The story of MpsAB started from previous work in our lab in which we investigated whether S. aureus possesses a type 1 electrogenic NADH:quinone oxidoreductase (also known as Ndh1), a cation-translocating proton pump. However, such a Ndh1 is not present in staphylococci. Instead, we identified a hypothetical protein in S. aureus showing sequence similarity to the Escherichia coli specific NuoL protein, one of the core components of the proton pumping mechanism in Escherichia coli. The corresponding nuoL-like gene in S. aureus was named mpsA (membrane potential-generating system) because its deletion mutant was severely affected in membrane potential generation and exhibited a small-colony-like (SCV) phenotype. mpsA  is the first part of an operon comprising three genes: mpsA, mpsB and mpsC. However, mpsC is separated by a weak transcription terminator from the upstream mpsB; later it turned out that mpsC is dispensable for the function. 

As an extension of the previous study, we constructed mpsA,mpsB, and mpsABC deletion mutants and studied their phenotypes. All mutants hardly grew under ambient air conditions. Then, for the sake of curiosity, we also tested the impact of CO₂ and bicarbonate (HCO₃^-). To our surprise, the growth of the mutants could be restored to near wild-type levels under elevated CO₂ or bicarbonate concentrations. This indicates that the mutants have a defect in the acquisition of sufficient CO₂ or bicarbonate. Uptake studies with radiolabeled NaH¹⁴CO₃ (bicarbonate) revealed that MpsAB represents a bicarbonate transporter. Unlike CO₂, which can diffuse passively in and out of the cells, the uptake of bicarbonate affords a transport system. We assume that  sodium ions are co-transported with hydrogencarbonate-ion (HCO₃⁻), like in mammalians.  

This reminds us of the reports of CO₂-dependent S. aureus isolated from patients in the 1970s. Back then, nothing was known about the molecular basis of this CO₂ dependency. Fast forward to nearly 50 years later, we think that we might be able to shed some light on the cause of the CO₂ dependency. MpsAB-like transporters are widespread in bacteria. However, there are some species, like E. coli, which do not have such a transporter; instead they have a carbonic anhydrase (CA) enzyme that equilibrates HCO₃⁻+ H⁺ with CO₂ + H₂O. 

In this regard, we wondered whether the Staphylococcus specific mpsAB can complement an E. coli CA mutant. This mutant is unable to grow under atmospheric air but can grow in the presence of high CO₂, a phenotype similar to that of S. aureus mpsABC deletion mutant. Indeed, the E. coli CA mutant could be complemented by Staphylococcus specific mpsAB, and the growth of the S. aureus mpsAB(C) mutant could be complemented by the E. coli specific CA gene. Although MpsAB and CA have different mechanisms, they represent a CO₂/bicarbonate concentrating system that can functionally replace each other. Surprisingly, the MpsAB encoded genes are widespread and found in almost all bacterial families, suggesting that they have a central metabolic function. 

Next comes the million-dollar question: why do non-autotrophic and Rubisco-lacking bacteria like staphylococci and many other bacterial families need a MpsAB-like transporter or CA to enrich the cytoplasmic concentration of dissolved inorganic carbon sources? Autotrophic and CO₂-fixing bacteria, such as cyanobacteria, emerged about 3.8 billion years ago, during which the atmospheric CO₂ content was high, while that of O₂ was low. In a CO₂-rich environment, the CO₂-fixing enzymes in autotrophic bacteria, such as Ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) could work properly despite its low affinity to bicarbonate and limited discrimination between CO₂ and O₂. However, due to oxygenic photosynthesis in the subsequent 2.5 billion years, the atmosphere became O₂-rich and CO₂-poor. As a consequence, the cytoplasmic CO₂ levels became too low for proper CO₂ fixation and assimilation. Therefore, it is assumed that autotrophic bacteria that grow in atmospheric conditions appear to largely have CO₂ concentrating systems (CCSs).

To complete the story, we performed further experiments. We found that in S. aureus, the mpsAB genes contribute to fitness and pathogenicity/virulence in animal models. Some highly pathogenic species even have both systems, MpsAB and CA homologs. This observation suggests that a bicarbonate/dissolved inorganic carbon concentrating systems play an important role in fitness and pathogenicity and could explain the role of MpsAB beyond the known carbon dioxide-fixing bacteria.

Our full paper:

https://www.nature.com/articles/s41467-019-11547-5