Microbroth dilution method for antibiotic susceptibility testing of fastidious and anaerobic bacteria of the urinary microbiome

ABSTRACT Bacterial isolates from the human urinary microbiome have been extensively studied for their antibiotic resistance; however, little work has been done on those isolates that are difficult to grow in vitro. This study was designed to qualify a serum-based medium, New York City Broth III (NYCIII), and a broth microdilution method to determine the antibiotic susceptibility of previously underreported or undescribed microbes that have a difficult time growing in standard Mueller-Hinton broth. Here, we demonstrate that NYCIII microbroth dilution can be an effective method for the determination of antibiotic susceptibility of species found in the human urinary microbiome. We show that this method serves well to characterize fastidious and anaerobic urinary microbes that have no Clinical and Laboratory Standards Institute (CLSI) guidelines, including several in the families Aerococcaceae, Lactobacillaceae, or Actinomycetaceae. Previous studies using expanded quantitative urine culture reveal that urine samples from clinical patients are commonly polymicrobial in composition. Thus, we test whether NYCIII can serve as a viable harmonized medium, capable of supporting antibiotic susceptibility testing in a range of fastidious, non-fastidious, and anaerobic urinary microbes. We propose this methodology to be standardized comparable to CLSI standards to allow for resistance testing in uncharacterized urinary bacteria. IMPORTANCE Antibiotic susceptibilities of fastidious and anaerobic bacteria of the human urinary microbiome are largely underreported due to difficulty in growing them in the lab environment. The current standard medium, Muller-Hinton broth, has difficulty supporting the growth of many of these species, leaving microbiologists without a standardized method. To address this need, this study offers a methodology to survey susceptibilities in a high-throughput manner of these understudied microbes with a proposed harmonized medium, NYCIII, which is capable of supporting the growth of both fastidious and non-fastidious urinary microbes. Broader standardization of this method can allow for the development of antibiotic-resistant breakpoints of the many uncharacterized urinary microbes. Antibiotic susceptibilities of fastidious and anaerobic bacteria of the human urinary microbiome are largely underreported due to difficulty in growing them in the lab environment. The current standard medium, Muller-Hinton broth, has difficulty supporting the growth of many of these species, leaving microbiologists without a standardized method. To address this need, this study offers a methodology to survey susceptibilities in a high-throughput manner of these understudied microbes with a proposed harmonized medium, NYCIII, which is capable of supporting the growth of both fastidious and non-fastidious urinary microbes. Broader standardization of this method can allow for the development of antibiotic-resistant breakpoints of the many uncharacterized urinary microbes.

The standard urine culturing methods designed to detect the most common uropathogens do not support the growth of fastidious and anaerobic bacteria (7).The resulting bias has led to underreporting and thus understudy of such species.Many of these species are now considered to be "emerging" uropathogens, a recognition made possible by recent advances in metagenomics and metaculturomics technologies (8)(9)(10)(11).Such methods include those such as expanded quantitative urine culture (EQUC) coupled with matrix-assisted laser desorption/ionization-time of flight mass spectrome try (MALDI-TOF MS) (12)(13)(14).Unfortunately, apart from the limited number of species listed by CLSI in M45-ED3:2016, there exist few standards or guidelines for the evaluation of AST in uropathogens (4).
Only in 2016 CLSI first published guidance and clinical breakpoints for several fastidious urine microbes, such as Aerococcus, but it has yet to further add to or expand upon them.As the growth medium for broth microdilution, CLSI recommends cation-adjusted Mueller-Hinton broth supplemented with lysed horse blood (CAMHB-LHB) (2.5% to 5% vol/vol) with incubation at 35°C with 5% supplemented CO 2 for 24 to 48 hours (5).Similarly, EUCAST recommendations for Aerococcus is Mueller-Hinton broth supplemented with 5% lysed horse blood and 20 mg/L β-NAD (MH-F broth) with incubation at 35°C ± 1°C in ambient air for 18 ± 2 hours (6).The common theme is the supplementation of the standard medium with lysed horse blood, which has been previously utilized to improve the growth of other fastidious microbes, such as Corynebacterium diphtheriae and Haemophilus influenzae (15,16).Although effective in promoting growth, the use of blood renders the media opaque and thus prevents the measurement of bacterial growth via optical density.To conduct AST on many strains, we sought a medium that could (i) support the growth of fastidious and anaerobic microbes and (ii) allow direct monitoring of bacterial response to antibiotics.
In evaluating media, a significant candidate is New York City Broth III (NYCIII).This medium was originally developed to promote the growth and isolation of fastidious Neisseria species from clinical specimens (17,18).Although now fallen out of favor compared with the use of Thayer-Martin media, NYCIII is still remarkable for its transpar ency and, consequently, its ability to support turbidity measurements via spectrometry.This property is attributed to the use of horse serum instead of lysed horse blood.Taking advantage of this ability, one study found that NYCIII was the most ideal of nine different media compositions to promote the growth and measurement of bacterial vaginosisassociated anaerobes (19).Looking to extend these findings, we evaluated NYCIII for evaluation of AST in fastidious and anaerobic species of the urinary microbiome.Herein, we investigate the use of this methodology alongside CLSI standards and evaluate AST in previously untested urinary microbes.We also expand this method to include testing in co-cultures to evaluate interactions between species that may increase or decrease resistance to certain antibiotics when grown together.

