15Bacteria are central to human health and disease, but the tools available for modulating and 16 editing bacterial communities are limited. New technologies for tuning microbial populations 17 would facilitate the targeted manipulation of the human microbiome and treatment of bacterial 18 infections. For example, antibiotics are often broad spectrum in nature and cannot be used to 19 accurately manipulate bacterial communities. Bacteriophages can provide highly specific 20 targeting of bacteria, but relying solely on natural phage isolation strategies to assemble well-21 defined and uniform phage cocktails that are amenable to engineering can be a time-consuming 22 and labor-intensive process. Here, we present a synthetic-biology strategy to modulate phage 23 host ranges by manipulating phage genomes in Saccharomyces cerevisiae. We used this 24 technology to swap multiple modular phage tail components and demonstrated that Escherichia 25 coli phage scaffolds can be redirected to target pathogenic Yersinia and Klebsiella bacteria, and 26 conversely, Klebsiella phage scaffolds can be redirected to target E. coli. The synthetic phages 27 achieved multiple orders-of-magnitude killing of their new target bacteria and were used to 28 selectively remove specific bacteria from multi-species bacterial communities. We envision that 29 this approach will accelerate the study of phage biology, facilitate the tuning of phage host 30 ranges, and enable new tools for microbiome engineering and the treatment of infectious diseases. 31 32 65 their manipulation within bacterial hosts. Finally, all existing approaches are limited in the 66 number of mutations that can be introduced simultaneously. Multiple rounds of mutations are 67 therefore often required, making the process inefficient. Here, we demonstrate a high-throughput 68 phage-engineering platform that leverages the tools of synthetic biology to overcome these 69 challenges and use this platform to engineer model phages with tunable host ranges. 70 71
RESULTS
72Yeast platform for bacteriophage genome engineering. 73 We used an efficient yeast-based platform (Jaschke et al., 2012; Lu et al., 2013) to create phages 74 with novel host ranges based on common viral scaffolds. Inspired by gap-repair cloning in yeast 75 (Ma et al., 1987) and the pioneering work of Gibson and co-workers (Gibson, 2012; Gibson et al., 76 2008; Gibson et al., 2009), we captured phage genomes into Saccharomyces cerevisiae, thus 77 enabling facile genetic manipulation of modified genomes that can be subsequently re-activated 78 or "rebooted" into functional phages after transformation of genomic DNA into bacteria (Figure 79 1A). The workflow is split into two parts. In the first part, the entirety of the viral genome to be 80 assembled in yeast is amplified by PCR in such a way that each adjacent fragment has homology 81 over at least 30 bp. The first and last fragments of the phage genome are amplified with primers 82 that carry "arms" that have homology with a yeast artificial chromosome (YA...