Azlocillin can be the potential drug candidate against drug-tolerant Borrelia burgdorferi sensu stricto JLB31

Lyme disease is one of most common vector-borne diseases, reporting more than 300,000 cases annually in the United States. Treating Lyme disease during its initial stages with traditional tetracycline antibiotics is effective. However, 10–20% of patients treated with antibiotic therapy still shows prolonged symptoms of fatigue, musculoskeletal pain, and perceived cognitive impairment. When these symptoms persists for more than 6 months to years after completing conventional antibiotics treatment are called post-treatment Lyme disease syndrome (PTLDS). Though the exact reason for the prolongation of post treatment symptoms are not known, the growing evidence from recent studies suggests it might be due to the existence of drug-tolerant persisters. In order to identify effective drug molecules that kill drug-tolerant borrelia we have tested two antibiotics, azlocillin and cefotaxime that were identified by us earlier. The in vitro efficacy studies of azlocillin and cefotaxime on drug-tolerant persisters were done by semisolid plating method. The results obtained were compared with one of the currently prescribed antibiotic doxycycline. We found that azlocillin completely kills late log phase and 7–10 days old stationary phase B. burgdorferi. Our results also demonstrate that azlocillin and cefotaxime can effectively kill in vitro doxycycline-tolerant B. burgdorferi. Moreover, the combination drug treatment of azlocillin and cefotaxime effectively killed doxycycline-tolerant B. burgdorferi. Furthermore, when tested in vivo, azlocillin has shown good efficacy against B. burgdorferi in mice model. These seminal findings strongly suggests that azlocillin can be effective in treating B. burgdorferi sensu stricto JLB31 infection and furthermore in depth research is necessary to evaluate its potential use for Lyme disease therapy.


Results eradication of B. burgdorferi persisters by Azlocillin and cefotaxime.
In the present study, we first assessed the potency of azlocillin and cefotaxime against B. burgdorferi in dose dependent manner in both log and also stationary phase cultures of B. burgdorferi along with standard Lyme antibiotic (doxycycline). We used mitomycin C as a positive control and determined viability by colony forming unit (CFU) counts throughout the entire study 19 . Our results showed that the both tested antibiotics, cefotaxime at high concentration 40 μg/ ml and azlocillin at very low concentration 2.5 μg/ml could able to completely (100%) kill log phase culture of B. burgdorferi respectively (Fig. 1A). Similarly, azlocillin at 20 μg/ml concentration also eliminated stationary phase B. burgdorferi persisters completely. However, cefotaxime at highest concentration of 80 μg/ml could able to kill (80%) of the stationary phase B. burgdorferi persisters (Fig. 1B). More importantly, cefotaxime at increased concentrations from 20 to 80 μg/ml did not vary much in killing of a small persister fraction of surviving cells. The doxycycline, a bacteriostatic couldn't able to kill both the log phase and stationary phase B. burgdorferi cultures at higher concentrations of 80 μg/ml. More than 1000 stationary phase cells were survived at doxycycline concentration of 80 μg/ml. The mitomycin C at 1.25 µg/ml concentration killed B. burgdorferi both in log and stationary phase persisters as reported earlier 19 . B. burgdorferi persisters induced by doxycycline are more tolerant to drugs. Doxycycline is most commonly used drug to treat Lyme disease among the antibiotics prescribed for Lyme disease. Many researchers have shown that doxycycline doesn't kill B. burgdorferi completely and some population of drug-tolerant B. burgdorferi still exists 19,25,28,29 . The human maximum drug concentration (Cmax) range of doxycycline is between 2.6-5.9 μg/ml and has a half-life of 14 to 24 h 19,30 . At this doxycycline Cmax concentration range (2.5-5.0 μg/ml), 1000-10000 cells/ml of log phase Borrelia and 10000-200000 cells/ml of stationary phase Borrelia still exists from the initial inoculam of 10 6 cells/ml ( Fig. 2A,B). The fraction of B. burgdorferi persisters survived against doxycycline is significantly high. Further, to find drugs that can effectively kill doxycycline-tolerant B. burgdorferi, we have tested both azlocillin and cefotaxime on doxycycline-tolerant B. burgdorferi persisters. The doxycycline-tolerant B. burgdorferi that survived at both log phase and stationary phase were treated with 20 and 40 μg/ml of azlocillin and also resuspended again with doxycycline for 7 days and plated on BSK-II agarose medium for CFU. In log phase cultures, the doxycycline-tolerant B. burgdorferi that survived at 2.5 and 5 μg/ml of doxycycline were reduced to <50 cells/ml by azlocillin. Azlocillin completely eliminated all the doxycycline-tolerant B. burgdorferi survived at 10 μg/ml of doxycycline treatment ( Fig. 2A). As expected and also shown by other groups, B. burgdorferi in stationary phase were more tolerant to doxycycline 11,19,29,31 . Azlocillin significantly reduced stationary phase drug-tolerant peristers survived at 2.5 and 5 μg/ml of doxycycline treatment to <300 cells. Stationary phase B. burgdorferi that persisted at 10 μg/ml of doxycycline were effectively killed to <200 by azlocillin (Fig. 2B). Azlocillin has effectively eliminated 99% of doxycycline-tolerant B. burgdorferi persisters both in log and stationary phase cultures.
