Isolation of Borrelia miyamotoi and other Borreliae using a modified BSK medium

Borrelia spirochetes are the causative agents of Lyme borreliosis (LB) and relapsing fever (RF). Despite the steady rise in infections and the identification of new species causing human illness over the last decade, isolation of borreliae in culture has become increasingly rare. A modified Barbour-Stoenner-Kelly (BSK) media formulation, BSK-R, was developed for isolation of the emerging RF pathogen, Borrelia miyamotoi. BSK-R is a diluted BSK-II derivative supplemented with Lebovitz’s L-15, mouse and fetal calf serum. Decreasing the concentration of CMRL 1066 and other components was essential for growth of North American B. miyamotoi. Sixteen B. miyamotoi isolates, originating from Ixodes scapularis ticks, rodent and human blood collected in the eastern and upper midwestern United States, were isolated and propagated to densities > 108 spirochetes/mL. Growth of five other RF and ten different LB borreliae readily occurred in BSK-R. Additionally, primary culture recovery of 20 isolates of Borrelia hermsii, Borrelia turicatae, Borrelia burgdorferi and Borrelia mayonii was achieved in BSK-R using whole blood from infected patients. These data indicate this broadly encompassing borreliae media can aid in in vitro culture recovery of RF and LB spirochetes, including the direct isolation of new and emerging human pathogens.


Results
In vitro isolation of B. miyamotoi RI13-2395. In vitro isolation initially focused on B. miyamotoi from U.S. sources, given the prior lack of isolation in BSK-H. No growth occurred when blood from a laboratory infected mouse was seeded into BSK-IIB. In contrast, growth of B. miyamotoi occurred when the same blood was seeded into a confluent monolayer of ISE6 tick cells in tick cell co-culture medium (TCCM) as well as into BSK-IIB and L-15B300, the primary component of TCCM, combined at a ratio of 1:1 (Fig. 1). In the latter case, subsequent passage into BSK-IIB:L-15B300 (1:1), with BALB/c mouse serum substituted for rabbit serum (BSK-IIB-MS), yielded a peak spirochete density of 3.0 × 10 7 spirochetes/mL. The recovered spirochetes were confirmed as B. miyamotoi by glpQ PCR and 16S rDNA sequencing (426 nucleotides, 100% identity to B. miyamotoi LB-2001). At this time, a strain name, RI13-2395, was assigned and either this actively growing culture or frozen stocks (passage 2; P2) used for subsequent experiments.

Inhibition of B. miyamotoi growth in BSK-IIB.
To determine if growth of B. miyamotoi RI13-2395 in BSK-IIB-MS:L-15B300 (1:1) was due to the addition of L-15B300 or from decreased concentrations of BSK-IIB ingredients, side-by-side growth comparisons were performed with BSK-IIB-MS:L-15B300 (1:1), BSK-IIB-MS alone, and BSK-IIB-MS mixed with H 2 0 at four different ratios (Fig. 2a). Actively growing spirochetes (P2) were inoculated into all media formulations at a final concentration of 1.6 × 10 5 spirochetes/mL. Notably, no growth of B. miyamotoi occurred in BSK-IIB-MS. Only when BSK-IIB-MS was diluted with H 2 0 did growth of B. miyamotoi ensue. Similar growth (8 × 10 6 spirochetes/mL; doubling time of 17 h) was achieved on day four for the 1:1,3:2 and 2:1 mixtures as compared to BSK-IIB-MS:L-15B300 (1:1) (10 7 spirochetes/mL; doubling time of 16 h). The highest peak densities were observed when BSK-IIB-MS was diluted with H 2 0 at ratios of 1:1 and 3:2.  (Fig. 2b). Individually reducing the concentration of proteinaceous components of BSK-IIB, Bacto neopeptone, bovine serum albumin (BSA), and Bacto TC yeastolate, and other components, by half did not relieve inhibition of B. miyamotoi growth to the same degree as was evidenced for CMRL 1066 ( Fig. 2b; data not shown). Continuous growth of spirochetes was not observed beyond day four when any single component was reduced by half.
