Refined control of CRISPR-Cas9 gene editing in Clostridium sporogenes: the creation of recombinant strains for therapeutic applications

Abstract

Despite considerable clinical success, the potential of cancer immunotherapy is restricted by a lack of tumour-targeting strategies. Treatment requires systemic delivery of cytokines or antibodies at high levels to achieve clinically effective doses at malignant sites. This is exacerbated by poor penetration of tumour tissue by therapeutic antibodies. High-grade immune-related adverse events (irAEs) occur in a significant number of patients (5-15%, cancer- and therapeutic-dependent) that can lead to lifelong issues and can exclude from treatment patients with pre-existing autoimmune diseases. Tumour-homing bacteria, genetically engineered to produce therapeutics, is one of the approaches that seeks to mitigate these drawbacks. The ability of Clostridium sporogenes to form spores that are unable to germinate in the presence of oxygen (typical of healthy tissue) offers a unique advantage over other vectors. However, the limited utility of existing gene editing tools hinders the development of therapeutic strains. To overcome the limitations of previous systems, expression of the Cas9 protein and the gRNA was controlled using tetracycline inducible promoters. Furthermore, the components of the system were divided across two plasmids, improving the efficiency of cloning and conjugation. Genome integrated therapeutic genes were assayed biochemically and in cell-based functional assays. The potency of these strains was further improved through rationally-conceived gene knock-outs. The new system was validated by demonstrating the efficient addition and deletion of large sequences from the genome. This included the creation of recombinant strains expressing two pro-inflammatory cytokines, interleukin-2 (IL-2) and granulocyte macrophage-colony stimulating factor (GM-CSF), and a pro-drug converting enzyme (PCE). A comparative, temporal in vitro analysis of the integrant strains and their plasmid-based equivalents revealed a substantial reduction of cytokine activity in chromosome-based constructs. To compensate for this loss, a 7.6 kb operon of proteolytic genes was deleted from the genome. The resultant knock-out strains showed an 8- to 10-fold increase in cytokine activity compared to parental strains.

Keywords: CRISPR-Cas9; Clostridium sporogenes; cancer; cytokines; immunotherapy; secretion.

Figure 1

Schematic representation of the two-plasmid gene editing system. (A) pTetR-P _IPL12 -Cas9 carries the native SpCas9, modified to remove a BsmBI restriction site. The tet repressor gene (tetR) is expressed constitutively and binds to the tet operator (tetO) sequence of the synthetic P _IPL12 promoter in the presence of (anhydro)tetracycline. This is the basis for control of cas9 expression. Both vectors utilised backbones from pMTL8000 vectors (54); gRNA, guide RNA consisting of spacer and scaffold sequences; LHA/RHA: left/right homology arm; lacZα, β-galactosidase counter-selection marker (for cargo cloning). (B) A representative workflow for gene editing using the CRISPR-Cas system. Conjugative transfer of pTetR-P _IPL12 -Cas9 into the host is conducted prior to and separately from that of p8222F-gX-HC. In parallel, a gRNA spacer sequence and the DNA (that is to be integrated) is cloned into p8222F-gX-HC. Following conjugative transfer of p8222F-gX-HC, viable colonies are restreaked to induction plates, which contain Erm, Tm, and aTc, and incubated for 24h. Cultures able to grow on induction plates are screened by PCR to detect genomic recombination. Plasmid loss is achieved by subculture in the absence of antibiotic selection. If Erm selection is maintained, pTetR-P _IPL12 -Cas9 is retained by the host and additional rounds of gene editing can be conducted. (B) created with BioRender.com.

Figure 2

The effect of the size and the composition of four different vectors on the conjugation efficiencies of two recipients: empty C. sporogenes WT and strain bearing the inducible Cas9-plasmid. p8222F (4,998 bp): plasmid with only P _fetO promoter; p8222F-g7 (5,128 bp): plasmid with addition of sgRNA targeting SLS operon; p8222F-g7-SLS (6,567 bp): plasmid targeting SLS operon with repair cassette; p8222F-g7-SLS-3kb (9,698 bp): plasmid targeting SLS operon with repair cassette and 3 kb λDNA cargo. The transconjugants were counted on selection plates: Erm/Cs for C. sporogenes WT (in turquoise) and Erm/Cs/Tm for C. sporogenes-pTetR-P _IPL12 -Cas9 (in orange) with and without inducer (aTc, 96 ng/ml). Approximate conjugation efficiencies were calculated as obtained transconjugant CFU/total C. sporogenes recipient CFU. Each bar represents the mean and standard deviation of data collected from two experiments performed using biological duplicates. Statistical analysis was performed using unpaired t-test, ns – not significant, ** P<0.01.

Figure 3

Characterisation of growth, sporulation, and enzymatic activity assay in C. sporogenes strains. (A) Growth of Clostridium strains was measured as a direct increase in absorbance at 600 nm throughout the course of 24-hour bacterial incubation in bovine free medium (PYT). The symbols represent the average of three independent experiments, and error bars indicate the standard errors of the means. (B) The summary of OD₆₀₀ measurements recorded at the times of pre-determined experimental time point sample collections (5h, 7h and 10h). (C) Spore titres after 120h incubation in PYT broth. Samples of heat-treated cultures (80°C, 20 min) were plated in serial dilution on agar plates and enumerated following 24-48h incubation. Bars represent the number of CFU (colony forming unit) per ml of culture. The data represent the average of three independent experiments and error bars indicate the standard error of the mean. The sporulation-deficient Clostridium sporogenes-NT-Δspo0A mutant was used as a negative control to rule out experimental error. The detection limit for colony counts was 50 CFU/ml. (D) Menadione reductase activity assays on C. sporogenes cell lysates containing either plasmid-based or chromosome-integrated nfrA gene. Results show specific menadione reductase activities of C. sporogenes NfrA-expressing strains soluble fractions obtained from 7h time points during the growth of bacterial cultures. The C. sporogenes-NT-ΔpyrE was used as background control. Activities were normalised to total protein concentrations, determined using BCA assay. Activities are expressed in units per mg. The data represent the average of three independent experiments and error bars indicate the standard errors of the means. Statistical analysis was performed using unpaired t-test, *** P<0.001.