The drug-tolerant B. burgdorferi persisters which persisted after treating with doxycycline were also tested with 40 and 80 μg/ml cefotaxime. Cefotaxime effectively kills the log phase doxycycline-tolerant B. burgdorferi to <10 cells (Fig. 3A). Furthermore, cefotaxime also killed stationary phase B. burgdorferi to <2500 cells, which are survived at 2.5 and 5 μg/ml doxycycline. Doxycycline-tolerant B. burgdorferi that persisted at 10 μg/ml doxycycline were killed effectively to <200 cells when treated with cefotaxime. When the doxycycline-tolerant B. burgdorferi that persisted were resuspended again with doxycycline, no drastic reduction of B. burgdorferi was observed. Though it is statistically significant for B. burgdorferi treated at 2.5 μg/ml, the decrease in no of B. burgdorferi cell growth is less with both azlocillin and cefotaxime treatment. At 5 and 10 μg/ml of doxycycline resuspension treatment no major reduction of B. burgdorferi growth was observed. At both log and stationary phase, significant growth reduction was also observed with azlocillin and cefotaxime treatment when compared to doxycycline resuspension (Figs. 2 and 3).
We also tested whether the colonies survived after doxycycline-tolerant B. burgdorferi persisters treated with azlocillin repeatedly had acquired any resistance towards drugs. So, we picked and regrown doxycycline-tolerant B. burgdorferi persisters that were treated with 10 μg/ml of azlocillin for 7 days in BSK-II medium. Then the regrown B. burgdorferi were tested by treating them again with 10, 20 μg/ml of azlocillin. These results observed were same as earlier and only less than 10 cells were survived (Fig. 4). From these results we have found that drug-tolerant B. burgdorferi peristers that survived, were treated with azlocillin did not acquire any antibiotic resistance mechanism. Our observation showed B. burgdorferi persisted might be typical stochastic persister cells which was shown earlier by other researchers 19,25,32 . Azlocillin and Cefotaxime combination increases efficacy. It is known that some antibiotics act synergistically or more effective when used in combinations 33 . In our further studies we have tested combinations of azlocillin and cefotaxime to identify whether they can increase the efficiency of killing drug-tolerant persisters cells. The azlocillin and cefotaxime combinations could able to kill drug-tolerant B. burgdorferi persisters formed at 5 and 10 μg/ml of log phase cultures. The azlocillin (20 μg and 40 μg) and cefotaxime (80 μg) combinations reduced 2.5 μg/ml of doxycycline treated persisters(log phase) to less than 10 cells/ml (Fig. 5A). Furthermore, combination of 40 μg/ml azlocillin and 80 μg/ml cefotaxime effectively could able to kill stationary phase doxycycline (2.5, 5 and 10 μg/ml) treated persister cultures. The drug-tolerant persisters survived in this combination were <10 cells/ml (Fig. 5B). Other drug combinations of azlocillin and cefotaxime (40 μg/ml azlocillin and 40 μg/ ml cefotaxime; 20 μg/ml azlocillin and 80 μg/ml cefotaxime) shown in Fig. 5B reduced the persisters generated by doxycycline (at concentrations 2.5, 5 and 10 μg/ml) to 10 to 100 cells/ml. Overall, the drug combinations of 40 μg/ ml azlocillin and 80 μg/ml cefotaxime is much more effective in killing persisters than using azlocillin alone. The azlocillin and cefotaxime combination significantly kills doxycycline-tolerant B. burgdorferi than doxycycline resuspension at both log and stationary phases. The exponential culture of B. burgdorferi culture was exposed to antibiotics for 5 days, and surviving cells were determined by CFU count. The culture was treated with cefotaxime, azlocillin, doxycycline and mitomycin C (n = 6). Error bars represent standard errors.