BSK-R. BSK-IIB-MS diluted 40% with water [BSK-IIB-MS:H 2 O (3:2)] served as the basis for further media optimization using frozen P2 stocks of B. miyamotoi RI13-2395. Lebovitz's L-15, a primary component of L-15B300, was added at final concentrations ranging from 4 to 10 g/L to look for growth augmentation. A final concentration of either 5 or 6.7 g/L [final concentration in BSK-IIB-MS:L-15B300 (1:1)] stimulated growth of B. miyamotoi by three-fold as compared to BSK-IIB-MS diluted 40% with water. A spirochete density of 3.0 × 10 7 cells/mL was achieved after six days of growth when 5 g/L Lebovitz's L-15 was included. At the uppermost concentration tested, 10 g/L, spirochete growth was impaired. Actively growing spirochetes were seeded into media formulations at a final concentration of 1. 6 × 10 5 spirochetes/mL. The percentages of BSK-IIB-MS to H 2 0 in the 1:1 and 3:2 ratios, was 50:50 and 60:40, respectively. (b) Side by side comparisons of cell densities in BSK-IIB:L-15B300 (1:1), BSK-IIB-MS alone, and BSK-IIB-MS with individual indicated components reduced by half. The final concentration of indicated components was 4.04 g/L CMRL 1066 10X w/o L-glutamine, 20.85 g/L BSA, 2.01 g/L Bacto neopeptone and 0.84 g/L Bacto TC yeastolate. Actively growing spirochetes were seeded into media formulations at a final concentration of 1. 6 × 10 5 spirochetes/mL. (a,b) Spirochete densities by day are indicated by color: day 0, black; day 4, grey; day 5, yellow; day 6, blue. www.nature.com/scientificreports/ Table 1 provides a recipe for the final media formulation, designated BSK-R, which includes 5 g/L Leibovitz's L-15 and components of BSK-IIB adjusted to 60%, with the exception of 4.3 g/L CMRL-1066 w/o L-glutamine, 5.0 g/L glucose as well as 2.2 g/L HEPES and 0.98 g/L sodium bicarbonate, which were adjusted to give a final pH of 7.5. Serum was added to a final concentration of 13% (7% mouse serum, 4% rabbit serum and 2% fetal bovine serum).
Comparison of B. miyamotoi growth in BSK-R, BSK-IIB-MS diluted 40% with water, and BSK-IIB-MS:L-15B300 (1:1), using a P2 frozen stock as the starting inoculum, indicated that B. miyamotoi strain RI13-2395 displayed an improved doubling time of 18 h in BSK-R as compared to 23 h in the original BSK-IIB-MS:L-15B300 (1:1) formulation used initially to propagate this strain. The spirochete density achieved at this time point (96 h) was also two-fold higher in BSK-R as compared to BSK-IIB-MS:L-15B300 (1:1). After eight days, a maximum density was reached, with 1.3 × 10 8 , 3.9 × 10 7 and 7.2 × 10 7 spirochetes/mL observed in BSK-R, BSK-IIB-MS diluted 40% with water and BSK-IIB-MS: L-15B300 (1:1), respectively. Direct isolation of B. miyamotoi from Ixodes scapularis ticks and Tamias striatus in BSK-R. Seventeen single nymphs, derived from five female adult B. miyamotoi infected I. scapularis ticks, were seeded into BSK-R with antibiotics. Borrelia miyamotoi spirochetes were successfully recovered in 14 of the 17 inoculated cultures, yielding a recovery rate of 82%. The average density without passage was 1.1 × 10 7 spirochetes/mL and the average time to this density was 11.5 days (range seven to 14 days). Spirochetes were confirmed as B. miyamotoi by glpQ PCR and/or 16S rDNA sequencing. Additionally, one isolate, CT13-2396, was genome sequenced, with this sequence made publicly available (GenBank accession numbers CP017126 to CP017140) 15 . When blood (collected with EDTA) from a field collected chipmunk (T. striatus) was inoculated into BSK-R with antibiotics, growth also occurred, albeit at slower rates as compared to I. scapularis ticks. Motile spirochetes (1 × 10 4 spirochetes/mL) were observed after 13 days. Upon passage of spirochetes to fresh BSK-R, a density of 7.7 × 10 6 spirochetes was achieved 11 days later. The recovered spirochete was ospA PCR negative and glpQ PCR positive.
