The addition of acetate to B. burgdorferi growth medium promotes the expression of RpoS and RpoS-dependent genes (e.g., OspC, DbpA, etc.; Xu et al., 2010; Van Laar et al., 2012; Richards et al., 2015). The acetate-triggered induction was presumed to be a result of increased levels of intracellular Ac-P, which, in turn, initiated the Rrp2–RpoN–RpoS regulator cascade (Xu et al., 2010). However, a recent study has shown that acetate can upregulate RpoN/RpoS-dependent proteins in the absence of Ac-P (Richards et al., 2015). The molecular mechanism and physiological significance of the acetate-triggered induction of virulence factors has remained unclear.

One consequence of exposing bacterial cells to membrane permeable acids, like acetate, is a drop in pH_i (acid stress) which can retard growth and, under extreme conditions, cause cell death (Foster, 1999; Wilks and Slonczewski, 2007; Slonczewski et al., 2009; Wilks et al., 2009). Since acetate-induced expression of OspC and RpoS was independent of Ac-P concentration, it seemed likely that this “acetate effect” might result from acetate directly decreasing pH_i. To test this possibility, we monitored the cytoplasmic pH of B. burgdorferi exposed to various membrane-permeable acids using the pH-responsive membrane-permeable dye pHrodo Green (effective pH range 4–7) (Miksa et al., 2009). B. burgdorferi B31 strain A3 spirochetes were grown in BSK-II media containing acetate, benzoate, lactate, or propionate, stained with pHrodo Green and visualized using fluorescence microscopy (Figure 1A). Spirochetes grown in BSK-II alone and exposed to pHrodo Green showed little fluorescence compared to background indicating a near neutral intracellular pH (Figure 1A, untreated). However, spirochetes grown in media containing acetate, benzoate, or propionate and exposed to the dye were demonstrably brighter, suggesting that these acids had acidified the cytosol. Cells exposed to lactate showed little change in fluorescence, indicating that this acid had a minor effect on intracellular pH (Figure 1A, lactate). We confirmed the fluorescence microscopy data using a more sensitive quantitative fluorescence microplate assay as described in “Materials and Methods” (Figure 1B). Cells exposed to acetate, benzoate, or propionate had a statistically significant increase in fluorescence compared to the untreated control (Figure 1B, untreated, acetate, benzoate, propionate) while lactate (Figure 1B, lactate) showed no change in fluorescence relative to the control. Although fluorescence microscopy showed a slight change in fluorescence for cells exposed to lactate, there was no change in the microplate assay, suggesting that subtle changes in fluorescence are difficult to quantify using this assay. Additionally, the pH of the buffered media did not change throughout the experiment. The result for lactate was predictable as B. burgdorferi is a homofermenter whose sole end-product of central metabolism is lactate. Thus, these bacteria have a highly efficient lactate secretion system that would actively excrete lactate as fast as it could enter the cell. These data suggest that membrane permeable acids can directly affect the pH_i of B. burgdorferi.

Having demonstrated that membrane-permeable weak acids decrease pH_i, we next sought to determine if the induction of RpoS and OspC in the presence of acetate (Xu et al., 2010; Van Laar et al., 2012; Richards et al., 2015) is due to an acid stress response in B. burgdorferi. We analyzed the effect of acetate, benzoate, lactate, and propionate on RpoS and RpoS-dependent virulence factors OspC and BBA66 during mid-logarithmic phase (1–5 × 10⁷ spirochetes/mL) to avoid growth-dependent (“stationary phase”) induction of RpoS (Yang et al., 2000). Cell lysates of spirochetes grown in BSK-II containing acetate, benzoate, lactate, or propionate were analyzed by immunoblot for production of RpoS, OspC, and BBA66. Benzoate and propionate had the greatest effect on RpoS expression, causing a marked increase in RpoS expression compared to the untreated control (Figure 1C). Acetate led to a moderate increase in RpoS levels while lactate did not increase RpoS expression. Benzoate and propionate, followed by acetate, resulted in the greatest increase in expression of the RpoS-dependent factors OspC and BBA66 (Figure 1C and Supplementary Figure S1). Lactate increased the expression of only OspC compared to the untreated control (Figure 1C and Supplementary Figure S1). Taken together, these data suggest that a decrease in pH_i promotes the expression of RpoS and RpoS-dependent virulence factors, with non-metabolizable acids having the greatest affect on both RpoS production and pH_i.

