As a further measure of validation, we co-injected cis and trans indole-isonitriles to samples where enzymatic product formation was observed (Figure 4, B8) and only the two product peaks that correlated to the retention times of cis and trans indole-isonitriles were observed. Finally, additional confirmation for indole-isonitrile biosynthesis was obtained through LC-MS analyses under negative ion-mode (Additional file 6). Overall,

the assay results validated the formation of cis and trans indole-isonitriles as the biosynthetic products of the pathway encoded by WelI1/I3. In contrast to AmbI1/3 and WelI1/3 from check details FA UTEX1903 and HW UTEXB1830 respectively, which only produce the cis isomer of the indole-isonitrile [7,8], assay mixtures containing WelI1/I3 from WI HT-29-1 produced both the cis and trans isomers of the indole-isonitrile when the assay is carried out over a 16 h duration. Because a mixture of cis and trans products are observed for the first time, these are exciting observations from a natural product biosynthesis point of view, as they lead to interesting questions about the biochemical mechanism

of WelI1/I3. It is probable that the enzymes are producing the trans isoform in concentrations below detection limit within the first 3 h, which then accumulate over time and can be detected after 16 h. However, it remains to be seen whether both of these isomers engage as substrates for downstream hapalindole-producing steps of the Neuronal Signaling inhibitor pathway. Figure 4 HPLC analyses of WelI1-WelI3 catalyzed indole-isonitrile formation. A) Biosynthetic steps catalyzed by WelI1 and WelI3 respectively. In vitro reconstitution assay for indole-isonitrile biosynthesis using cell lysates of E. coli BL21(DE3) heterologously expressing WelI1 and WelI3. Models of WelI1 and WelI3 were built based on homology to PvcA and PvcB X-ray structures [34] using Phyre2.0. B) HPLC was analyzed at 310 nm with a UV detector. X-axis – retention time in minutes (min). Y-axis – intensity in arbitrary units. Presented as a stacked Y-plot and is drawn to L-gulonolactone oxidase relative intensity units. B1) Synthesized cis indole-isonitrile

only (tR = 8.8 min). B2) Synthesized trans indole-isonitrile only (tR = 13.1 min). B3) Co-injection of synthetic standards of cis and trans indole-isonitrile. B4) Control for enzyme assay where cell lysates of E. coli BL21(DE3) were subjected to assay conditions without WelI1 and WelI3. B5) WelI1 and WelI3 enzyme assay after 16 h incubation at 25°C. B6) Control sample (4) spiked with cis indole-isonitrile after 3 h incubation. B7) Control sample (4) spiked with cis indole-isonitrile after 16 h incubation. B8) Co-injection of cis and trans indole-isonitrile with enzyme assay mixture. Peaks show only relative intensities and are not normalized for concentration of metabolites. Until now, direct evidence of the presence of indole-isonitriles from cyanobacterial cultures has remained elusive.