No growth of B. miyamotoi occurred in BSK-R when nine single minced I. pacificus nymphs, derived from a field collected female adult tick (glpQ PCR positive, ospA PCR negative), were incubated for 30 days, including two blind passages into fresh media. Seeding of blood (20 µL) from a SCID mouse infected with B. miyamotoi via I. pacificus feeding, also yielded no growth in BSK-R, however, motile spirochetes (3.5 × 10 4 spirochetes/mL) were detected microscopically in ISE6 co-cultures and subsequently passaged into BSK-R. The recovered B. miyamotoi isolate, CA17-2241, was genome sequenced with the chromosome and large linear plasmid sequences publicly available (GenBank accession numbers CP021872 and CP021873) 16 .

BSK-R supports growth of both relapsing fever and Lyme borreliae. Side-by-side comparisons
demonstrated three other RF borreliae, B. hermsii, Borrelia turicatae, and Borrelia recurrentis, grew to high densities without passage in BSK-R and BSK-IIB, with ~ 10 8 spirochetes/mL achieved for all ( Table 2). Two B. burgdorferi sensu lato genospecies, B. burgdorferi and B. mayonii, were similarly propagated in both BSK-R and BSK-IIB ( Table 2). As expected, the B. miyamotoi isolates recovered in this study, LB-2001 and three B. miyamotoi strains originating from Japan all grew in BSK-R with maximum concentrations ranging from 3.6 × 10 7 to 1.5 × 10 8 spirochetes/mL after 6-10 days. In contrast, BSK-IIB was unable to support growth of any North American B. miyamotoi strain, with the exception of B. miyamotoi CA17-2241, which reached a maximum cell density of 4.2 × 10 6 spirochetes/mL after 19 days. For the three B. miyamotoi strains originating from Japan, spirochete motility and growth rate were improved in BSK-R as compared to BSK-IIB. Twelve additional strains encompassing Borrelia   Direct isolation of Borrelia species from human blood and field collected ticks using BSK-R. When 21 human whole bloods (50-500 µL collected with EDTA) from infected patients were inoculated into BSK-R with antibiotics, isolates of both RF (B. miyamotoi, B. hermsii, B. turicatae) and Lyme borreliae (B. mayonii and B. burgdorferi) were successfully recovered (Table 3). For all five species, spirochetes were detected by periodic microscopy after three to 15 days incubation without passage. A single B. mayonii culture required blind-passage on day eight into fresh BSK-R at a dilution of 1% (v/v) for detection of spirochetes; this sample was not received until 38 days post collection with considerable hemolysis evident. In all cases, recovered spirochetes were subsequently propagated to densities ≥ 5 × 10 6 spirochetes/mL. Starting inocula were estimated at < 10 3 to 10 7 spirochetes based on microscopic quantitation of the original blood samples. A microscopic limit of detection of 8.3 × 10 3 spirochetes/mL blood (one spirochete in 600 fields of blood diluted 1:10) precluded determination of spirochete numbers lower than 10 3 . The average time for microscopic detection of spirochetes in B. hermsii, B. turicatae and B. mayonii recovered cultures was six, seven, and nine days, respectively. The single B. miyamotoi isolate required 15 days of growth for detection. The quality of the blood specimen (hemolysis, contamination, length of time from collection to culture), the level of spirochetemia and the infecting Borrelia species were associated with the time to microscopic detection of spirochetes in culture. Lower inocula and low-quality specimens correlated with longer recovery times. One B. hermsii isolate was successfully recovered from blood collected two days after azithromycin treatment, while the B. burgdorferi isolate was recovered from a heavily contaminated blood specimen. All recovered isolates were verified to species level by PCR, eight housekeeping gene multi-locus sequence typing or 16S sequencing. Direct recovery of B. burgdorferi and B. miyamotoi was also achieved when single field-collected I. scapularis adult ticks (n = 55) were homogenized and inoculated into BSK-R with antibiotics. Twenty-six isolates of B. burgdorferi were readily obtained within seven days of growth and confirmed by PCR (ospA positive, glpQ negative). Markedly, both B. burgdorferi and B. miyamotoi were simultaneously recovered in BSK-R within seven

Discussion