Synthesis of RpoS and RpoS-dependent virulence factors requires activation of the Rrp2–RpoN–RpoS signal transduction cascade (Yang et al., 2003; Fisher et al., 2005; Burtnick et al., 2007) and the transcriptional regulator BosR (Ouyang et al., 2009). To determine if acid stress promotes production of OspC and BBA66 through the Rrp2–RpoN–RpoS signaling pathway, B. burgdorferi strains B31-A3 (wild-type, WT), B31-A3ΔrpoN (ΔrpoN), B31-A3ΔrpoS (ΔrpoS), and B31-A3ΔbosR (ΔbosR) were grown in medium containing benzoate and we determined the level of OspC and BBA66 by immunoblot (Figure 1D). Unlike the WT strain, in the ΔrpoN, ΔrpoS, and ΔbosR mutant strains, production of OspC and BBA66 were abrogated (Figure 1D). Therefore, in response to membrane permeable acids, synthesis of RpoS and RpoS-dependent virulence factors likely proceed through the Rrp2–RpoN–RpoS pathway and this response requires BosR. These results suggest a generalized response to membrane-permeable acids where a decrease in pH_i activates the Rrp2–RpoN–RpoS/BosR signaling cascade resulting in virulence factor synthesis in B. burgdorferi.

Regulation of the Rrp2–RpoN–RpoS cascade in B. burgdorferi is complex, with multiple components regulated transcriptionally, post-transcriptionally, or post-translationally (Yang et al., 2003; Fisher et al., 2005; Burtnick et al., 2007; Lybecker and Samuels, 2007; Dulebohn et al., 2014; Richards et al., 2015). Having established that RpoS increases in response to acid stress, we next sought to determine if rpoS, or other key components of the Rrp2–RpoN–RpoS cascade, are transcriptionally regulated in response to membrane permeable acids. Using quantitative reverse transcriptase PCR (qRT-PCR) we analyzed the level of rrp2 and rpoN in response to membrane permeable acids and found they were relatively unchanged under acid stress conditions (Figure 2A). Conversely, rpoS transcript levels demonstrated a statistically significant increase under the same conditions, suggestive of transcriptional regulation of rpoS under acid stress and consistent with the immunoblot data demonstrating an increase in RpoS (Figures 1C, 2C). Additionally, B. burgdorferi strain B31-A3 harboring a plasmid containing the rpoS promoter fused to the lacZ_Bb gene (Hayes et al., 2010) exhibited a significant increase in β-galactosidase activity in the presence of benzoate or propionate (Figure 2B).

The level of the rpoS transcript expressed in B. burgdorferi is intricately linked to the BosR (Ouyang et al., 2009) and BosR is required for expression of OspC and BBA66 in response to acid stress (Figure 1D). We quantified the level of bosR transcript and found that expression remained relatively unchanged in the presence of propionate, but increased significantly in spirochetes exposed to benzoate (Figure 2A). Having observed a statistically significant increase in bosR transcript levels in response to benzoate we further evaluated BosR and the BosR-regulated protein NapA/BB0690 by immunoblot. In response to benzoate, both BosR and NapA levels increased compared to the untreated control (Figure 2C). Taken together, these data suggest that rpoS and bosR transcript levels increase in response to acid stress, leading to increased levels of these key regulators that may affect expression of both RpoS-dependent and BosR-regulated genes, suggestive of a possible overlap between an acid stress response and the oxidative stress response in B. burgdorferi.

Since RpoS and BosR levels increased under acid stress, we quantified RpoS-dependent and BosR-regulated transcripts in response to acid stress by qRT-PCR. These included the RpoS-dependent transcripts ospC, bba66, bbj24, bba34 and the BosR-regulated genes napA/bb690 and cdr/bb728 (Figure 2D). B. burgdorferi was grown in BSK-II supplemented with benzoate or propionate and transcript levels indicated that the RpoS-dependent genes ospC, bba66, bbj24, and bba34, were induced in response to propionate or benzoate compared to the untreated control (Figure 2D). Consistent with these data, β-galactosidase activity was significantly increased in spirochetes containing a plasmid with the lacZ_Bb gene fused to the ospC promoter in the presence of acetate, benzoate, or propionate (Supplementary Figure S3). Transcript levels of the BosR-regulated gene (Boylan et al., 2003; Hyde et al., 2006, 2009; Ouyang et al., 2009) napA/bb690 remained relatively unchanged, however, the level of cdr/bb728, encoding CoA reductase, was significantly increased in response to both propionate and benzoate (Figure 2D). These data indicate that RpoS-dependent and a BosR-dependent genes are induced in response to acid stress, also suggestive of an overlap between an acid stress response and the oxidative stress response in B. burgdorferi.

