The regulon of Brucella abortus two-component system BvrR/BvrS reveals the coordination of metabolic pathways required for intracellular life

Brucella abortus is a facultative intracellular pathogen causing a severe zoonotic disease worldwide. The two-component regulatory system (TCS) BvrR/BvrS of B. abortus is conserved in members of the Alphaproteobacteria class. It is related to the expression of genes required for host interaction and intracellular survival. Here we report that bvrR and bvrS are part of an operon composed of 16 genes encoding functions related to nitrogen metabolism, DNA repair and recombination, cell cycle arrest, and stress response. Synteny of this genomic region within close Alphaproteobacteria members suggests a conserved role in coordinating the expression of carbon and nitrogen metabolic pathways. In addition, we performed a ChIP-Seq analysis after exposure of bacteria to conditions that mimic the intracellular environment. Genes encoding enzymes at metabolic crossroads of the pentose phosphate shunt, gluconeogenesis, cell envelope homeostasis, nucleotide synthesis, cell division, and virulence are BvrR/BvrS direct targets. A 14 bp DNA BvrR binding motif was found and investigated in selected gene targets such as virB1, bvrR, pckA, omp25, and tamA. Understanding gene expression regulation is essential to elucidate how Brucella orchestrates a physiological response leading to a furtive pathogenic strategy.

mediated control of metabolic systems (as well as other important virulence-related systems) in B. abortus. Overall, the authors have performed a robust analysis of BvrR binding to DNA elements in B. abortus, and while many of the conclusions are supported by the data, there are some concerns about the data and conclusions.
The specific concerns are: -The authors have previously demonstrated that the BvrR/BvrS system is a transcriptional activator of virB. Here, the authors suggest that BvrR binds to the virB promoter at approximately -12 from the transcriptional start site. Mechanistically, this is difficult to understand. How does binding to that site promote transcription?
In the same vein, in lines 558-568 the authors discuss the different potential binding sites for BvrR, but the data in To our knowledge, transcription could indeed be promoted from unusual sites and multiple binding sites could be needed for optimal binding. Some activators are known to bind to unusual regions and induce a promoter activity.
Since these questions are similar to question #2 from Reviewer #2, we added a line in the results section to introduce the idea that transcriptional regulation is a complex process of which we know very little, as follows:  "These observations suggest that regulation of genes important for virulence is complex, and that bacterial transcription factors do not behave as per the textbook operon model with interactions between the different BvrR-P binding sites probably according to BvrR-P concentration in a given moment, and in relation to additional transcription factors that might be involved in this process, as has been described for virB (see below)." This idea is described in detail in the discussion section as follows: Lines 677-698: "To our knowledge, transcription could indeed be promoted from unusual sites and multiple binding sites could be needed for optimal binding. Some activators are known to bind to unusual regions and induce promoter activity. as it has been described for other bacterial pathogens [100,101]. PhoP of B. subtilis, is a response regulator for phosphate starvation, which induces activation of pstS by binding to an upstream region (-40 to -132) and a coding region (+17 to +270) required for complete promoter activity. In addition, the coding region box had a low affinity for PhoP-P, suggesting a dynamic DNA-protein binding, in which the regulator is required to start transcription [102].
Global regulators are known to bind to a collection of sites, and the regulatory effect on each binding site would be dependent on the protein concentration at any given moment, its affinity, and in relation to additional transcription factors. Hence, they can be activators, repressors, have dual regulatory roles or no described regulatory function [103][104][105][106][107].
In Salmonella, the global response regulator OmpR activates the expression of SsrAB, a two-component system located on the pathogenicity island 2 (SPI-2).
The BvrR binding sites described in this work should be considered bona-fide putative gene regulation sites that deserve further investigation. Additionally, to our knowledge, very few Brucella promoter regions have been functionally characterized and hence, this essential information to properly unveil the mechanisms of gene regulation is missing. In this sense, confirmation of the role of each BvrR-P binding site, by itself, or in combination with other BvrR-P binding sites and/or additional regulatory mechanisms as well as gene promoter characterization certainly will shed some light to understand this complex phenomenon." - Fig. 1 -This is a very minor point, but it is difficult to determine which gels/lanes correspond to the map and primer sets. Moreover, the authors have included a control to demonstrate that transcription R/ We thank the reviewer for this comment. Fig.1 has been revised to clarify the reviewer´s concerns. We reconstructed this figure, the regions interrogated have stops that the 3' end of the message (i.e., the primer set represented by the black bar). Why is a similar control not included on the 5'-end of the message? been re-numbered to 16 regions, and the gels have been labelled accordingly. We think this time the info is clearer. Moreover, as suggested by reviewer 2, the genes in the map have been color-coded according to their annotated function. Therefore, the colors of the genes in Fig. 2, have also been modified to match Fig.1.
The figure legend has been modified to clarify that the intergenic region between the 5'-gene pckA and bvrR has not been tested because it was already known that pckA transcribed independently from bvrR, bvrS and the PTS genes (Dozot et al., 2010), unlike the case of the 3´-gene folC whose relationship with the transcription of bvrR, bvrS and the PTS genes was unknown.The last primer set interrogated was not a control per se, because we did not previously know if it was co-transcribed with bvrR, bvrS and the PTS genes or not. In total, 31 primer pairs were tested to span 16 overlapping regions of no more than 400bp. Only one representative RT-PCR product per region is shown.
All amplicons were sequenced to corroborate their identity. The results shown correspond to the log phase of the growth curve in TSB and are also representative of the co-transcription events observed at the stationary growth phase in the same medium." - Fig. 4 -Overall, the EMSAs are convincing, but there are some issues with some of them. For example, the binding to the virB

