We previously reported that SLC19A1 is an importer of the immunotransmitter 2'3'cyclic-GMP-AMP (cGAMP) 1 by performing a genome wide screen in U937 cells. Soon after, Lutejin et al. reported similar findings by conducting a screen in THP-1 cells 2 . While the conclusions of these two studies largely overlap, we arrived at significantly different conclusions regarding how broadly SLC19A1 is used by different cell types. Our study suggests that in addition to SLC19A1, many cultured and primary cell types use alternative, unidentified transporters to import cGAMP and other cyclic dinucleotides (CDNs). This conclusion was based on our findings that inhibition of SLC19A1 did not significantly reduce extracellular cGAMP signaling in multiple cell types, including primary CD14 + peripheral blood mononuclear cells (PBMCs) from most donors. In contrast, Luteijn et al. concluded that SLC19A1 is the major CDN importer in humans, largely based on their use of a radiolabeled [ 32 P] cGAMP uptake assay. Using this assay, they showed that inhibition of SLC19A1 abolishes [ 32 P] uptake in total PBMCs. However, they did not test whether inhibition of SLC19A1 affects extracellular cGAMP signaling in these cells. Here, we highlight an important issue with the [ 32 P] cGAMP uptake assay used by Luteijn et al. and demonstrate that measuring extracellular cGAMP signaling through the STING pathway is currently the best method for evaluating cGAMP import. We also show that inhibition of SLC19A1 has no effect on extracellular cGAMP signaling in total PBMCs, confirming that this cell type relies on other transport mechanisms for cGAMP import.
ResultscGAMP is not stable in serum-containing tissue culture media (hereafter referred to simply as media) due to the presence of its hydrolase ENPP1 in serum 3 , and it is degraded into AMP and Pi, both of which are reported substrates of SLC19A1 4 . Our analysis using thin layer chromatography (TLC) confirmed that in media a small percentage of [ 32 P] cGAMP is degraded into [ 32 P] AMP and [ 32 P] Pi while STF-1623, an ENPP1 inhibitor we developed 5 , efficiently blocks degradation ( Fig. 1a). Therefore, an uptake assay without ENPP1 inhibitors could be measuring uptake of cGAMP degradation products. This may explain why Luteijn et al. were able to measure [ 32 P] uptake for up to five hours without reaching equilibrium 2 , despite reports that uptake of other SLC19A1 substrates reaches equilibrium within an hour 4,6 . To investigate further, we found that preincubation of [ 32 P] cGAMP in media before performing an uptake assay increased [ 32 P] uptake, with a linear correlation between the preincubation time and the amount of [ 32 P] uptake. Furthermore, adding STF-1623 to the preincubation abolished the increase in [ 32 P] uptake, indicating that the increased [ 32 P] uptake is due to degradation of [ 32 P] cGAMP ( Fig. 1b). Importantly, STF-1623 did not inhibit uptake of [ 3 H] methotrexate (MTX), a model substrate of SLC19A1, indicating that STF-1623 does not inhibit cGAMP uptake through [ 3...