In response to acid stress, many bacteria activate an acid stress response (Marquis et al., 1987; Bearson et al., 1997; Foster, 1999, 2004; Slonczewski et al., 2009; Kanjee and Houry, 2013) to stabilize pH_i and retain the PMF (Bearson et al., 1997; Slonczewski et al., 2009; Wilks et al., 2009). B. burgdorferi is thought to lack “typical” bacterial stress responses (Caimano et al., 2004) and is missing many of the enzymes and transporters characteristically utilized to mitigate acid stress (Fraser et al., 1997). However, the genome does encode a vacuolar-type V_oV₁ ATPase, an arginine deiminase system (ArcA-ArcB-ArcD), and several proton/ion pumping transporters/symporters that could play a role in maintaining pH_i and the PMF in response to acid stress. To determine if B. burgdorferi responds to acid stress by increasing the transcription of genes encoding protective enzymes or transporters associated with an acid stress response we monitored select genes in response to spirochete growth under acid stress conditions. To generate acid stress conditions, we exposed spirochetes to propionate, as it activates the Rrp2–RpoN–RpoS cascade (Figure 1C) and unlike lactate and acetate, is not actively exported from Borrelia or utilized in undecaprenyl-P synthesis (Fraser et al., 1997). Unable to be metabolized like acetate or exported like lactate, propionate may accumulate in the cytoplasm, enhancing its ability to decrease pH_i. Additionally, the tick microbiome does not harbor bacteria predicted to generate benzoate, however, several species of bacteria found in the I. scapularis midgut could potentially produce propionate as a metabolic end-product, likely exposing B. burgdorferi to this acid during the infectious cycle (Narasimhan et al., 2014).

The Borrelia V_oV₁ ATPase (Fraser et al., 1997) is hypothesized to hydrolyze ATP to translocate protons out of the cytoplasm to maintain the PMF and drive proton-coupled transport (Radolf and Samuels, 2010). This is uncommon in bacteria but vacuolar V_oV₁ ATPases are often utilized by eukaryotic cells to acidify intracellular compartments through the ATP-driven translocation of protons (Nishi and Forgac, 2002; Mindell, 2012). To determine if the V_oV₁ ATPase was playing a role in the acid stress response we quantified the transcript levels of the V_oV₁ ATPase components (bb090, bb091, bb092, bb093, bb094, bb096). We found that all six components of the V_oV₁ ATPase had increased expression in response to propionate (Figure 3A). The most significant increases were in the V₁ ATP-binding components (bb093, bb096) and the V_o proton translocation (bb091) component of the V_oV₁ ATPase. We also quantified the expression of the arginine deiminase pathway genes (arcA/bb841-arcB/bb842-arcD/bb843), whose products can protect bacteria from acid stress through the ArcA-mediated conversion of arginine to citrulline and ammonia (Marquis et al., 1987; Ryan et al., 2009; Cusumano and Caparon, 2015). All of these genes had increased expression levels in response to propionate with statistically significant increases in both arcB/bb842 and arcD/bb843 transcript levels (Figure 3A).

In response to acid stress, bacteria must maintain both the pH_i and their PMF to survive (Bearson et al., 1997; Foster, 1999, 2004; Slonczewski et al., 2009). B. burgdorferi encodes several transporters/symporters that could play important roles in maintaining PMF homeostasis under acid stress conditions through the transport of protons/ions or monocarboxylic acids. We found that specific transport systems were induced in response to acid stress. Specifically, a Na/H⁺ transporter (bb638), glutamic acid transporter (bb729), potassium transporters (bb724, bb725), and a magnesium transporter (bb380) were all significantly induced in response to propionate (Figure 3B). However, the Na⁺, Ca⁺⁺, K⁺ transporter bb164 and the Na⁺/H⁺ transporter bb447 remained relatively unchanged (Figure 3B). Taken together, these results suggest that in response to acid stress, B. burgdorferi initiates an acid stress response that promotes the transcription of select genes encoding proteins that can potentially mitigate the detrimental effects of excess protons in the cytoplasm.

