Environmental arsenic poisoning affects roughly 200 million people worldwide. The toxicity and mobility of arsenic in the environment is significantly influenced by microbial redox reactions, with arsenite (As ) being more toxic than arsenate (As ). Microbial oxidation of As to As is known to be regulated by the AioXSR signal transduction system and viewed to function for detoxification or energy generation. Here, we show that As oxidation is ultimately regulated by the phosphate starvation response (PSR), requiring the sensor kinase PhoR for expression of the As oxidase structural genes aioBA. The PhoRB and AioSR signal transduction systems are capable of transphosphorylation cross-talk, closely integrating As oxidation with the PSR. Further, under PSR conditions, As significantly extends bacterial growth and accumulates in the lipid fraction to the apparent exclusion of phosphorus. This could spare phosphorus for nucleic acid synthesis or triphosphate metabolism wherein unstable arsenic esters are not tolerated, thereby enhancing cell survival potential. We conclude that As oxidation is logically part of the bacterial PSR, enabling the synthesis of the phosphate analog As to replace phosphorus in specific biomolecules or to synthesize other molecules capable of a similar function, although not for total replacement of cellular phosphate.

ArsR is a well-studied transcriptional repressor that regulates microbe-arsenic interactions. Most microorganisms have an arsR gene, but in cases where multiple copies exist, the respective roles or potential functional overlap have not been explored. We examined the repressors encoded by arsR1 and arsR2 (ars1 operon) and by arsR3 and arsR4 (ars2 operon) in Agrobacterium tumefaciens 5A. ArsR1 and ArsR4 are very similar in their primary sequences and diverge phylogenetically from ArsR2 and ArsR3, which are also quite similar to one another. Reporter constructs (lacZ) for arsR1, arsR2, and arsR4 were all inducible by As(III), but expression of arsR3 (monitored by reverse transcriptase PCR) was not influenced by As(III) and appeared to be linked transcriptionally to an upstream lysR-type gene. Experiments using a combination of deletion mutations and additional reporter assays illustrated that the encoded repressors (i) are not all autoregulatory as is typically known for ArsR proteins, (ii) exhibit variable control of each other's encoding genes, and (iii) exert variable control of other genes previously shown to be under the control of ArsR1. Furthermore, ArsR2, ArsR3, and ArsR4 appear to have an activator-like function for some genes otherwise repressed by ArsR1, which deviates from the well-studied repressor role of ArsR proteins. The differential regulatory activities suggest a complex regulatory network not previously observed in ArsR studies. The results indicate that fine-scale ArsR sequence deviations of the reiterated regulatory proteins apparently translate to different regulatory roles. IMPORTANCEGiven the significance of the ArsR repressor in regulating various aspects of microbe-arsenic interactions, it is important to assess potential regulatory overlap and/or interference when a microorganism carries multiple copies of arsR. This study explores this issue and shows that the four arsR genes in A. tumefaciens 5A, associated with two separate ars operons, encode proteins exhibiting various degrees of functional overlap with respect to autoregulation and cross-regulation, as well as control of other functional genes. In some cases, differences in regulatory activity are associated with only limited differences in protein primary structure. The experiments summarized herein also present evidence that ArsR proteins appear to have activator functions, representing novel regulatory activities for ArsR, previously known only to be a repressor. In reaction to arsenic in their environment, microorganisms orchestrate an organized response that may involve arsenite [As(III)] oxidation, arsenate [As(V)] reduction, or both. These redox reactions serve to detoxify or protect the organism or to generate energy, depending on the organism and the genes involved. Current models depict As(III) being taken up into the cell via aquaglyceroporins (e.g., reviewed in references 1, 2, and 3), where it then interacts with a DNA-binding repressor protein, ArsR, resulting in a conformational change in ArsR and causing it...