High-throughput genetic engineering of nonmodel and undomesticated bacteria via iterative site-specific genome integration

Abstract

Efficient genome engineering is critical to understand and use microbial functions. Despite recent development of tools such as CRISPR-Cas gene editing, efficient integration of exogenous DNA with well-characterized functions remains limited to model bacteria. Here, we describe serine recombinase-assisted genome engineering, or SAGE, an easy-to-use, highly efficient, and extensible technology that enables selection marker-free, site-specific genome integration of up to 10 DNA constructs, often with efficiency on par with or superior to replicating plasmids. SAGE uses no replicating plasmids and thus lacks the host range limitations of other genome engineering technologies. We demonstrate the value of SAGE by characterizing genome integration efficiency in five bacteria that span multiple taxonomy groups and biotechnology applications and by identifying more than 95 heterologous promoters in each host with consistent transcription across environmental and genetic contexts. We anticipate that SAGE will rapidly expand the number of industrial and environmental bacteria compatible with high-throughput genetics and synthetic biology.

Fig. 1.. Iterative integration of multiple genetic constructs into the chromosome using SAGE.

This schematic depicts the SAGE cycle for rapid, efficient, unidirectional, and site-specific integration of multiple DNA fragments into bacterial chromosomes. To construct a base strain, attB sequences (here, a 10-sequence poly-attB cassette) are integrated into the target microbe’s genome using standard methods (e.g., allelic exchange or transposase integration). Next, one of the system’s 10 recombinases is transiently expressed from a nonreplicating helper plasmid, catalyzing recombination between its cognate attP and attB sequences, located on a cotransformed target plasmid and the host chromosome, respectively. Recombination integrates the attP target plasmid into the genome unidirectionally, generating attL and attR sequences. To excise unneeded plasmid features and prime the strain for additional SAGE cycles, ΦC31 integrase is transiently expressed by transforming a nonreplicating backbone excision plasmid. ΦC31 integrase catalyzes recombination between ΦC31 attP and attB sites (P and B, respectively, in blue) of the integrated target plasmid, excising the selection marker and the E. coli origin of replication from the chromosome.

Fig. 2.. SAGE enables stable, highly efficient integration of plasmid DNA into the genomes of engineered bacteria.

(A) Diagram of genome-integrated 10× poly-attB cassette, including terminators for transcriptional insulation. Each attB sequence is indicated by a color-coded box and is flanked by a random 20-nt DNA spacer sequence. (B) Diagram of Tn5 poly-attB insertion cassettes. Cassettes are flanked by Tn5 direct repeat (DR) sequences and contain an antibiotic resistance cassette (either nptII or aac) upstream of the poly-attB cassette from (A). A cognate pair of ΦC31 att sites flanks the resistance cassette, allowing its unidirectional excision by electroporation of the ΦC31 integrase helper plasmid pJE1817. (C) Plasmid maps of SAGE plasmids used for efficiency experiments in (D). (D) Transformation efficiency when the poly-attP target plasmid pGW60 is transformed with or without an integrase-expressing helper plasmid or when the positive control plasmid is transformed. Control plasmids are as follows: pJE354 (SBW25, TBS10, and Gpo1), pEYF2K (CGA009), or pSET152 (RHA1). Error bars indicate the two-sided SD in three or more biological replicates. Dots indicate individual samples. The accuracy of integration represents the fraction of colonies in which pGW60 recombined into the poly-attB cassette rather than a pseudo-att site, as determined by PCR. With the exception of BT1 in Gpo1 and RV/F370 in RHA1 (which used 12, 20, and 18 samples for screening, respectively), integration accuracy represents the fraction of 24 samples with colony PCR screening results indicating insertion at the intended attB site. n.d. indicates samples not assayed by PCR because of low numbers of transformants across plate replicates. CFU, colony-forming units.

Fig. 3.. Transient expression of ɸC31 integrase excises the attP plasmid backbone and enables selection marker recycling.

(A) Diagram of ΦC31 integrase–mediated excision of an attP target plasmid backbone. Specifically, pJE1818 was inserted at the Bxb1 attB of a chromosomally integrated poly-attB cassette. ΦC31 integrase was expressed from the nonreplicating helper plasmid pJE1817. Colony PCR primers are indicated by arrows and red numbers. (B and C) Colony PCR validation of plasmid backbone excision in sucrose-resistant (B) P. fluorescens JE4621– and (C) R. palustris JE4632–based strains following SAGE integration and incubation on the sucrose-containing medium. Expected band sizes are as follows: JE4866 [4741 base pairs (bp)], JE4866 with backbone excision (1152 bp), JE4621 (no band), JE4885 (4964 bp), and JE4855 with backbone excision (1375 bp).

Fig. 4.. SAGE enables stable expression of heterologous genes in the absence of selection.

Flow cytometry measurements of fluorescence in populations of P. fluorescens JE4621 containing either no heterologous expression construct (parent strain), a pBBR-based replicating mNeonGreen expression vector (pJE354), or a nonreplicating mNeonGreen expression vector that has been chromosomally integrated at the indicated attB site by the corresponding recombinase. Starter cultures, with the exception of the parent strain that was grown in the absence of antibiotic, were cultivated in medium containing kanamycin sulfate (50 μg/ml) for plasmid selection. Each passage was inoculated with a 1024-fold dilution of its precursor culture (starter → passage 1 → passage 2) to enable 10 generations of exponential growth before reaching stationary phase. The x axis indicates the relative fluorescence (excitation at 488 nm and emission at 530 nm) of each cell. The y axis represents the abundance of cells in the population with a given relative fluorescence unit (RFU).

Fig. 5.. Development and high-throughput analysis of genome-integrated promoter libraries.

(A) Overview of promoter library construction and DNA/RNA sequencing (RNA-seq) barcode sequencing fragments. (B) Example promoters with different classes of 5′UTR (level of barcode-associated “noise”) and condition sensitivity. Large, outlined circles represent mean values, and small dots represent individual samples. (C) Chart displaying mean RTA for consistent (5′UTR and condition insensitive) promoters in P. fluorescens SBW25. Relative strength of the promoters used in (D) is indicated. Error bars represent two-sided SD of between 36 and 80 samples (see source data file and Supplementary File D1 for exact numbers). (D) Promoter performance for a small subset of pLibrary promoters in microtiter plate growth assays. Error bars represent two-sided SD in three replicates. Relative promoter activity is calculated by comparing mean RFU/OD₆₀₀ values across all carbon sources. (E) Correlation between relative expression levels determined by RTP and fluorescent protein reporter assay for the set of promoters used in (D). Linear equation and coefficient of determination between the same promoters using RTA data from (C) and fluorescent plate reader data from (D), as determined by Pearson correlation.

Fig. 6.. Promoter activity correlates with phylogenetic distance.

(A) Coefficient of determination (R²) generated when comparing RTA of promoters between promoters that are characterized in each organism pair. R² values are provided for both sample sets including all promoters or for a subset of promoters that only includes promoters that were consistent in both organisms. (B) Example scatterplot of data used to perform Pearson correlation and the linear equation and coefficient of determination generated using Pearson correlation using consistent promoters from P. fluorescens and P. frederiksbergensis.