Study isolates
Standardized reference strains for quality control (QC) were selected as outlined by CLSI and were procured from the American Type Culture Collection (ATCC) (4,5).Urinary isolates for resistance testing were procured from our IRB-approved biorepository (LU 215192).A complete list of strains used in this study can be found in Table S1.Each strain tested was grown from frozen stock, either purchased from ATCC or frozen in brucella media at −80°C.Strains were struck out on trypticase soy agar (TSA) with 5% sheep blood, and identity was confirmed via MALDI-TOF MS.

Antibiotic and media preparation
Preparations of stock antibiotics were as outlined in CLSI M100-ED33:2023.Preparation of CAMHB-LHB broth was according to CLSI instructions.NYCIII broth was prepared as outlined by ATCC medium #1685 with modification: 4.2 mL HEPES buffer solution, 7.5 g proteose peptone no. 3, 2.5 g glucose, 2.5 g NaCl, and 1.875 g yeast extract were added to 450 mL Millipore water before autoclaving at 121°C for 15 minutes.After cooling, 25 mL heat-deactivated horse serum and 25 mL heat-deactivated newborn calf serum were added via vacuum filter filtration (0.2 μm) to form the final 500 mL solution.For culturing anaerobic bacteria, the media were conditioned in an anaerobic chamber overnight.A comparison of the ingredients of CAMHB-LHB versus NYCIII can be found in Table 1.A complete list of brand components can be found in Table S2.To compare the growth of microbes between the two different media, single colonies struck on TSA plates were inoculated into 5 mL of test media and allowed to incubate for 48 hours (72 hours for anaerobic species).Afterward, cultures were evaluated for growth via colony-forming unit per milliliter (CFU/mL) enumeration.

Susceptibility testing
Strains were started in tubes with 5 mL of the test growth media until mid-log phase at 37°C with 5% supplemented CO 2 and shaking at 200 rpm.For the microbroth dilution method, 20 µL of this inoculum was added to 180 µL of fresh test media with or without antibiotics within the wells of a flat-bottom transparent 96-well plate.This plate was placed within a BioTek Epoch 2 Microplate Spectrophotometer, which incubated the bacteria at 37°C with 5% supplemented CO 2 and shaking at 200 rpm (double orbital).Absorbance was read every 15 minutes at 600 nm for 72 hours.Growth curves were analyzed to determine the minimal inhibitory concentration (MIC).For quality control testing, antibiotic dilutions were made via the standard twofold dilutions.For exploring susceptibility ranges in the fastidious and anaerobic urinary strains, antibiotic dilutions were conducted in 10-fold dilutions to capture a wider range of possible antibiotic susceptibility in unknown, undomesticated strains.All experiments were conducted in duplicate.

Comparison of standard media with NYCIII
Although CAMHB can support the growth of common uropathogens, it does a poor job of supporting the rarer fastidious urinary microbes (Fig. 1).For example, no growth is observed at all for the pathogen Streptococcus pneumoniae and the commensal Lactobacillus crispatus for this medium.Upon supplementation with lysed horse blood, the fastidious microbes have enhanced growth.Similar growth is maintained in NYCIII compared to CAMHB-LHB for most common uropathogens.Importantly, growth is strongly observed in NYCIII for the species that demonstrated no growth in CAMHB.