We next sought to define conditions that would inhibit acidification of the cytoplasm or inhibit permeable acid-mediated activation of the Rrp2–RpoN–RpoS pathway. To test if arginine could prevent the acidification of the B. burgdorferi cytoplasm, either through direct, or indirect buffering through the deamination of arginine and production of ammonia, we again monitored pH_i using the pH-sensitive fluorescent dye pHrodo Green. B. burgdorferi was grown in media containing benzoate, benzoate and arginine, or benzoate and alanine, stained with pHrodo Green and intracellular fluorescence was evaluated by fluorescence microscopy (Figure 4A). Benzoate has been used to study the acid stress response in E. coli due to its ability to decrease pH_i and induce a robust acid stress response (Wilks and Slonczewski, 2007; Slonczewski et al., 2009), and in B. burgdorferi benzoate had the greatest effect on pH_i and OspC and BBA66 expression. Therefore, we utilized benzoate to promote acid stress and to make any reduction in fluorescence or protein expression easily detectable in the presence of arginine. Spirochetes grown in BSK-II alone showed weak fluorescence, indicating a near neutral pH (Figure 4A, untreated) and spirochetes grown in BSK-II containing benzoate were markedly brighter, suggesting a decreased pH_i (Figure 4A, benzoate). However, spirochetes grown in both benzoate and arginine had fluorescence levels similar to the untreated controls, suggesting a near neutral pH_i (Figure 4A, benzoate + arginine). To test whether this buffering effect was specific to arginine, spirochetes were grown in BSK-II containing benzoate and alanine and fluorescence levels were found to be equivalent to the benzoate treated samples (Figure 4A, benzoate + alanine). To better characterize the effects of arginine on permeable acid-mediated changes to pH_i, we next measured pH_i changes using pHrodo Green in a quantitative microplate assay. Similar to what we had seen by fluorescence microscopy, spirochetes grown in the presence of benzoate had significantly higher relative fluorescence levels in the microplate assay (Figure 4B). Unlike spirochetes grown in benzoate, spirochetes grown in both benzoate and arginine had relative fluorescence levels similar to the untreated controls (Figure 4B). The relative fluorescence levels of spirochetes grown in media containing benzoate and alanine were equivalent to the benzoate treated samples (Figure 4B). These results indicate that supplementing with arginine can inhibit the acidification of the B. burgdorferi cytoplasm that occurs in response to benzoate.

If the Rrp2–RpoN–RpoS cascade is activated due to a decrease in pH_i, then the addition of arginine should inhibit induction of the Rrp2–RpoN–RpoS cascade. To test this hypothesis, we grew B. burgdorferi in the presence of both benzoate and arginine and monitored OspC and BBA66 levels by immunoblot. To ensure that any affect attributed to arginine was specific to arginine, WT B31-A3 spirochetes were cultured in BSK-II containing benzoate and alanine. Levels of OspC and BBA66 were again increased in response to benzoate, however, not in the presence of both benzoate and arginine (Figures 4C,D). The addition of alanine was unable to inhibit the permeable acid-mediated induction of OspC and BBA66 and arginine alone did not promote activation of the Rrp2–RpoN–RpoS cascade (data not shown). Taken together with our previous data, these results suggest that pH_i and virulence factor expression are coupled in B. burgdorferi. A decrease in pH_i can activate the Rrp2–RpoN–RpoS cascade and promote virulence factor expression.