BSK-R, a diluted BSK-II derivative supplemented with Lebovitz's L-15, mouse and fetal calf serum, was developed and successfully used for the direct in vitro isolation of 15 freezer-retrievable isolates of B. miyamotoi originating from animal and tick sources in the eastern or upper midwestern U.S. This media also broadly supported growth of 15 other Borrelia species encompassing the RF group as well as the LB group, all of which grew from frozen stocks to unpassaged cell densities of > 10 7 spirochetes/mL. Notably, primary cultures (21 isolates) of B. miyamotoi, B. hermsii, B. turicatae, B. burgdorferi and B. mayonii were successfully recovered from the blood of infected patients using BSK-R. These include the first isolations from human cases in the U.S. for the two emerging pathogens, B. miyamotoi (this report) and B. mayonii 11 . Unexpectedly, we discovered North American B. miyamotoi only grew in BSK-IIB-MS when all components were diluted with water. This finding yielded the single most important media modification for successful growth of B. miyamotoi. The primary component responsible for growth inhibition was CMRL 1066. Decreasing its final concentration by 50% to 4.04 g/L, with all other components held constant, enabled initial growth of B. miyamotoi, although it was not sufficient for sustained propagation. The concentration of motile bacteria arrested at day four, suggesting at least one other component of BSK-IIB-MS also negatively impacts B. miyamotoi growth. CMRL 1066 was originally incorporated into Stoenner's enrichment of Kelly's medium [final concentration estimated to be 4.87 mg/mL (5% of a 10X solution)] to improve antigen expression and growth of B. hermsii 17 . The concentration of CMRL 1066 in BSK-II, which was optimized for growth of B. burgdorferi, is ~ 7.1 mg/ mL 18 . The concentration of CMRL 1066 in BSK-H is ~ 10 mg/mL. We postulate the higher concentration of CMRL 1066 and other components is why B. miyamotoi from ticks and patients in the U.S. was not cultivable in this medium 13,14 . Although Lebovitz's L-15 augmented growth of B. miyamotoi, it was not able to substitute for CMRL 1066 (data not shown). Lebovitz's L-15, like CMRL 1066, is comprised of amino acids and vitamins, but without other factors present in CMRL 1066. It is buffered with salts, free base amino acids and galactose to help maintain physiological pH control in CO 2 free systems 19 .
The sensitivity of Borrelia species to serum complement varies by animal source and has been linked to reservoir host competency in nature 20 . Initial incorporation of mouse serum into media formulations was based on evidence suggesting B. miyamotoi is resistant to mouse serum complement. Borrelia miyamotoi was initially detected in mice in nature 5 . Additionally, mouse serum (CD1 and BALB/c) was not bactericidal for B. miyamotoi when tested in an in vitro assay, whereas rabbit serum was (data not shown). The negative effects of rabbit sera on B. miyamotoi could be decreased, but not eliminated, when serum was inactivated at 0.5 h at 56 °C. The www.nature.com/scientificreports/ observed sensitivity to rabbit serum complement is consistent with a prior study where 50% human serum, but not rabbit serum, was suitable for propagation of B. miyamotoi 21 . Further inactivation of rabbit serum for 45 min and incorporation into BSK-R as the sole source of serum (13%) was found to support growth of B. miyamotoi, albeit with lower spirochete motility and slower growth as compared to mouse serum. Nonetheless, this finding indicates rabbit serum, a less expensive alternative to mouse serum, can substitute when using BSK-R for propagation of Borrelia stock cultures. Continued work is needed to further define serum requirements for the direct in vitro recovery of North American B. miyamotoi. Primary culture recovery of Borrelia species in BSK-R was successfully achieved from bloods, of infected patients, collected in EDTA and with spirochete concentrations estimated at < 10 3 and of low quality (collected 39 days prior, no motile spirochetes, heavily contaminated, and post-antibiotic treatment). Initially, blood was seeded into BSK-R at dilutions no higher than 1% (v/v), with no removal of blood components. Subsequently, this protocol was modified to include removal of EDTA and components released from hemolyzed blood cells, given their inhibitory effect on growth of B. miyamotoi and/or B. burgdorferi [22][23][24] . With this modification, larger blood volumes were successfully used for recovery of cultures.