www.nature.com/scientificreports www.nature.com/scientificreports/ time kill studies of B. burgdorferi by Azlocillin. We have observed that azlocillin kills 100% of normal B. burgdorferi cells and more than 99% of drug-tolerant B. burgdorferi persisters at 20 and 40 μg/ml azlocillin concentrations. To determine the rate of antimicrobial activity of azlocillin sodium with time, 10 6 /ml B. burgdorferi (JLB31 strain) was exposed to concentrations of 20 and 40 μg/ml azlocillin drug. In 2 hrs the initial B. burgdorferi inoculum decreased more than 1-log10-unit at both the concentrations of azlocillin sodium (Fig. 6). By 24hrs B. burgdorferi load is decreased to 10000 cells/ml and by 48 hrs bacteria were reduced to 100 cells/ml (99.9%). By 96 hrs, azlocillin has killed all B. burgdorferi persisters at both concentrations in stationary phase cultures. The doxycycline Cmax concentration range of 5 μg/ml and 10 μg/ml (2x of Cmax) was used for effectivity comparision. In control, the B. burgdorferi growth increased to 10 8 cells/ml.
In silico analysis of azlocillin binding to pBps and clp proteases. Since the 3D structure for the pencillin-binding protein of Borrelia burgdorferi (PBP-3Bb) has not been resolved, its 3D structure was predicted using homology modeling. Penicillin-binding protein 1 A (PBP-1A) of Pseudomonas aeruginosa (PDB ID: 5DF7), for which 3D structure is available (PDB ID: 4OON), was found to have 39% sequence similarity with PBP-3Bb 34 . Therefore, 4OON was used as the template to build the 3D structure of PBP-3Bb; in the modeled structure of PBP-3Bb, 87.1% of the amino acids were observed to be in the most favored region while 11.4% were in addition allowed region of the Ramachandran plot (supplementary Figure 1). In addition, the modeled structure showed RMSD value 0.235 A° with that of the template structure despite the moderate sequence similarity observed between these two proteins. These results indicate the reliability of the modeled structure for further After 5 days the cultures were taken, washed, diluted, and plated on semisolid BSK-II medium for CFU counts (n = 3). Statistically significant difference between groups by two-way ANOVA followed by Tukey's multiple comparisons test are indicated by ****p < 0.0001, ***p < 0.0003, and **p < 0.0088. Hash symbol represent eradication to the limit of detection. In the figure legend, Doxy persisters (doxycycline persisters), Doxy-tolerant Bb (doxycycline-tolerant B. burgdorferi), doxy-resuspension (B. burgdorferi treated again with doxycycline), Azlo 20 (20 μg/ml of azlocillin) and Azlo 40 (20 μg/ml of azlocillin).
As clpP was reported to be upregulated by doxycycline-tolerant B. burgdorferi, we speculated that azlocillin mode of action against B. burgdorferi through inhibition of any of the Clp protease subunits, because we found azlocillin is active against doxycycline-induced persistors of B. burgdorferi 32 . Based on these reports, we have performed molecular modeling and molecular docking on five different protease proteolytic subuntis, ClpP, ClpP1, ClpP2, ClpA and ClpX of B. burgdorferi. Azlocillin was found to bind with these five protease subunits with similar binding affinity as that of with PBP-3Bb (ClpX: −9.9, ClpP: −9, ClpP2: −9, ClpP1: −8.9, ClpA: −8.8 (kcal/ mol). As the 3D structure for clpX of B. burgdorferi has not been resolved, its 3D structure was predicted using homology modeling based on clpX of E. coli (PDB ID 3HTE) (supplementary Figure 2). Five amino acids from clpX namely, LYS123, GLY122, THR124, LYS128 and ARG301 were found to interact with azlocillin by hydrogen bonding while four more amino acids namely, GLY120, ARG364, GLU109, GLN330 made with van der Waals interactions with azlocillin ( Fig. 7c,d). stationary phase culture. The B. burgdorferi were treated different concentrations of doxycycline (2.5, 5 and 10 μg/ml) for 3 days, pelleted, washed and then treated again with doxycycline and cefotaxime (40 and 80 μg/ ml). After 5 days the cultures were taken, washed, diluted, and plated on semisolid BSK-II medium for CFU counts (n = 3). Statistically significant difference between groups was evaluated by two-way ANOVA followed by Tukey's multiple comparisons test are indicated by ****p < 0.0001, ***p = 0.0001, **p < 0.0067 and *p = 0.0163. Hash symbol represent eradication to the limit of detection. In the figure legend, Doxy persisters (doxycycline persisters), Doxy-tolerant Bb (doxycycline-tolerant B. burgdorferi), doxy-resuspension (B. burgdorferi treated again with doxycycline), Cefotaxime 40 (40 μg/ml of cefotaxime) and Cefotaxime 80 (80 μg/ ml of cefotaxime). (2020) 10:3798 | https://doi.org/10.1038/s41598-020-59600-4 www.nature.com/scientificreports www.nature.com/scientificreports/ In vivo testing of drugs in C3H/HeN mice. From