We next focused on the molecular mechanism underlying the activation of the Rrp2–RpoN–RpoS cascade under acid stress. The response regulator Rrp2 is essential in B. burgdorferi (Groshong et al., 2012) and its cognate histidine kinase-2 (Hk2) has no known function. We tested whether Hk2 might play a role in activation of the Rrp2–RpoN–RpoS cascade under acid stress. Strains B31-A3 (WT) and B31-A3Δhk2(Δhk2) B. burgdorferi (Xu et al., 2010) were grown in BSK-II media supplemented with benzoate or benzoate and arginine and immunoblots of cell lysates demonstrated that OspC and BBA66 were increased in response to benzoate, but not in the presence of benzoate and arginine (Figure 4D). Additionally, we analyzed the level of NapA/BB690 levels to determine what affect benzoate and arginine would have, if any, on the oxidative stress response. Similar to what we observed for OspC and BBA66, NapA/BB690 levels increased in the presence of benzoate and levels were similar to the untreated sample in the benzoate and arginine treated samples (Figure 4D). These data indicate that Hk2 does not play a role in the permeable acid-mediated induction of RpoS-dependent factors (Figure 4D), or in promoting the expression of NapA/BB690. These results are consistent with what has been previously shown for the acetate-triggered induction of RpoS (Xu et al., 2010) and Hk2 does not play a role in the arginine-mediated inhibition of the Rrp2–RpoN–RpoS cascade or in expression of NapA/BB690.

During the infectious cycle, B. burgdorferi survives in complex environments that require changes in gene expression to ensure survival and transmission. In response to the weak organic acid propionate, spirochetes initiate an acid stress response, promoting the expression of genes whose products can mitigate the detrimental effects of excess protons in the cytoplasm. Moreover, several monocarboxylic acids trigger the acid stress response that promotes RpoN- and RpoS-dependent virulence factor expression and a portion of the oxidative stress response regulon. It has long been supposed that B. burgdorferi does not have typical stress response mechanisms (Caimano et al., 2004); however, these data suggest that the acid stress response and virulence factor expression are likely linked, possibly to promote transmission or to facilitate long-term survival in the tick.

Membrane-permeable acids can promote the expression of RpoS-dependent virulence factors and trigger an acid stress response in B. burgdorferi. However, whether membrane-permeable acids are present throughout the infectious cycle is unknown. Although the biochemical composition of some relevant biological tissues where B. burgdorferi resides during the infectious cycle is known, the composition of the I. scapularis midgut is poorly defined. The membrane-permeable acids acetate, butyrate, and propionate have been identified in the midgut of spirochete-containing arthropods (Odelson and Breznak, 1983; Gijzen and Barugahare, 1992) and bacteria identified in the tick midgut, including B. burgdorferi, produce membrane-permeable acids as metabolic end-products (Fraser et al., 1997; Narasimhan et al., 2014). Therefore, to assess the potential role for membrane-permeable acids on signaling during the infectious cycle, we determined the concentration of membrane-permeable acids in the midgut of I. scapularis and in relevant biological fluids of host animals.

We measured the levels of monoprotic acids (lactate, acetate, propionate, butyrate, and benzoate), using reverse phase high-pressure liquid chromatography (HPLC) and mass spectrometry (MS), in mouse sera, rabbit sera, and I. scapularis (B. burgdorferi-infected) midgut contents for the presence of these acids. Samples were isolated and processed as described previously and monocarboxylic acids were separated by HPLC and detected at 210 nm (Figures 5A–C). Peaks corresponding to acetate and lactate were identified based on their relative retention times compared to authentic standards (Figure 5D) and their identity was confirmed by MS. Mouse and rabbit sera had significant levels of lactate (∼12.1 ± 2.43 and ∼6.7 ± 0.17 mM, respectively) and acetate (∼7.8 ± 1.53 and ∼10.6 ± 0.3 mM, respectively). These two monoprotic acids were the predominant species detected in these sera samples. Midgut contents harvested from replete, adult I. scapularis ticks that had fed upon the same rabbit whose sera we measured had a concentration of lactate (∼9.3 ± 0.8 mM) similar to levels measured in the rabbit sera (Figure 5E), while the acetate concentration was significantly higher (38.0 ± 4.3 mM). We were unable to detect butyrate, propionate, or benzoate in any of the samples analyzed.

Initially, we had expected the levels of acetate and lactate to be increased in the tick midgut compared to levels initially detected in rabbit sera. That expectation was based upon the hypothesis that the removal of water from the blood meal, to facilitate saliva production in the feeding tick, would effectively increase the concentration of these acids in the midgut contents. While this might be the case for acetate, it appears that lactate, which is the end-product of B. burgdorferi metabolism is not increasing, suggesting that this acid might be consumed by the tick or by the resident midgut microbiota (Narasimhan et al., 2014).