The time to detection of B. miyamotoi in primary cultures was slowed and the unpassaged spirochete density lower when blood specimens versus tick homogenates were used as the starting inoculum. The two isolates recovered from animal and human blood both required ~ 14 days of growth in BSK-R for their detection, whereas the recovery time from tick homogenates was as low as seven days, even when B. burgdorferi was concurrently isolated. This suggests something about the inocula themselves, such as a higher spirochete number in ticks, an inhibitory component in blood or a growth component in ticks, might be responsible for this difference. PCR quantitation has been used to estimate the number of B. miyamotoi genome equivalents in blood from infected U.S. patients at 10 4 -10 5 /mL 25 ; whether the number of spirochetes in I. scapularis ticks is higher requires further investigation. Of note, primary culture recovery of B. miyamotoi from the plasma of patients in Russia also required incubation in MKP-F for 12-20 days 22 . The slowed time for primary recovery of B. miyamotoi is consistent with its longer doubling time as compared to other borreliae. A doubling time of ~ 18 h was determined for U.S. B. miyamotoi in BSK-R and in MKP supplemented with 50% human serum 21 , whereas the doubling time for B. mayonii in BSK-R is ~ 9 h. Given primary isolation of B. miyamotoi from blood requires ~ 14 days, it's likely media components critical for growth deteriorate with age and why growth could be improved upon detection of spirochetes and transfer into fresh media.
Borrelia miyamotoi was not isolated from I. pacificus in BSK-R, although it was successfully propagated in BSK-R, once an isolate was recovered via co-culture of infected SCID mouse blood with an embryonic tick cell line, ISE6. The reason for this is not clear and requires further work. Challenges occurred with transmission and feeding of I. pacificus on both immunocompetent and SCID mice and may have been due to the condition of the ticks. Mold was present in the maintained ticks and few spirochetes were observed in tick midguts. As sequences of B. miyamotoi from I. pacificus and I. scapularis have been shown to differ, it is also possible strain differences might be associated with recovery in BSK-R 41 .
While this work was in progress, two tick-derived isolations of North American B. miyamotoi were reported. M1029 and LB-2001 were isolated from SCID mouse plasma in MKP diluted with 50% human serum and in MKP-F (which is diluted with 10% fetal calf serum), respectively 21,42 . We postulate that growth in these two cases is due to a reduced concentration of CMRL 1066; 3.2 and 6.3 g/L for MKP containing 50% human serum and MKP-F, respectively, as well as other components. In the case of MKP supplemented with 50% human serum, the volume of serum added would also have reduced the concentrations of all other MKP ingredients by half 21 . While the addition of 50% human serum likely had a positive effect on growth, our finding that B. miyamotoi grew in BSK-IIB-MS diluted 50% with water, suggests the reduction of MKP components by half may be the primary reason for growth of North American B. miyamotoi in this media.
Given the lengthy turnaround times and specialized growth requirements of Borrelia species, use of culture as a standard diagnostic test is limited. Nonetheless, its value should not be understated. Isolation of Borrelia species is paramount to a better understanding of the natural biology and pathogenesis of newly identified human pathogens or those which represent distinct geographic and ecological niches. Additionally, the public health importance of Borrelia species dictates the use of culture as a diagnostic benchmark for a true understanding of the geographic regions these infections occur. The development of a media formulation that enables growth and direct isolation of both RF and LB Borrelia species, should help to improve our understanding of these important tick-transmitted pathogens. Given that batch/lot differences of BSK ingredients have previously been shown to influence spirochete recovery 10 further research directed towards optimization and standardization of BSK-R as well as simplification is important. www.nature.com/scientificreports/ plaster of Paris and activated charcoal and allowed to hatch to larvae. Larvae (5-100 per mouse) were allowed to feed to repletion on CD1 mice and molt to nymphs. Blood was collected in EDTA tubes from small mammals live trapped in Minnesota 35 . All work with laboratory mice was approved by the Centers for Disease Control (CDC) and Prevention Division of Vector-Borne Diseases Animal Care and Use Committee in accordance with the Animal Welfare Act. Human EDTA whole blood specimens were submitted for Borrelia diagnostic testing or were residual samples from research studies.

Mouse infection.
Approximately 200 I. scapularis eggs, were surface sterilized by submerging in H 2 O 2 for one minute, rinsed with sterile saline, submerged in 70% ethanol for 2 min, rinsed again and ground in a 1.7 mL Eppendorf tube containing 200 µLs sterile saline using a plastic pestle. BALB/c mice were inoculated subcutaneously with 100 µLs of egg homogenate. To infect SCID mice, I. pacificus nymphal ticks (3 per mouse) were allowed to feed to repletion. Mice were euthanized and blood collected by cardiac puncture without the use of anticoagulant.