our in vitro studies we have found that azlocillin and cefotaxime have killed B. burgdorferi effectively at low MIC and MBC concentration. Based on these results, the efficacy of compounds azlocillin and cefotaxime were tested in 5-6 week old female C3H/HeN mice in comparision to one of the currently prescribed drug doxycycline. In order to identify more effective drugs, we have infected C3H mice with infective dose of B. burgdorferi (2 × 10 5 ) than the normal efficacy studies 31,35 . The higher concentration of B. burgdorferi dosage increases the infectivety rate. Mice were treated with the drugs azlocillin (50 mg/kg), doxycycline (50 mg/kg) and cefotaxime (30 mg/kg) once a day for five days post 7, 14 and 21 days of B. burgdorferi infection. Both azlocillin and doxycycline cleared B. burgdorferi completely in the 7 day infected mice. No B. burgdorferi growth was observed from the ears cultured in all the mice treated with azlocillin and doxycline groups. In addition to it, B. burgdorferi specific DNA by qPCR was not observed in azlocillin and doxycline treated mice. In the cefotaxime treated mice, 1 of the 4 mice showed positive for ear cultures. The B. burgdorferi DNA was detected by qPCR in 2 of 4 mice from ear and also in urinary bladder of all the 4 mice of cefotaxime group (Table 1). In saline treated (control) group, all the mice showed positive for ear cultures and also for qPCR analysis of ear and urinary bladder. As expected, all the mice of the naïve group shown negative for ear cultures and for qPCR analysis of ear and urinary bladder.
Both azlocillin and doxycycline treatment has shown similar efficacy in the mice infected for 7 days. To identify a highly effective molecule that can clear B. burgdorferi infection we have extended infection time periods to 14 and 21 days. The B. burgdorferi infection reaches high titre between 14 and 21 days of postinfection and spreads to different organs of body like heart, spleen, kidney and joints etc 36 . For all the mice treated with drugs after 14 and 21 days of B. burgdorferi infection, ear and urinary bladders were collected for whole-DNA extraction and quantitative PCR (qPCR) analysis. In all the 4 mice infected for 14 days and treated with cefotaxime, high amount of B. burgdorferi DNA was detected in both ear and urinary bladder. After 14 days of infection and doxycycline treatment, the B. burgdorferi DNA was detected in 3 of 7 mice in ear tissue and 1 of 7 mice in urinary bladder. In one of the doxycycline treated mice B. burgdorferi DNA was present in both ear (high amount) and urinary bladder. However, for azlocillin treatment, only 2 of 8 mice has shown B. burgdorferi DNA in ear. No B. burgdorferi DNA was present in urinary bladder of all the 8 mice treated with azlocillin ( Table 2). In all the mice treated with azlocillin after 21 days of infection, no B. burgdorferi DNA was detected in both ear and urinary bladder. But in the doxycycline group, 1 of 3 mice was detected positive for B. burgdorferi DNA in ear tissue and no DNA was found in urinary bladders ( Table 3)  burgdorferi persister formation is not heritable. The colonies that were recoverd after doxycyclinetolerant B. burgdorferi (5 μg/ml) treated with 10 μg/ml azlocillin were used for the studies. The recovered persister colonies grown in BSK-II medium were treated with azlocillin (10 and 20 μg/ml) and plated on semisolid BSK-II medium. The colonies recovered after treating with 10 μg/ml azlocillin were cultured with fresh BSK-II medium for 7 days. Then the B. burgdorferi were treated again with azlocillin (10 and 20 μg/ml). After 5 days of treatment the cultures were taken, washed, and plated on semisolid BSK-II medium for CFU counts (n = 3). Statistically significant difference between groups was evaluated by two-way ANOVA followed by Tukey's multiple comparisons test are indicated by ****p < 0.0001 and **p = 0.0085. Hash symbol represent eradication to the limit of detection. (2020) 10:3798 | https://doi.org/10.1038/s41598-020-59600-4 www.nature.com/scientificreports www.nature.com/scientificreports/

Discussion
Like many bacteria B. burgdorferi also forms persisters due to external stimuli like depriving of nutrients, pH and antibiotics etc 29 . The mechanism by which B. burgdorferi form persisters is unknown sofar. Many bacteria uses different type of mechanisms for the formation of persisters. The E. coli and S. Typhimurium bacteria uses identified redundant toxin-antitoxin (TA) modules for the persister formation 29,37 . Toxins helps in persister formation by decreasing the energy level of cells or by inhibiting protein synthesis 29,38,39 . The B. burgdorferi persisters generated by antibiotics are biphasic, with a small subpopulation of surviving cells which are not genetically modified 19 .