Taken together, these data indicate that: (i) a weak organic acid could trigger a significant acid stress response in B. burgdorferi by affecting the intracellular pH of the cells, (ii) B. burgdorferi could mitigate these effects by generating or transporting neutralizing compounds (e.g., NH₄, arginine) or “pumping” protons out of the cells (V_oV₁ ATPase), (iii) the acid stress response and virulence factor expression (e.g., OspC) was regulated by RpoN and RpoS, and (iv) acetate was present in the tick midgut at concentrations high enough to promote virulence factor expression via RpoS-dependent gene regulation in vivo to possibly promote long-term survival in the tick midgut.

Borrelia burgdorferi strains B31-A3, B31-A3ΔrpoS, B31-A3ΔrpoN, B31-A3ΔbosR-L9, and B31-A3Δhk2 have been described previously (Elias et al., 2000; Fisher et al., 2005; Xu et al., 2010; Katona, 2015). Membrane-permeable acids were tested as previously described (Wilks and Slonczewski, 2007) with minor modifications. B. burgdorferi strains were grown at 34°C in BSK-II medium pH 7.0 under microaerobic conditions (3–5% O₂) to mid-logarithmic phase (5–7 × 10⁷/mL). Spirochetes were subsequently diluted to 1 × 10⁶/mL in BSK-II medium pH 7.0 or BSK-II medium pH 7.0 supplemented with 20 mM sodium acetate, sodium benzoate, sodium lactate, sodium propionate or in combination with 20 mM L-arginine or L-alanine (Sigma-Aldrich, St. Louis, MO, United States).

Spirochetes were grown to 1–5 × 10⁷/mL, harvested by centrifugation, washed twice with HEPES–NaCl buffer (20 mM HEPES, pH 8.0; 50 mM NaCl) (HN) and whole cell lysates were analyzed by SDS-PAGE and immunoblots as previously described (Dulebohn et al., 2014). Antibodies were used at the following concentrations: αOspC 1:1000 (Tilly et al., 2007), αBBA66 1:4000 (Clifton et al., 2006), αNapA/BB690 1:5000 (Boylan et al., 2003), αRpoS 1:500 (Dulebohn et al., 2014), αFlagellin 1:100 (Barbour et al., 1986), and αFur/BosR 1:1000 (Boylan et al., 2003). Immunoblots were incubated in 1× TBS-T with an HRP (horseradish peroxidase) conjugated recombinant protein A (Thermo Fisher Scientific, Grand Island, NY, United States) at 1:4000 for 1 h at RT with shaking for the detection of OspC, NapA/BB690, RpoS, and BosR. For detection of flagellin, an anti-mouse secondary antibody-HRP conjugated antibody was used at 1:10,000 (Thermo Fisher Scientific) and detection of BBA66 was performed with a secondary HRP conjugated αChicken antibody at 1:50,000 in 1× TBS-T as previously described (Clifton et al., 2006). Immunoblots were developed using SuperSignal West Pico chemiluminescent substrate kit (Thermo Fisher Scientific) and X-ray film (Lab Scientific Inc., Livingston, NJ, United States). Quantification of OspC and BBA66 (Supplementary Figure S1) was performed using Image J software (Schneider et al., 2012).

qRT-PCR was performed as previously described (Dulebohn et al., 2014) with minor modifications. Briefly, spirochetes were grown in BSK-II medium with or without benzoate or propionate, harvested at 1–5 × 10⁷/mL, and washed with HN buffer. Spirochetes were resuspended in 1 mL of TRI reagent (Sigma-Aldrich) and RNA harvested using a Direct-zol RNA mini prep kit (ZYMO Research, Irvine, CA, United States). RNA (5 μg) was treated for 1 h with Turbo DNase (Thermo Fisher Scientific) and 1 ng subsequently used for cDNA synthesis using High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific). cDNA was then used in qPCR reactions using the 5′ nuclease Prime Time Assay from IDT on a Viia 7 instrument (Thermo Fisher Scientific). Primers and probes are listed in Supplementary Table S1 and statistical analysis done with GraphPad PRISM software (GraphPad Software Inc., La Jolla, CA, United States).