R/The experiments shown in panels A and B are different
and independent from each other (direct EMSA and promoter is highly variable between panels A and B. Why is this? It is understandable that differences exists between experiments, but in this case, the data are very difficult to interpret in terms of the competition controls when the control for those experiments looks nothing like the results in the panel A. competitive EMSA respectively). Therefore, they are not meant to be compared in between. Each gel, either on panel A or B, has its own negative control (probe without protein) to compared with. In panel B, each gel has its own positive control (probe with protein) to compare with.
In the case of the virB probe: "Competitive EMSA were performed as described (10). Briefly, the digoxigenin-labeled probes tested in the direct EMSA for tamA, omp25 and virB1, were incubated with BvrR-P (0.6 µM) and either an excess of the respective non-labeled probe as competitor, or separately, with an excess of non-labeled negative control probe (rplL or dhb) as competitor. Samples were then processed as described for direct EMSAs.
-Line 68 -Brucella replicates in a vacuole composed of (or associated with) the ER, and thus it may be incorrect to say that the bacteria replicate "within the ER." We have also modified the discussion accordingly, so the convey message is that the binding sites found should be describing the genes controlled directly by this TCS and under conditions that mimic the intracellular environment confronted by B. abortus while trafficking to its replicative niche.' This is not correct, the data show binding of Bvr-P to DNA, not regulation of gene expression. While there is some evidence that BvrR/S controls expression some genes, including virB and omp25, yo support this claim, it is essential that the authors provide data for the new set of genes that they claim to be controlled by BvrR/S. Lines 120-122:

R/The sentence has been modified according to the
"We expand our knowledge of the BvrR/BvrS regulon, describing genomic regions directly bound by BvrR under conditions that mimic the intracellular environment confronted by B. abortus while trafficking to its replicative niche." 2 The authors say that most BvrR-P binding sites are in regions upstream of the target genes. They also find binding sites in the virB4 and virB5 genes, several thousand bases into the operon. This is not at all discussed or commented on in the manuscript. How does this work? Are there internal promotors? This should be clarified. Table   4, most of the BvrR-P binding sites are within coding regions, as has been described in several cases (Bonocora RP, et al. 2015PLoS Genet. Lybecker M,et al. 2014. Transcription, Fitzgerald DM, et al 2014. The fact that there are still many general questions to answer related to transcriptional regulation in prokaryotes (Mejía-Almonte C, et al. 2020. Nat Rev Genet), and even more within the Brucella genus, precludes the possibility for establishing hypothesis that could be easily proven.