To identify drugs that can completely eliminate B. burgdorferi we have tested drugs azlocillin and cefotaxime on log and stationary phase cultures. The main criteria for selection of these drugs are based on their safety and good efficacy with low MIC and MBC values. The cefotaxime has killed log phase culture at 40 μg/ml and unable to kill all the stationary phase B. burgdorferi persisters even at 80 μg/ml. The azlocillin has eradicated B. burgdorferi log phase culture at 2.5 μg/ml and stationary phase culture at 20 μg/ml. So far, azlocillin is the FDA approved drug shown that killed both B. burgdorferi log phase and stationary phase cultures completely. Azlocillin is an acylated form of ampicillin which is similar to the antibiotics mezlocillin and piperacillin. It is a broad range β-lactam antibiotic kills many pathogens which also show high activity against Pseudomonas aeruginosa 40,41 . It is well known that cell wall-acting drugs do not kill nongrowing bacteria. One of the reason azlocillin killing B. burgdorferi persisters is that the stationary phase culture represents a steady state of growing and dying cells. It is shown recently that the stationary phase B. burgdorferi synthesize peptidoglycan at the poles and also in the middle of the cell 19,31 . Due to the production of peptidoglycan in stationary phase B. burgdorferi, other cell wall acting drugs like ceftriaxone and vancomycin were killing stationary phase B. burgdorferi very efficiently 31 . As azlocillin targets cell wall synthesis, this might be one of the reason that it is killing drug-tolerant persisters. It is also shown that β-lactams killing nonreplicative M. tuberculosis bacteria effectively when used in combination with meropenem 42 .
The serum levels of azlocillin and cefotaxime in the blood are higher than the MBC values to kill the B. burgdorferi. The Cmax concentration of cefotaxime is 125 µg/ml and azlocillin is 236.55 + /−12.9 µg/ml, which is 10 times more than the MBC value 43,44 . The Cmax concentration of doxycycline is at the range of 2.6-5.9 μg/ml and  www.nature.com/scientificreports www.nature.com/scientificreports/ has a half-life of 14 to 24 h 30 . It is shown that at this Cmax concentration of doxycycline all the B. burgdorferi cannot be cleared, still some subpopulation persists and exists 13,19,25,28,29 . Researchers has shown when the stationary phase B. burgdorferi were treated with Cmax concentration (2.6-5.9 μg/ml) of doxycycline nearly 10 4 -10 6 cells still survived 19,29 . At this Cmax concentration of doxycycline, we also found 10 3 -10 4 of log phase Borrelia and 10 4 -10 5 cells/ml of stationary phase Borrelia still exists. The fraction of B. burgdorferi subpopulation survived with doxycycline treatment is significantly high. The resuspension of doxycycline-tolerant persisters again with doxycycline did not show much effect as like previous studies 29 . This is because the Cmax concentrations of doxycycline used to kill B. burgdorferi is not sufficient 11,19 . In our study, we have shown for the first time that azlocillin, an FDA approved drug is killing nearly 99.9% of doxycycline-tolerant B. burgdorferi persisters that formed at Cmax concentrations, both in log phase and stationary phase cultures. In comparision to azlocillin, cefotaxime did not eliminate doxycycline-tolerant B. burgdorferi persisters as efficiently and nearly 10 3 cells/ml B. burgdorferi persisters were survived when treated with 40 μg/ml and 80 μg/ml of cefotaxime. So, the results showed azlocillin effectively kills the drug-tolerant B. burgdorferi persisters. Though both antibiotics are β-lactams, azlocillin might use different mechanism in killing persisters. When used drug combination of azlocillin and cefotaxime, it killed B. burgdorferi persisters more effectively than using alone. The drug combination of 40 μg/ml azlocillin and 80 μg/ ml cefotaxime kills doxycycline persisters more effectively (less than 10 cells) compared to other combinations showed in the Fig. 5. Further, we also tested whether the doxycycline-tolerant persisters population that were treated with azlocillin have acquired resistance to drugs or persisters formed stochastically tolerating antibiotic stimuli. As our studies were conducted in in vitro, several factors like influence of host, immune response and tissue penetration etc were not taken into account. From our observation, we found persisters survived stochastically and didn't acquire any antibiotic resistance, the same phenomenon was also observed by other researchers in B. burgdorferi 19,29 . Some researchers also observed resistant B. burgdorferi to erythromycin drugs but we did not observe any resistant mutants from our results shown in Fig. 