Discussion
In summary, we have collected evidence that support the view that the residues required to sustain a promiscuous reaction in serum albumin could have evolved to hold a biological function. We have found that those residues involved in the reaction (Lys 199 and Arg 222 in HSA) are located in a large and hydrophobic cavity and are conserved during evolution (Figure 2) Interestingly, the pka shifts in these amino acids are also conserved to produce polarized “acid” and “basic” residues independently of being Lys or Arg. As we previously showed, the swapping in the Lys/Arg (HSA) and Arg/Lys (BSA) key residues (Table 1) in the natural sequences analyzed is statistically different from the occurrence by chance of the simple replacements of Lys and Arg. Evolutionary models as those used in this work (see Methods) consider residue replacements during evolution in a site-independent fashion. However, conservation of pka shifts in species also indicates physicochemical environmental constraints indicating a putative functional adaptation. Another piece of evidence is the finding of a common tunnel in the three structures analyzed. Tunnels allow the transit of ligands between cavities and protein surfaces. As shown in Figure 4, three important residues lie at the bottom of the tunnel (Lys 199, Arg 218, and Arg 222 in HSA numbering). Among these, Lys 199 and Arg 222 were previously suggested to be involved in promiscuous reactions41,48,49. Also, Arg 218 is 100% conserved in serum albumins and its replacement has been related to the occurrence of human diseases. Replacement of Arg 218 to His or Pro produces abnormal albumin with increased affinity for serum thyroxine found in an autosomal dominant condition called Familial dysalbuminemic hyperthyroxinemia50,51. This condition is caused by an abnormal albumin molecule with an increased affinity for the hormone thyroxine. Supporting the putative role of this tunnel is the fact that two flanking residues (Arg 222 and His 440 in HSA see Figure 4) are evolving under positive selection.
Our results offer a biological interpretation of the observed promiscuous catalytic activity in serum albumins. In addition, a new perspective is offered in the classification of albumins as possible enzymes, in light of the observation of the selective pressure in the key positions for catalysis, typically observed during functional divergence52. From this point of view, the catalytic properties of the serum albumins documented in the present work could be involved in an unknown biological process rather than a promiscuous behavior. These results not only allow us to better characterize the family but also open up interesting questions about the origin of promiscuous behavior and the evolution of protein function.