Validation of AST method
Furthering our evaluation of NYCIII, we next sought a methodology to conduct AST on a large number of isolate-antibiotic combinations utilizing spectrometry.This method is based on the twofold broth microdilution method as outlined by CLSI with modification to be conducted in a 96-well format.The output of bacterial growth curves when grown in the presence of increasing concentrations of antibiotic was analyzed to determine the MIC.This method was first applied to the non-fastidious microbes Staphylococcus aureus ATCC29213, Enterococcus faecalis ATCC29212, Escherichia coli ATCC25922, and Pseudomo nas aeruginosa ATCC27853 grown in CAMHB at 37°C in ambient air with shaking and the fastidious microbe S. pneumoniae ATCC49619 grown in CAMHB with 5% horse serum (vol/vol) at 37°C with 5% supplemented CO 2 with shaking.Horse serum was used over lysed horse blood to enable turbidity readings at 600 nm.Quality control MIC ranges according to CLSI standards are as outlined in Table S3.
With the QC MIC ranges as a reference, these microbes then were grown in their CLSI standard medium and compared to growth in the experimental NYCIII medium in the presence of antibiotics (Fig. 2; Table S4).Observed MIC values of reference microbes agreed for both media for all antibiotics tested with three exceptions.First, E. faecalis ATCC29212 exhibited susceptibility to ampicillin in the standard CAMHB condition at 0.25 µg/mL, below the 0.5 µg/mL QC range; however, the same strain behaved within the QC range when grown in NYCIII.Second, E. coli ATCC25922 exhibited increased resistance beyond the QC range for gentamicin when grown in NYCIII (>1 µg/mL).Similarly, S. aureus ATCC29213 exhibited increased resistance to ampicillin when grown in NYCIII (>2 µg/mL).

Determination of MIC of fastidious urinary isolates
Utilizing this method of AST, we evaluated the MICs of fastidious clinical isolates of the human urinary tract.We first selected species that were capable of growing in ambient air supplemented with 5% CO 2 .These included Actinomyces turicensis, Actinotignum timonense, Aerococcus urinae complex, Corynebacterium aurimucosum, Streptococcus agalactiae, Streptococcus anginosus, and two coagulases negative Staphylococci (CoNS):  Staphylococcus lugdunensis and Staphylococcus simulans (Table 2; Table S5).For each species, at least four different strains were tested in duplicate.

Evaluation of urinary anaerobes
Isolation and growth of anaerobic bacteria from the urinary tract are not common as the equipment to foster anaerobic environments is rarely found in clinical settings.However, their presence often can be detected via enhanced culture methods, such as EQUC, and metagenomic sequencing techniques, such as via 16S rRNA gene and shotgun metagenomic sequencing.Thus, we determined whether our methodology also could be used to evaluate AS of these microbes.These include Anaerococcus sp., Gardnerella vaginalis, Lactobacillus crispatus, Lactobacillus iners, Lactobacillus jensenii, Lactobacillus gasseri, Peptoniphilus harei, and Propionimicrobium lymphophilum.As a proof of concept, we evaluated susceptibility to tetracycline in an anaerobic environment with incubation for 72 hours (Table 3).