Analysis of β-galactosidase activity was performed as previously described (Dulebohn et al., 2014). Briefly, B. burgdorferi strain B31-A3 containing the pBHospCp-lacZ_Bb (Hayes et al., 2010) or pBSV2G-rpoSp₉₂-lacZ_Bb shuttle vector, harboring the ospC or rpoS promoter, respectively, were grown to 1–5 × 10⁷ spirochetes/mL as stated above and β-galactosidase activity was assayed in three biological replicates as previously described (Dulebohn et al., 2014). The pBSV2G-rpoSp₉₂-lacZ_Bb shuttle vector was constructed by amplifying the rpoS promoter from B. burgdorferi B31-A3 gDNA using primers rpoSP92-F and rpoSP-R (Supplementary Table S1) and the 92 bp fragment (rpoSp₉₂) was ligated immediately upstream of the lacZ_Bb gene in pBHlacZ_Bb (Hayes et al., 2010). Relevant sequences were confirmed by sequencing.

Borrelia burgdorferi strain B31-A3 was grown to 1–2 × 10⁷ spirochetes per mL in BSK-II (pH 7.0) and cultures were split and treated with either sodium acetate, sodium lactate, sodium propionate, sodium benzoate or a combination of sodium benzoate and arginine, or sodium benzoate and alanine (20 mM final concentration of each). Cultures were then incubated at 35°C for 48 h under microaerobic conditions (3–5% O₂) and equivalent numbers of spirochetes (3 × 10⁹ spirochetes) were harvested by centrifugation at 3,000 rpm for 10 min and washed twice with Live Cell Imaging Solution (LCIS) (Thermo Fisher Scientific). pHrodo Green AM (Thermo Fisher Scientific) was then used according to the manufacturer’s instructions, samples incubated at 37°C for 30 min and then washed twice with LCIS. Fluorescence was quantified using a BioTek synergy spectrophotometer (BioTek, Winooski, VT, United States) using Excitation/Emission filters of 485 nm/560 nm, respectively. At least two independent biological replicates were performed for each condition tested and each biological replicate was analyzed in duplicate. Extended imaging led to increased background fluorescence so ReadyProbes^TM Backdrop^® Green Background Suppressor reagent (Thermo Fisher Scientific) was added to each sample and then imaged on a Nikon Eclipse E80 epifluorescence microscope (100× magnification, CY3 filter). All images were captured with identical exposure times (1 s).

Mouse infection and tick feeding were performed as previously described (Bontemps-Gallo et al., 2016). Briefly, laboratory-raised I. scapularis larvae (Oklahoma State University) were fed to repletion on B. burgdorferi B31-A3 infected mice, collected and allowed to molt to nymphs. Nymphs were then fed to repletion on naïve RML mice, collected and allowed to molt into adults. Adult ticks were then fed to repletion on New Zealand White rabbits. Midgut material was collected as previously described (Bontemps-Gallo et al., 2016) and diluted 1:2 in sterile water. Blood samples from New Zealand White rabbits or RML mice were clarified by centrifugation at 2000 × g for 10 min and the serum fraction recovered. All samples were cleared by centrifugation using a Microcon-10 (EMD Millipore, Billerica, MA, United States) at 10,000 × g until completion. The filtrate was then analyzed using an Agilent 1200 Series liquid chromatography system coupled to a 6120 Series Single Quad mass spectrometer equipped with an ESI source (Agilent Technologies, Santa Clara, CA, United States). Twenty microliters of sample or standard was injected and separated on an Aminex HPX-87 300 mm × 7.8 mm column (Bio-Rad, Hercules, CA, United States) following the manufacturers protocol. Absorbance data was collected from 200 to 750 nm to monitor COOH groups at 210 nm and aid in compound identification and peak purity analysis. Negative ion mass data was acquired as previously described (Ibanez and Bauer, 2014), with a 350°C nitrogen gas temperature and select ion monitoring for the [M-H]^- of acetic acid (59.00), lactic acid (89.1), propionic acid (73.1), and butyric acid (88.1). The mass spectrometer was tuned to the following internal mass reference ions (m/z): 112.99, 601.98, 1033.99, and 1633.95. Sample identifications were made by retention time relative to a standard and confirmed by MS. Quantification of lactate and acetate was performed with biological triplicates for mouse sera and tick midgut contents. One rabbit blood sample was taken and sample preparation and HPLC/MS analysis was performed in triplicate. Quantification of lactate and acetate was performed by monitoring UV absorbance at 210 nm and calculating the area under the peak relative to a standard curve for each compound.