Tick cell co-culture. I. scapularis embryonic (ISE6) cells 43 were cultivated in tissue culture flasks with plug seal caps at 34 °C with caps tight. Tick Cell Co-culture Media (TCCM) consisted of 90% L-15B300, prepared according to Munderloh 44 and 10% BSK-IIB, a modification of the media used to culture Borrelia lonestari 45 . For seeding confluent cell cultures, 20-250 µL of infected mouse blood was added to flasks. Media was replaced every 7-10 days, by decanting, followed by centrifugation at 3300(g) for 8 min. The supernatant was discarded, the resultant pellet resuspended in 5-6 mL fresh media and transferred back into the flask. For inoculation of BSK-R, 50 µL of EDTA whole blood from field collected animals was added, followed by gentle mixing via pipetting. Red blood cells (RBCs) were allowed to settle for 48 h and the supernatant transferred to a new tube. EDTA was not removed. Ixodes ticks were surface sterilized by 3% H 2 O 2 (5 min) and 70% ethanol (11 min) washes and air dried on bibulous paper. Nymphs minced with scalpels on glass slides and adults homogenized using Wheaton Tenbroeck tissue grinders (DWK Life Sciences Millville, NJ) were inoculated into cultures. 50-500 µLs of EDTA whole blood from patients was used to seed cultures. If blood volumes larger than 100 µLs were inoculated, red blood cells (RBCs) were allowed to settle overnight. To remove EDTA and red blood cell components, the resulting supernatant was removed and centrifuged at 3300(g) for 8 min, with the pellet resuspended in fresh media.
All cultures were incubated in 5 mL or 14 mL snap cap round bottom polypropylene Falcon tubes or 5 mL Eppendorf centrifuge tubes at 34 °C with tubes filled to 90% capacity and caps pressed on tight, creating a microaerophilic environment. BSK-R seeded with blood from infected animals or patients was refreshed after approximately two weeks of culture; supernatants were removed, and centrifuged at 3,300(g) for 8 min, with the pellet resuspended in fresh media. Tubes were examined periodically by dark field microscopy for spirochete growth. Microscopy. Wet mounts for dark field microscopy were prepared by pipetting 5 µLs of culture or a 1:10 dilution of whole blood onto a glass slide and covering with a 22 mm × 22 mm coverslip. Wet mounts were visualized using an Axio Imager.A2 microscope (Ziess, Thornwood, NY) at a total magnification of 400X. The number of spirochetes per field was determined by counting motile spirochetes in at least 20 fields. The exception was for human blood specimens, where both motile and non-motile spirochetes were counted. The number of spirochetes per mL was determined as follows: average number of spirochetes per 400X field × (5 × 10 5 spirochetes/mL) dilution factor. The conversion factor was determined as previously described 11 . Spirochete motility was assessed based on observation of rapid corkscrewing movement.  46 : BmglpQ-F GAC AAT ATT CCT GTT ATA ATGC, BmglpQ-R CAC TGA GAT TTA GTG ATT TAA GTT C, and BmglpQ-P FAM-CCC AGA AAT TGA CAC AAC CAC-BHQ1. The B. burgdorferi ospA primers were modified from Ivacic et al. 47 : OspA-F AAT ATT TAT TGG GAA TAG GTC TAA , OspA-R CAC CAG GCA AAT CTA CTG A and TTA ATA GCA TGT AAG CAA AAT GTT AGCA-BHQ1. Each 20 µL PCR reaction contained 5 mM MgCl 2 , 500 nM each primer, 100 nM probe, 1 × PerfeCta qPCR FastMix II Low ROX (Quanta Biosciences, Beverly, MA), 5 µL DNA extract. Cycling conditions on the 7500 FAST Dx (Applied Biosystems, Foster City, CA) were 95 °C 3 min, and 50 cycles of 95 °C for 10 s and 57 °C for 30 s. PCR assays for B. turicatae and B. hermsii were as described 48 . For sequencing, primers were glpQ-F GGA GCT GAC TAC CTA GAA C, glpQ-R GGG TAT CCA AGG TCC AAT TCC, 16 s-F CCT GGC TTA GAA CTA ACG , and 16 s-R TTC