Structural and biochemical characterization of c-di-AMP synthesizing enzymes

One major concern of today’s life is the increase of antimicrobial resistance and the rising num-ber of multi drug resistant bacterial species. There is an urgent need of identifying new antibi-otic drug targets since resistances threaten the repertoire of available antibiotics. Cyclic di-AMP (c-di-AMP) is the only known essential second messenger mainly found in Gram-positive bacteria of which several are known as human pathogens (Woodward et al. 2010; Luo Y and Helmann 2012; Mehne et al. 2013; Gundlach et al. 2015a; Commichau et al. 2017; Gundlach et al. 2017a; Gundlach et al. 2017b; Commichau et al. 2019). It is involved in many cellular processes like cell wall metabolism and DNA integrity scanning (Corrigan Rebecca M and Gründling 2013; Commichau et al. 2019). c-di-AMP is synthesised by proteins containing diadenylate cyclase domains (DAC) (Romling 2008; Corrigan Rebecca M and Gründling 2013; Blötz et al. 2017; Commichau et al. 2019). CdaA is the sole DAC in the hu-man pathogen Listeria monocytogenes, which is also conserved in many other human patho-genic bacteria (Rosenberg et al. 2015; Heidemann et al. 2019; Tosi et al. 2019). Since c-di-AMP is also essential for the growth of these pathogenic bacteria, CdaA seems to be an attrac-tive target for the development of novel antibiotic compounds (Corrigan R. M. and Gründling 2013; Zheng et al. 2014; Rosenberg et al. 2015; Opoku-Temeng and Sintim 2016a; Commichau et al. 2019). This work is focused on functionality and regulation of the most prevailing DAC class CdaA from L. monocytogenes and the c-di-AMP receptor DarB/YkuL from B. subtilis. Here we report new crystal forms of CdaA from Listeria monocytogenes in its apo-state, post-catalytic state with bound c-di-AMP and two catalytic Co2+ ions as well as in complex with AMP. The comparison of the determined crystal structures revealed a tyrosine side chain (Tyr187) positioned in different orientations but locking the adenine ring after ATP binding. A mutation of Tyr187 to Ala unveiled its essential role during catalysis. This data enables the sug-gestion of a slightly different mechanism compared to its homolog DisA which is also known to synthesize c-di-AMP. Each monomer of CdaA needs to bind an ATP molecule, subsequent-ly a dimer is formed in order to perform catalysis. We suggest that this dimer formation is only transiently existing which is different to the stable associated DisA dimer comprising DAC domains positioned in a head to head assembly. This is an interesting and important assumption and helps to develop CdaA specific inhibitors that prevent dimerization. In order to identify potential compounds that reduce the CdaA activity we used crystallographic fragment screen-ing. This approach requires well diffracting (~ 2.0 Å) crystals of CdaA in its apo-state. Ob-tained apo CdaA crystals belong to the space group P212121 and diffracted up to 1.7 Å resolu-tion. Furthermore, the active site of apo CdaA in the crystal is exposed to solvent channels making it suitable for fragment screening. The results of our first crystallographic fragment screen unveiled three fragment binding sites in CdaA. Some fragments could be used to design inhibitors capable of preventing CdaA dimerisation and ATP binding due to π-π stacking inter-actions with the side chain of the Tyr187. Additional in vitro and in silico experiments are need-ed to come up with a potential CdaA inhibitor. Furthermore, we showed that the phosphoglucosamine mutase GlmM inhibits CdaA under hyperosmotic conditions. A phenylalanine 155 which is exposed on the surface of GlmM im-portantly contributes to the CdaA regulation. Further biochemical and structural experiments on how GlmM inhibits CdaA might also help to develop CdaA inhibitor which is based on the inhibitory mechanism of GlmM. Beside studying the regulation and inhibition of CdaA it is not less interesting to understand more abound the function of c-di-AMP. Up to the present time a plethora of c-di-AMP binding proteins have been discovered. Many of these proteins are involved in potassium or osmolyte uptake (Gundlach et al. 2019). c-di-AMP was identified to bind to conserved domains like the RCK_C domain or the CBS domain. It has been reported that B. subtilis carries 16 CBS domain containing proteins. c-di-AMP binding assays resulted in the identification of the CBS domain protein DarB/YkuL as a c-di-AMP receptor (Gundlach et al. 2019). Homologs have been identified in different Gram-positive bacteria like L. monocytogenes (CbpB) (Sureka et al. 2014). In order to get further insights into its function we crystallized DarB from B. subtilis in presence of c-di-AMP. Here we report new crystal forms of DarB in its apo-state and in complex with either c-di-AMP, 3’3’cGAMP or AMP. All determined crystals diffracted to a resolution of 1.5 - 1.8.4 Å and exhibit the same crystal packing (P212121). The crystal structures revealed two DarB monomers in the asymmetric unit forming a disk-like dimer. Surprisingly, the difference electron density map of each complex crystal suggested one of the described ligands in each of the supposed nucleotide binding site. This is different to the CBS domain containing protein OpuCA which is known to bind only one c-di-AMP in an extended conformation (Schuster et al. 2016). 2021-05-2