Figure 4

Validation of C. sporogenes-NT-ΔpyrE strains harbouring either plasmid-based or integrated murine cytokines (mIL-2 and mGM-CSF) at the SLS locus in the quantitative and functional assays. (A) Results of commercial ELISA tests indicating the quantities of mIL-2 and mGM-CSF cytokines present in the supernatants of cytokine bearing-C. sporogenes strains at 5-,7-, and 10-hour growth in PYT media. (B) Results of MTT and AlamarBlue functional assays following the incubation of cytokine specific T-cells in the presence of culture supernatants of C. sporogenes strains. Recombinant murine IL2 and GM-CSF were used to prepare mIL-2 and mGM-CSF standard curves., (C) denotes C. sporogenes-NT-ΔpyrE control, (BDL) - below detectable levels. Statistical analysis was performed using unpaired t-test, * P<0.05, ** P<0.01, *** P<0.001.

Figure 5

Schematic representation of nprM proteolytic operon knock-out in C. sporogenes mediated by inducible two-plasmid CRISPR-Cas9 system. (A) Deletion of six proteolytic genes has been exemplified with C. sporogenes-NT-ΔpyrE strain and subsequently repeated using two cytokine-bearing integrant strains. Triangles represent indicative alignment of screening primers as presented in Table S2 . (B) Gel electrophoresis showing PCR screens of all three proteolytic knock-out strains conventionally labelled as ΔPR1, where S1: CspNTΔPΔPR1 (C. sporogenes-NT-ΔpyrEΔPR1), S2: CspNTΔPΔPR1::mIL-2 (C. sporogenes-NT-ΔpyrEΔPR1::mIL-2) and S3: CspNTΔPΔPR1::mGM-CSF (C. sporogenes-NT-ΔpyrEΔPR1::mGM-CSF). Left image: Deletion of the operon detected in colonies subjected to a PCR reaction using primers flanking the deletion locus. Right image: Subsequent confirmation of proteolytic operon deletion in selected clones using internal (junction) and flanking primers. (L) denotes DNA marker, (C1-C3): equivalent C. sporogenes-NT-ΔpyrE controls.

Figure 6

Analysis of proteolytic activity in C. sporogenes strains. (A) Qualitative analysis of proteolytic activity by gelatin zymography of the C. sporogenes-NT-ΔpyrE and C. sporogenes-NT-ΔpyrEΔPR1 supernatants at four different time points (5h, 7h, 10h and 12h). In total, 20 μl of samples were loaded on a SDS polyacrylamide gel containing 0.05% of gelatin. After electrophoresis, proteins were renaturated in the gel and the proteolytic activity was revealed after an overnight incubation required for gelatin degradation. CB10 - C. butyricum control supernatant collected at 10h growth. Samples were normalised according to OD₆₀₀ of 1.2 prior to loading. M: Pre-stained Protein Ladder PageRuler™ (Thermo Scientific™); P⁺: C. sporogenes-NT-ΔpyrE samples (PR1 positive); P^-: C. sporogenes-NT-ΔpyrEΔPR1 (PR1 negative); CB10 - C. butyricum control supernatant collected at 10h growth. (B) Quantitative analysis of proteolytic activities detected in C. sporogenes strains using Pierce™ colorimetric protease assay (Thermo Scientific™) and sample supernatants collected at three different time points (5h, 7h, and 10h). CspNTΔP (P⁺) and CspNTΔPΔPR1 (P^-) – PR1-positive and PR1-negative backgrounds, respectively. Standard curve for the assay has been constructed using serially diluted TPCK trypsine stock solution. The absorbance measurement has been carried out at 450 nm. The data represent the average of two independent experiments, and error bars indicate the standard errors of the means. Statistical analysis was performed using unpaired t-test, *** P<0.001.

Figure 7

Validation and comparison of C. sporogenes-NT-ΔpyrE and C. sporogenes-NT-ΔpyrEΔPR1 strains harbouring integrated murine cytokines (mIL-2 and mGM-CSF) in the quantitative and functional assays. (A) Results of commercial ELISA tests indicating the quantities of mIL-2 and mGM-CSF cytokines present in the supernatants of cytokine bearing-C. sporogenes strains at 5-,7-, and 10-hour growth in PYT media. (B) Results of MTT and AlamarBlue functional assays following the incubation of cytokine specific T-cells in the presence of culture supernatants of C. sporogenes strains. Recombinant murine IL-2 and GM-CSF were used to prepare mIL-2 and mGM-CSF standard curves. (C) denotes C. sporogenes-NT-ΔpyrEΔPR1 control, (BDL) – below detectable levels. The data represent the average of three independent experiments and error bars indicate the standard errors of the means. Statistical analysis was performed using unpaired t-test, ** P<0.01, *** P<0.001.