GCC TCT GGT ATT CTT CC. Reactions were as previously described 34 . Consensus sequences were assembled with Clone Manager 9.3 (Sci-Ed, Denver, CO) and aligned to Borrelia sequences in GenBank. One to eight housekeeping genes were amplified, sequenced, and analyzed for identification of B. mayonii 25 . Media formulations. To prepare 600 mL of an in-house formulation for B. hermsii growth, (here termed BSK-IIB), 500 mL Milli-Q water and 25.0 g universal grade Bovine Serum Albumin (BSA) (Millipore Sigma, St. Louis, MO) was added to a 1 L flask and stirred slowly until fully dissolved. The following reagents were next added, one at a time, ensuring each reagent was fully dissolved before addition of others; 4.85 g CMRL 1066 10X w/o L-glutamine, powder, (US Biological, Salem, MA), 2.5 g Gibco Bacto neopeptone (Fisher Scientific, Pittsburgh, PA), 3.0 g HEPES sodium salt (Fisher Scientific), 0.35 g sodium citrate (Millipore Sigma), 3.5 g D-( +)-Glucose (Millipore Sigma), 0.4 g sodium pyruvate, sodium salt (Millipore Sigma), 0.2 g N-acetyl-D-glucosamine (MP Biomedicals, Irvine, CA), 1.1 g sodium bicarbonate, anhydrous (Millipore Sigma), 1.0 g Gibco Bacto TC yeastolate (Fisher Scientific), and 0.14 g tissue culture grade L-cysteine HCl monohydrate (Fisher Scientific). Once all reagents were fully dissolved, 72 mL of heat inactivated (56 °C for 45 min) sterile, trace hemolyzed rabbit serum (Pel-Freez, Rogers, AR) was added followed by 0.04 g Invitrogen UltraPure dithiothreitol (DTT) (Fisher Scientific). Finally, the pH was adjusted to 7.5 and antibiotics were added followed by filter sterilization using a 0.22 µm Millipore Express PLUS filtration system (Millipore Sigma). This same formulation with substitution of heat inactivated (56 °C for 30 min) BALB/c mouse serum (Innovative Research, Novi, MI) in place of rabbit serum was termed BSK-IIB-MS.
To prepare 600 mL of BSK-R, 300 mL Milli-Q water and 15.0 g BSA (universal grade) was added to a 1 L flask and stirred slowly until fully dissolved. Simultaneously, 200 mL Milli-Q water and 3.0 g Gibco Leibovitz's L-15 medium, powder (Fisher Scientific) was added to a 500 mL flask and stirred slowly until fully dissolved. To the 1 L flask containing fully dissolved BSA, the following reagents were added one at a time ensuring each reagent was fully dissolved before addition of next one; 2.6 g CMRL 1066 10X w/o L-glutamine (powder), 1.5 g Gibco Bacto neopeptone, 1.3 g HEPES sodium salt, 0.21 g sodium citrate, 3.0 g D-( +)-Glucose, 0.24 g sodium pyruvate, sodium salt, 0.12 g N-acetyl-D-glucosamine, 0.59 g sodium bicarbonate, 0.6 g Gibco Bacto TC yeastolate and 0.084 g L-cysteine HCl monohydrate. After reagents were fully dissolved, 24 mL heat inactivated (56 °C for 45 min) trace hemolyzed, sterile rabbit serum, 42 mL heat inactivated (56 °C for 30 min) BALB/c mouse serum and 12 mL heat inactivated (56 °C for 30 min) Corning premium fetal bovine serum (FBS) (Fisher Scientific) were added to the same flask, followed by 0.024 g Invitrogen UltraPure DTT and the fully dissolved Leibovitz's L-15 medium. Finally, the pH was adjusted to 7.5, antibiotics were added and the media filter sterilized as above.
Single use aliquots of complete media (including serum and antibiotics) were prepared in 5 mL or 14 mL falcon tubes and stored at − 20 °C in the dark for up to one year.
Ethics statement. The secondary use of residual samples described herein was performed in accordance with the relevant guidelines and regulations under a protocol approved by the CDC Institutional Review Board with an approved waiver of informed consent. The views and opinions expressed herein are those of the authors alone and do not represent the official position of the Centers for Disease Control and Prevention.