Cell Free Bacteriophage Synthesis from Engineered Strains Improves Yield

Abstract

Phage therapy to treat life-threatening drug-resistant infections has been hampered by technical challenges in phage production. Cell-free bacteriophage synthesis (CFBS) can overcome the limitations of standard phage production methods by manufacturing phage virions in vitro. CFBS mimics intracellular phage assembly using transcription/translation machinery (TXTL) harvested from bacterial lysates and combined with reagents to synthesize proteins encoded by a phage genomic DNA template. These systems may enable rapid phage production and engineering to accelerate phages from bench-to-bedside. TXTL harvested from wild type or commonly used bacterial strains was not optimized for bacteriophage production. Here, we demonstrate that TXTL from genetically modified E. coli BL21 can be used to enhance phage T7 yields in vitro by CFBS. Expression of 18 E. coli BL21 genes was manipulated by inducible CRISPR interference (CRISPRi) mediated by nuclease deficient Cas12a from F. novicida (dFnCas12a) to identify genes implicated in T7 propagation as positive or negative effectors. Genes shown to have a significant effect were overexpressed (positive effectors) or repressed (negative effectors) to modify the genetic background of TXTL harvested for CFBS. Phage T7 CFBS yields were improved by up to 10-fold in vitro through overexpression of translation initiation factor IF-3 (infC) and small RNAs OxyS and CyaR and by repression of RecC subunit exonuclease RecBCD. Continued improvement of CFBS will mitigate phage manufacturing bottlenecks and lower hurdles to widespread adoption of phage therapy.

Keywords: CRISRPi; T7; TX-TL; cell-free bacteriophage synthesis (CFBS); cell-free expression systems; gene expression.

Figure 1

Gene expression background in the TXTL source influences CFBS yields. The number of progeny resulting from bacteriophage T7 infection of host E. coli BL21 is influenced by the genetic background of the host at the time of infection. Knockdown (KD) or overexpression (OX) of certain genes may have a positive, negative, or neutral impact on T7 progeny yield relative to wild-type BL21 (WT) (a). Cell-free bacteriophage synthesis (CFBS) can be modulated by modifying the genetic background of the source of transcription and translation (TXTL) machinery derived from cell lysates (b).

Figure 2

Inducible CRISPRi of trxA knockdown mRNA and T7 efficiency of plating (EOP). Induced expression of dFnCas12a (a) with crRNA targeting the trxA promoter results in repression of trxA by 90 ± 2.4% (b), which is associated with a significant efficiency of plating (EOP) reduction to 12 ± 7.0% (c). Data represented as mean ± SD (n = 3). Rhamnose induction significantly lowers trxA expression vs the NT-control (p < 0.0001, 2-way ANOVA). Welch’s two-tailed t test was performed indicating significant reduction of EOP (p < 0.05). p ≤ 0.05, *; p ≤ 0.01, **; p ≤ 0.001, ***.

Figure 3

CRISPRi-mediated gene repression in E. coli BL21 has varied impact on T7 lysis onset time and mean lysis time (T50%). CRISPRi induced by 2% (w/v) L-rhamnose for 4 h changes lysis profiles of T7 infecting log-phase E. coli BL21 depending on gene target and whether crRNA targets promoters (odd numbers) or coding sequences (even numbers). A nontargeting control was included in each experiment (g037). Optical densities (OD) were normalized with a max OD = 1 for clearer comparisons. Shaded regions represent the max and min OD of triplicate experiments performed on different days. Representative curve effects (a): neutral (NT-control), shift (trxA (g001)), widening (mukB (g012)), and shoulder (infC (g020)). Lysis onset time (b) and mean lysis time (c) are represented as mean ± SD (n = 3). Welch’s two-tailed t test was performed indicating significant change in lysis and mean lysis timing for as a result of CRISPRi induction (p < 0.05). p ≤ 0.05, *; p ≤ 0.01, **.

Figure 4

Gene overexpression and CRISPRi-mediated gene repression modulates efficiency of plating (EOP). Log-phase E. coli BL21 knockdown (a) and overexpression (b) strains were infected with phage T7 at MOI = 0.0001. After 30 min (roughly one lysis cycle) at 37 °C, further phage replication was stopped by adding chloroform to each culture followed by pipet mixing. Phage yields were enumerated by plaque counting on double-layer agar plates containing E. coli BL21 as a propagation host. Efficiency of plating (EOP) represents the fraction change in assembled phage between gene-repressed CRISPRi strains and their nonrepressed counterparts carrying the same plasmid without CRISPRi induction (a) or fraction change in assembled phage between overexpressing strains and their noninduced counterparts carrying the same plasmid (b). Data represented as mean ± SD (n = 3). Welch’s two-tailed t test was performed indicating significant change in EOP as a result of CRISPRi induction (p < 0.05). p ≤ 0.05, *; p ≤ 0.01, **; p ≤ 0.001, ***; p ≤ 0.0001, ****.

Figure 5

T7 genome pJl1-T7-Pr-sfGFP cascade phage-dependent TXTL mechanism. A cell-free expression system synthesizing sfGFP was used to troubleshoot CFBS reaction. 0.5 nM T7 genome was included in a standard CFES reaction using sfGFP expression controlled by T7 RNA polymerase (RNAP) (a). Transcription of sfGFP occurs only if T7 RNAP (T7 gene 1) itself is transcribed by endogenous E. coli RNAP (b). sfGFP expression reflects functional endogenous TXTL capacity and T7 RNAP synthesis and transcription activity.

Figure 6

Cell-free protein and bacteriophage synthesis yields are influenced by genetic background of the TXTL donor. Biological duplicates (1 and 2) of cell lysates were prepared from E. coli BL21 overexpression (UP ▲) or repression (DOWN ▼) of various effector genes. Protein synthesis (a) was carried out using the T7 genome pJl1-T7-Pr-sfGFP cascade (Figure 5). CFBS yields (b) were measured by plaque count assays using wild-type BL21 as a propagation host. Gray bars indicate protein and phage yields at 4 h. Black dotted lines indicate 4 h yields in wild-type BL21 lysates and colored lines indicate 20 h yields. Data show mean ± SD (n = 3). Welch’s two-tailed t test was performed, indicating significant increase in sfGFP fluoresce or T7 titer relative to the wild-type E. coli BL21 baseline (p < 0.05). p ≤ 0.05, *; p ≤ 0.01, **; p ≤ 0.001, ***; p ≤ 0.0001, ****.