DISCUSSION
The human lower urinary tract features a variety of ecological niches that can foster the growth of a wide diversity of bacterial species.Simulating these diverse environmental conditions outside of the human body to study these bacteria remains one of the greatest challenges for microbial ecologists, let alone the average clinician.As such, using components that most resemble the native environment for these microbes while balancing accessibility and costs for the clinical lab is always an ongoing struggle.In this study, we evaluated the usage of NYCIII media, originally developed for isolation of Neisseria species, as a potential harmonized medium to test AST in urinary isolates.While there exists a variety of tests specifically developed to evaluate AST in certain species, no generalized testing medium exists that can evaluate non-fastidious, fastidious, and anaerobic bacteria of the human urinary tract.The major advantage of such a harmonized medium would be the ability to conduct a direct comparison of values between species without having to account for differences in media or methodol ogies.Furthermore, a harmonized medium can allow for polymicrobial AST as it has been demonstrated that isolates of a polymicrobial culture behave differently together than when tested in isolation (21).
While MHB remains the standard for most bacterial AST as recommended by CLSI and EUCAST guidelines, the medium does not support the growth of many fastidious bacteria and most anaerobic bacteria of the urinary tract.Only with supplementation with lysed horse blood is growth rescued.However, the blood renders the medium opaque preventing turbidity measurements of bacterial growth via spectrometry.Thus, we evaluated NYCIII, a transparent medium using horse serum instead of horse blood, to evaluate AST.From our results, we demonstrate that NYCIII can support the growth  S3. of both fastidious and non-fastidious bacteria to levels similar to that of CAMHB-LHB.When testing AS ranges for QC standard strains, NYCIII can meet similar values, indicating direct translatability between standard MHB and NYCIII MIC values.Thus, NYCIII may be a feasible medium to conduct AST in a 96-well format via plate-reader spectrometry.We propose further testing for this medium with the goal toward standardization, such that AST guidelines may be established for the many emerging uropathogens.Indeed, with this medium, many urinary strains can be simultaneously compared at once to evaluate relative susceptibility.While screening selected specimens, we were able to identify several isolates that exhibited susceptibilities beyond the tested range, potentially indicating resistance.It is important to note that no resistance cutoffs have been established for these bacteria and, thus, the prevalence of resistance remains unknown.
As long as these emerging uropathogens remain uncharacterized, physicians will continue to prescribe antibiotics without guidance.The implications of these are several-fold including prolonged therapies, over-prescription, and poor antibiotic stewardship.Accordingly, urinary microbes that are not under active surveillance are potentially harboring antibiotic resistances.Even non-pathogenic bacteria can serve as reservoirs of resistance for pathogenic bacteria, such as through horizontal gene transfer (22).As such, establishing AST patterns in newly identified bacteria, and especially for emerging uropathogens, may be easier to conduct with our proposed harmonized methodology.Additionally, it may be of value to conduct AST of commensals, particu larly bacteria species considered protective against disease, to establish their vulnerabil ity to antibiotics.
As for limitations, only a small number of strains were tested for each species in this study, and thus, these MICs should not be used as reference values for clinical purposes.Rather, values found in this study should be interpreted as an evaluation of the feasibility of employing NYCIII as a valid AST methodology such that more in-depth studies may be conducted in the future.Furthermore, it may be important to take MICs of sulfonamides and trimethoprim when using NYCIII with a grain of salt.Media that contain relatively high concentrations of thymidine are known to interfere with the inhibitory effect of sulfonamides and trimethoprim (23).However, media that use horse blood have an Actinotignum timonense 5 5 Aerococcus urinae complex b 5 1 4 Corynebacterium aurimucosum 5 5

Streptococcus anginosus 6 6
a "No." refers to the number of isolates tested per species.
b Aerococcus urinae has recently been revealed to be a complex of at least four species: A. urinae, A. tenax, A. mictus, and A. loyolae (20).The type strain of each species was tested here.
advantage of containing thymidine phosphorylase borne from lysed horse erythrocytes (24).This enzyme converts thymidine to inactive thymine, allowing both antibiotics to work uninhibited.Since NYCIII replaces lysed horse blood with horse serum, this advantage has been lost.

Conclusion
In this study, we evaluate the use of a proposed harmonized medium, NYCIII, in conducting AST of clinical urinary isolates.The medium offers the ability to conduct AST in a 96-well format, allowing for the rapid screening of relative AST of both fastidious and anaerobic urinary bacteria.Further standardization of this medium may greatly improve the evaluation of resistance of emerging uropathogens in the clinical setting.

FIG 2
FIG 2 Validation of NYCIII as a growth medium for AST evaluation.MICs were determined for test microbes in both the standard medium and NYCIII and compared.If both test conditions yielded MICs that were within QC range, then the cell is shaded green.If MICs were outside QC range, then the cell is shaded red.For actual MIC values, see TableS3.

TABLE 2
Summarized susceptibility ranges of selected fastidious urinary bacteria in NYCIII

TABLE 2
Summarized susceptibility ranges of selected fastidious urinary bacteria in NYCIII (Continued)