R/According to the info presented in Supplementary
Information regarding promoter structure, position of transcriptional binding sites and transcriptional regulators in Brucella is scarce. As mentioned in the response to the first question, the evidence presented in this manuscript suggests that BvrR/BvrS gene regulation is a complex process that at DNA level probably involves binding of BvrR-P molecules in more than one single site and in noncanonical E. coli regions including coding regions and/or downstream promoter regions (Liu et al. 1998. Mol Microbiol. 119-130, Shimada T et al. 2008 or even in a promoter region, located in a coding region (Fitzgerald DM, et al. 2018. Mol Microbiol 108:361-378.). The information presented in this manuscript is intended to present a first glance of such gene regulation process, in relation to BvrR/BvrS. Hence, the fact that binding sites found within operons, as is the case of the virB4 and virB5 opens many possibilities, from additional promoter regions to binding regions with no regulatory function at all. The answer to this question is certainly relevant, and out of the scope of this manuscript.
We have introduced three paragraphs in the discussion section to address the reviewer concern, in a general way, to keep the manuscript script (Lines 676-697 ), and described in the first answer to reviewer #2.
Additionally, we replaced the following sentence in the results section: "Furthermore, through ChIP-Seq, we detected five different binding sites for virB1 (Table 3 and Fig 5F), suggesting that additional TSS located within the coding region could be expressed under different conditions [8,78]." For the following sentence: (lines 493 to 496) "Furthermore, through ChIP-Seq, we detected five different binding sites related to virB (Table 3 bioRxiv 2022.01.18.476748;doi: https://doi.org/10.1101/2022 To keep the manuscript as simpler as possible, we  Table), an observation that has also been described in other pathogens' TCSs (53)." In total, 31 primer pairs were tested to span 16 overlapping regions of no more than 400bp. Only one representative RT-PCR product per region is shown.
All amplicons were sequenced to corroborate their identity. The results shown correspond to the log phase of the growth curve in TSB and are also representative of the co-transcription events observed at the stationary growth phase in the same medium." bp-long. In total, 31 primer pairs were tested to span 16 overlapping regions of no more than 400bp. Only one representative RT-PCR product per region is shown. All amplicons were sequenced to corroborate their identity. The results shown correspond to the log phase of the growth curve in TSB and are also representative of the co-transcription events observed at the stationary growth phase in the same medium." 6 The authors write 'These cotranscription events happened at log and stationary growth phases…' however they do not show data for different growth phases. What was the growth phase tested in Fig 1? R/ Co-transcription was observed at both log and stationary growth phases, only the results of the former were chosen to construct Fig. 1, to avoid an oversized and repetitive figure. As mentioned, the legend has been modified as described above to clarify the reviewer´s concern: Lines 303-310, and consequently removed the following sentence from the results section:, L285; "As demonstrated, cotranscription is independent of the growth stage". 7 The introduction is rather confused. Paragraph from L72-Are you talking about TCS in general or Brucella and BvrR/S? The refs suggest the latter the text the former.

L73 Define TCS
R/ According to the reviewer´s suggestion, we changed the references related to TCS in general, as follows: ....." The phosphorylated form of this protein shows an increased affinity for DNA binding sites, activating or repressing a particular set of genes, which constitute a direct regulon [7,8].
Is now replaced for (starting at line 82): ....." The phosphorylated form of this protein shows an increased affinity for DNA binding sites, activating or repressing a particular set of genes, which constitute a direct regulon [6].
According to the reviewer´s suggestion, the previous definition of TCSs in L73: "The transition from an extracellular to an intracellular milieu requires a highly coordinated gene expression. This is achieved through several regulatory mechanisms, including TCSs that allow