FIGURE 2 Design and selection of NMN biosynthetic pathways. (a) Design of three pathways for NMN biosynthesis. (b) Standard Gibbs free energy changes for NMN biosynthesis pathways. (c) Using 8 pathway combinations to produce NMN. HPM, HsRbk + PcPrs + MrNampt; EPM, EcRbk + PcPrs + MrNampt; HHM, HsRbk + HsPrs + MrNampt; EHM, EcRbk + HsPrs + MrNampt; HPH, HsRbk + PcPrs + HsNampt; EPH, EcRbk + PcPrs + HsNampt; HHH, HsRbk + HsPrs + HsNampt; EHH, EcRbk + HsPrs + HsNampt. See Table S2 for details. Error bars represent the standard deviation of three biological replicates. Adk, adenosine kinase; Apt, adenine phosphoribosyltransferase; Rbk, ribokinase; Prs, phosphoribosyl pyrophosphate synthetase; Nampt, nicotinamide phosphoribosyltransferase. PPi, pyrophosphate.
3.2 Adaptation of split GFP assay to monitor protein expression in CFPS
The protein production of CFPS was usually monitored and quantified by using radioactive amino acid incorporation (Jewett & Swartz, 2004; Karim & Jewett, 2018), which allowed researchers to control and assess pathway performance in a precise manner in the CFPS-ME framework (Grubbe et al., 2020; Karim et al., 2020; Rasor et al., 2022). However, radioactive incorporation was unavailable for many laboratories and its procedure was laborious. To overcome the limitations of radioactive incorporation, we adapted a split GFP assay to monitor CFPS protein production . In brief, a 12-amino acid linker and a 16-amino acid GFP-derived peptide tag (GFP11) were cloned to the C terminus of the gene encoding the target protein. The fusion protein was produced via CFPS, while a truncated, non-fluorescing detector GFP protein (GFP1-10) was produced separately. Combining the CFPS cell lysates which contained the GFP11-tagged protein with the detector for a short time would elicit a quantifiable fluorescent signal. Hence, protein production of CFPS could be monitored simply by fluorescence determination (Figure 3a).
As a proof-of-principle, the split GFP assay was applied to monitor the expression levels of 6 enzymes included in the initial enzyme sets for NMN synthesis. An 84 nucleotide sequence encoding the 12-amino acid linker and 16-amino acid “GFP11” tag was added directly at the end of the coding sequence of Hs/MrNampt, Hs/PcPrs, and Hs/EcRbk before the stop codon. The constructed sequences were cloned into the pET-28a vector at the Nde I and Xho I sites (Figure S1) by the synthesis company. CFPS reactions were performed by using these enzyme expression plasmids as DNA templates and incubated for 16 h at 30°C. When the CFPS reactions were complete, the enzyme-enriched CFPS cell lysates were mixed and incubated with a GFP1-10 detector solution. Compared to the “Blank” CFPS reaction (using water instead of expression template), a combination of six enzyme-enriched cell lysates and the GFP1-10 detector all emitted significant fluorescence as expected. More importantly, the fluorescence was directly proportional to the amount of added enzyme-enriched cell lysates (Figure 3b). To further assess the sensitivity and accuracy of the assay, serial dilutions of purified GFP11-tagged HsRbk (HsRbk-LG) were prepared, and each dilution was then added to the GFP1-10 detector solution to regenerate GFP fluorescence. As shown in Figure 3c, there always was a strong correlation (R2>0.99) between the complementation fluorescence and the amount of HsRbk at different incubation time points (ranging from 8 to 16h). In addition, the amount of detected protein could reach the picomolar level, which indicated that this approach was highly sensitive and accurate. Taken together, these results suggested that the split GFP assay allowed to monitor protein expression in CFPS in a convenient and sensitive way.