4 45 . The very peculiar feature of B. burgdorferi is that it doesn't show any resistance to antibiotics. The severeal attempts made by researchers to raise mutants to amoxillin and ceftriaxone were not successful 19 . The time kill studies showed that by 48 hrs 99.9% of B. burgdorferi were killed when treated with 20 and 40 μg/ml azlocillin. By 96 hours, all the B. burgdorferi from stationary phase cultures were completely killed by azlocillin. We report here for the first time that a FDA approved drugs azlocillin  www.nature.com/scientificreports www.nature.com/scientificreports/ and cefotaxime more effectively killing both log and stationary phase B. burgdorferi cultures at clinically relevant concentrations in in vitro. Among these two drugs, azlocillin proved to be the most effective in eradicating the B. burgdorferi even at very low concentrations. www.nature.com/scientificreports www.nature.com/scientificreports/ It is shown that during doxycycline-persister formation many genes are differentially expressed compared to normal growth cycle of B. burgdorferi. A total of 35 genes were up-regulated two-fold and 33 genes were down-regulated more than two-fold. The up-regulated genes consisted of different genes including transporter genes, bacterial envelope protein coding genes, genes encoding chemotaxis proteins and the Clp protease gene 32 . It is known that β-lactams binds to ClpP protease and inhibits degradation of misfolded proteins leading to cells death 46 . Inactivation of the ClpXP protease in Staphylococcus aureus even leads to β-Lactam resistance 47 . As azlocillin and cefotaxime are killing doxycycline-tolerant B. burgdorferi persisters, we speculate these drugs may bind to B. burgdorferi ClpP protease. Though there is no evidence that ClpP protease binds to azlocillin, we hyphothesized based on available published data 32 . When we docked the azlocillin with five different Clp subunits, azlocillin was found to bind with maximum affinity with ClpX (−9.9 kcal/mol). The ClpX in bacteria has been known as important for protein unfolding, degradation and regulation of protein quality and turnover through controlled proteolysis 48,49 . The stronger binding of azlocillin with B. burgdorferi ClpX suggest that ClpX might be a potential target for azlocillin in addition to PBP-Bb. However, this observation needs to be experimentally verified in B. burgdorferi.
At the same time, the recent findings showed biosynthesis of bacterial cell wall takes place in non-growing B. burgdorferi stationary phase cultures by continious synthesis of peptidoglycan 31 . Penicillin-binding proteins (PBPs) play a major role in transglycosylation and transglycosylation steps for peptidoglycan synthesis 50 . The β-lactam drugs bind to PBPs and prevent either transglycosylation or transglycosylation steps which leads to disruption of bacterial of cell wall synthesis 51 . The crystallography structural studies shows azlocillin binds strongly to penicillin-binding protein 3 (PBP 3) of Pseudomonas aeruginosa 34 . Considering these studies, we have performed in silico analysis of azlocillin interacting with pencillin-binding protein of Borrelia burgdorferi (PBP-3Bb). We have found that azlocillin binds strongly with PBP-3Bb with binding affinity of −8.5 kcal/mol. Based on the co-relation of our in silico studies with the experimental findings of azlocillin binding to PBPs in other bacteria we speculate that PBP-3Bb can be one of potential target for azlocillin. In addition to it, studying of azlocillin binding to alternative targets cannot be excluded.
In this current study, we examined the efficacy of azlocillin and cefotaxime post 7, 14 and 21 infection in C3H mice model. We have inoculated higher dose of B. burgdorferi (2 × 10 5 ) because they replicate fastly in high titres and spreads to tissue sites effectively. Cefotaxime treatment failed to eradicate the B. burgdorferi infection completely in mice infected for 7 and 14 days. We found both doxycycline and azlocillin cleared Borrelia infection at 7 days of infection. In order to study the relative efficacies of doxycycline and azlocillin in mice we have Many studies have also shown that even after longterm treatment with doxycycline and ceftriaxone still some B. burgdorferi DNA was detected in mice and rhesus maca ques 13,16,35,52,53 . It is well known that in 10-20% of the people treated with Lyme disease have PTLDS. The PTLDS might be either due to presence of persistence bacteria or due to its residual particles left over after the infection. Irrespective of causing factors for PTLDS, effective treatment of Lyme patients with more potent drugs can minimize exposure of patients to infection which can prevent PTLDS.
