DRs (grey) respectively. The ds spacer₂‐repeat unit consists of the 30‐nt spacer₂ sequence (green) with six and five nucleotides of its upstream and downstream DRs (grey) respectively. The trimmed‐DR unit consists of the DR sequence lacking the nucleotides that are present in the spacer‐repeat units. After digestion with BsaI, all dsDNA oligonucleotides display compatible overhangs that specify their order and directionality as they are ligated to the target pGCRi vector via Golden Gate cloning

Figure 5

Titratable downregulation of msf·gfp expression with CRISPRi using inducible expression systems to drive SpdCas9 expression A. Schematic representation of derivatives of plasmid pCRi_gfp (plasmids pCCRi, pDCRi and pMCRi). All plasmids were separately introduced into P. putida KT·BG42, and three different expression systems were adopted: XylS/P_m (induced by addition of 3‐methylbenzoate), ChnR/P_chnB (induced by addition of cyclohexanone) and CprK1/P_DB3 (induced by addition of 3‐chloro‐4‐hydroxyphenylacetic acid). B. CRISPRi experiment on fluorescent proteins. P. putida KT·BG42 was used as a control [indicated in the figure as (−) CRISPRi], and grown on M9 minimal medium supplemented with 0.2% (w/v) glucose and 20 μg ml⁻¹ gentamicin. P. putida KT·BG42 transformants, harbouring the different CRISPRi vectors indicated, were re‐grown on M9 minimal medium supplemented with 0.2% (w/v) glucose, 100 μg ml^–1 streptomycin, 20 μg ml⁻¹ gentamicin and 1 mM of the corresponding inducer [indicated in the figure as (+) CRISPRi]. Bacterial growth and msfGFP fluorescence (λ_excitation/λ_emission = 485 nm/516 nm) were continuously measured during 15 h in a Synergy HI plate reader (BioTek Instruments, Inc., Winooski, VT, USA) using microtiter 96‐well plates incubated at 30°C. Fluorescence readings were normalized to the bacterial growth (estimated as the optical density measured at 600 nm). Each data point represents the mean value of the percentage of normalized fluorescence ± standard deviation from at least three biological replicates

Figure 6

Tunable effect of the CRISPRi system in P. putida KT·BG42 cells harbouring plasmid pMCRi with different inducer concentrations. A. Strain KT·BG42 cells transformed with plasmid pMCRi_gfp, harbouring the msf·gfp‐specific spacer [indicated in the figure as (+) CRISPRi] or with vector pMCRi_non‐target [harbouring a non‐target‐specific unique spacer, and indicated in the figure as (–) CRISPRi], were grown on M9 minimal medium with 0.2% (w/v) glucose, 100 μg ml^–1 streptomycin, 20 μg ml⁻¹ gentamicin and different concentrations of 3‐methylbenzoate (3‐mBz) in the 0 to 1 mM range. Bacterial growth and msfGFP fluorescence (λ_excitation/λ_emission = 485 nm/516 nm) were continuously measured during 15 h in a Synergy HI plate reader (BioTek Instruments, Inc., Winooski, VT, USA) using microtiter 96‐well plates incubated at 30°C. Fluorescence readings were normalized to the bacterial growth (estimated as the optical density measured at 600 nm). Each data point represents the mean value of the percentage of normalized fluorescence ± standard deviation from at least three biological replicates. B. Microscope pictures showing tunable, inducer‐dependent morphology changes of cells during CRISPRi‐mediated downregulation of ftsZ in wild‐type strain KT2440. Pictures were taken after 15 h with a Leica 2000 LED microscopy system (Leica Microsystems GmbH, Germany) at 100 × resolution (F1 type emission oil)

Figure 7

CRISPRi‐mediated downregulation of multiple gene targets (mCRISPRi) in P. putida KT·YFP·mCherry. For simultaneous repression of the expression of mCherry, yfp and ftsZ, cells containing the corresponding CRISPRi vector were grown in M9 minimal medium with 0.2% (w/v) glucose, supplemented with 100 μg ml^–1 streptomycin, 50 μg ml^–1 kanamycin and induced with 3‐methylbenzoate at 1 mM. In the figure, (–) mCRISPRi represents cells harbouring a non‐target‐specific pMCRi vector, and (+) mCRISPRi represents cells containing pCRiMs vector. A. Schematic representation of vectors pCRiMs and pCRiMc, harbouring target‐specific spacers. (B) Bacterial growth and mCherry fluorescence (λ_excitation/λ_emission = 567 nm/610 nm) and YFP fluorescence (λ_excitation/λ_emission = 495 nm/527 nm) were measured at 15 h in a Synergy HI plate reader (BioTek Instruments, Inc., Winooski, VT, USA) using microtiter 96‐well plates incubated at 30°C. Fluorescence readings were normalized to the bacterial growth (estimated as the optical density measured at 600 nm). Basal levels of fluorescence detected in P. putida KT2440 were also subtracted from the reading. Each bar represents the mean value of the percentage of normalized fluorescence ± standard deviation from at least three biological replicates. B. Microscope pictures showing morphology changes of cells during mCRISPRi‐mediated downregulation of ftsZ in wild‐type strain KT2440 after 15 h. Pictures were taken with a Leica 2000 LED microscopy system (Leica Microsystems GmbH, Germany) at 100 × resolution (F1 type emission oil)

Figure 8

CRISPRi‐based, targeted downregulation of pyrF expression in P. putida. Wild‐type KT2440 and the streamlined P. putida strain EM383 were transformed with either non‐target‐specific vectors pGCRi or pMCRi, or the corresponding target‐specific vectors (indicated with the suffix pyrF). The resulting strains were grown in 96‐well plates in M9 minimal medium supplemented with 0.2% (w/v) glucose, 100 μg ml^–1 streptomycin and 1 mM 3‐methylbenzoate. Bacterial growth was monitored in a Synergy HI plate reader (BioTek Instruments, Inc., Winooski, VT, USA) for 20 h at 30°C with shaking by periodically measuring the optical density at 600 nm (OD₆₀₀). Gene repression experiments are indicated for cells transformed with (A) plasmid pGCRi (no target and pyrF‐specific) and (B) plasmid pMCRi (no target and pyrF‐specific). Each data point represents the mean value of OD₆₀₀ readings ± standard deviation of at least three biological replicates