As it is not convenient to use intravenous drugs like azlocillin for treatment compared to oral available drugs, experiments are ongoing for developing oral delivery of azlocillin. Further efficacy studies should be done by long term treatment with azlocillin to check whether it can clear the B. burgdorferi in all the infected mice completely. These results are very encouraging to conduct further in vivo studies on different B. burgdorferi strains. Though azlocillin showed good efficacy in C3H mice model, further studies should be done in in vivo persisters model to prove whether these drugs have potential to eliminate persisters or not. In addition to it, further additional preclinical and clinical studies should conduct for repurposing these drug molecules and clinical acceptance, thereafter.

Materials and Methods
Bacterial strains and culture. The low passage B. burgdorferi sensu stricto strain JLB31 was (generously provided by Dr. Linden Hu Tufts University, Boston, MA, USA) cultured in Barbour-Stoenner-Kelly II (BSK-II) complete medium supplemented with 6% rabbit serum (Sigma, St.Louis, MO, USA). The cultures were incubated in sterile 50 mL falcon tubes (Corning Incorporated, Corning, NY, USA) at 33 °C for 3-7 days in 5% CO 2 incubator without antibiotics.
Semisolid plating was chosen to obtain the exact count of the growing borrelial colonies as colony forming units (CFU). We performed semisolid plating procedure as described by Jenny A. Hyde etal 54 . The 2X BSK-II medium was prepared in the following manner. To the 500 ml of CMRL-1066 medium: 50 g of bovine serum albumin (Sigma), 5 g neopeptone (BD), 6.6 g HEPES acid (Sigma), 0.7 g sodium citrate (Sigma), 5 g glucose (Sigma), 2 g yeastolate (BD), 2.2 g sodium bicarbonate (Fisher), 0.8 g sodium pyruvate (Sigma), 0.4 g N-acetyl-glucosamine (Sigma) were added and mixed thoroughly. Finally pH of the medium was adjusted to 7.6 and filtered through 0.2 µm filter units. For plating the medium is mixed in the following way. The 100 ml of 2X BSK-II medium prewarmed at 55 °C was mixed with 100 ml 1.7 ml of agarose (55 °C) and 14 ml sterilized rabbit serum and equilibrated to 55 °C. Then 10 ml of equilibrated BSK-II medium was dispensed into 60-mm petri dishes as bottom agar and allowed to solidify. Finally, the sample was resuspended in 0.5 ml fresh BSK-II medium and mixed with 10 ml of BSK-II agarose medium (55 °C) and poured as a top agar. The plates were incubated in the incubator with 5% CO2 at 33 °C for minimum of 21 days. The white visible colonies were counted after 21 days for the analysis. www.nature.com/scientificreports www.nature.com/scientificreports/ Antimicrobial agents. The 10 mM of drug stocks of Azlocillin sodium salt (Cayman chemicals, Ann Arbor, MI), Cefotaxime acid (Cayman chemicals) and Doxycycline (sigma) were prepared by dissolving in sterile distilled water. The 10 mM Mitomycin C stock was prepared by solubilizing in DMSO.
Dose-dependent killing of B. burgdorferi. The dose-dependent killing of B. burgdorferi was performed with log phase (3 days old) and stationary phase (7 days to 10 days old). To determine efficacy of drugs, 2 × 10 6 / ml of Borrelia were used from log and stationary phase cultures and grown in BSK-II medium in 48-well plates with drugs at varying concentrations ranging from 1.25, 2.5, 5, 10, 20, 40 and 80 μg/ml. The Borrelia cultures were incubated at 33 °C with 5% CO 2 for 5 days. After 5 days the cultures were centrifuged at 13,000 rpm for 10 minutes. Then the pellets were washed, resuspended in 0.5 ml of fresh BSK-II medium and serially diluted. Finally, the cultures diluted in 0.5 ml BSK-II medium were mixed with 10 ml of BSK agarose and poured as top agar. Plates were incubated at 33 °C with 5% CO 2 up to 21 days and visible colonies were counted. The experiments were done atleast thrice in triplicates.
Determining efficacy of drugs on doxycycline tolerant persisters. The 10 6 / ml of Borrelia taken from log and stationary phase were cultivated with varying doxycycline concentrations of 2.5, 5 and 10 μg/ml. The doxycycline cultivated Borrelia cultures were incubated in 48-well plates for 3 days at 33 °C with 5% CO 2 . Then the cultures were centrifuged for 10 minutes at 13,000 rpm. The remaining doxycycline tolerant Borrelia pellets were washed and incubated with 1 ml of BSK-II medium containing drugs of varying concentrations for 7 days. As a control doxycycline tolerant Borrelia obtained were also resuspended again with doxycycline concentrations of 2.5, 5 and 10 μg/ml. After incubation for 7 days the cultures were pelleted, washed and resuspended in 0.5 ml of fresh BSK-II medium. Then the semisolid plating was done by mixing cultures diluted in 0.5 ml BSK-II medium with 10 ml of BSK agarose and poured as top agar. The agar plates were incubated at 33 °C with 5% CO 2 up to 21 days. The white visible colonies were counted for generating persister curve. All the experiments were done atleast thrice with triplicates.
time-dependent killing studies. Time kill studies were performed with Borrelia isolate JLB31 (B. burgdorferi s.s.) to determine the rate of antimicrobial activity of azlocillin. The10 6 per mL Borrelia were grown in BSK-II medium with azlocillin at concentrations 20 and 40 μg/ml and doxycycline at concentrations 5 and 10 μg/ ml. BSK-II medium with no drugs was used as a control. At time intervals of 2, 6, 10, 24, 48, 96 and 144 hours B. burgdorferi cultures were centrifuged for 10 minutes at 13000 rpm and the pellet was resuspended in 0.5 ml of fresh BSK-II medium, serially diluted, mixed with 10 ml of BSK agarose and poured as top agar. After 21 days antibacterial activity was analyzed by counting bacteria colonies at all the time points performed. The experiment was done once with triplicates.
In vivo testing of drugs in C3H/HeN mice. Four weeks old female C3H/HeN mice, were purchased from Charles River Laboratories, Wilmington, Massachusetts. All mice were maintained in the pathogen-free animal facility according to animal safety protocol guidelines at Stanford University under the protocol ID APLAC-30105. All experiments were in accordance with protocols approved by the Institutional Animal Care and Use Committee of Stanford University. The mice were infected intradermally with 0.1 mL BSK medium containing 2 × 10 5 B. burgdorferi JLB31. On the 7, 14 and 21 days of infection, the mice were intraperitoneally administered a daily dose of drugs, azlocillin (50 mg/kg), cefotaxime (30 mg/kg) and doxycycline (50 mg/kg) for 5 consecutive days (Tables 1, 2 and 3). After 48 hours of the last dose of administering compounds, the mice were sacrificed and their urinary bladders, ears, and hearts were cultured in BSK-II medium. The cultures were evaluated for the presence of motile spirochetes after 21 days using the dark-field microscopy 51 . The DNA was extracted from urinary bladder and ear. If the B. burgdorferi was observed in any one of the organ in the mice, the animal was considered as infected. The absence of B. burgdorferi propagation marked the effectiveness of the treatment in these organisms. Quantitative Real-time pcR (qpcR) analysis of B. burgdorferi DnA from tissues. Urinary bladder, ear punches (20 mg per sample) were homogenized and DNA was extracted using the NucleoSpin tissue kit according to the manufacturer's instructions (Düren, Germany). qPCR from above tissues were performed in blinded samples using three oligonucleotides, two primers and a probe for B. burgdorferi Fla-B gene. These primers were listed as follows: Fla-B primers Flab1F 5′-GCAGCTAATGTTGCAAATCTTTTC-3′, Flab1R 5′-GCAGGTGCTGGCTGTTGA-3′ and TAMRA Probe 5′-AAACTGCTCAGGCTGCACCGGTTC-3' according to the published protocol. The amplification protocol consisted of 10 min at 95 °C, followed by 40 cycles of amplification (95 °C for 15 s and 60 °C for 1 min). A negative result was assigned where no amplification occurred or if the threshold cycle (CT) was greater than 36. Reactions were performed in duplicate for each sample. Results were plotted as the number of Borrelia per microgram of tissue. The lower limit of detection was 1 to 100 copies of B. burgdorferi Fla-B DNA per mg of tissue. In addition to standard laboratory measures to prevent contamination, negative controls (containing PCR mix, Fla-B primers, and Taq polymerase devoid of test DNA) were included. Molecular docking. The three-dimensional structure of pbp of B. burgdorferi was not available, so it was built using homology modelling by employing SWISS-MODEL 55 . The pencillin-binding protein (PBP) of P. aeruginosa was identified as the homologous protein using BLASTP and hence it was used as the template for building the three-dimensional structure for the borrelia PBP. The quality of the modeled structure was verified using Ramachandran plot (Laskowski et al., 1993 and 56 and superimposition of modeled structure with that of the template structure was carried using Chimera 57 . Azlocillin was docked with modeled structure of borrelia PBP using AutoDock Vina 58 . Molecular interaction between Azlocillin and borrelia PBP was analyzed using Discovery