Clinical isolates of M. catarrhalis (n = 14) were prepared as previously described [19] . The isolates used were kindly provided by J. Tjhio (Loyola University Stritch School of Medicine, Maywood, IL). All isolates tested elaborated a beta-lactamase by nitrocefin assay (data not shown).

All commercial media were prepared per label directions. Supplements were prepared individually, and in combination (as indicated) then filter sterilized (Table 1). For maximal level of achievable growth and generation time, each medium was inoculated with M. catarrhalis to a final concentration of ~10⁵ CFU/ml then incubated (35˚C) aerobically and anaerobically (Whitley Anaerobic Workstation A35). After incubation (48 h), samples were removed for standard viability count (Muller Hinton: MH agar; aerobic conditions). The optimal medium for anaerobic growth was determined after testing five media alone and with seven different supplements, in various combinations.

^aAll media were tested with all combinations of supplements; all supplements were made in yeast nitrogen base medium without amino acids or ammonium chloride, pH 7.0, Difco; ^bMuller Hinton and Brain Heart Infusion with supplements A-E (BHI-S) were also tested aerobically with and without NaNO₃; no significant differences in final absorbance_600nm or CFU/ml in MH and BHI-S media with and without nitrate were measured; ^cMinus sign (-) indicates no growth; plus sign (++) indicates ~1-3 × 10⁹ CFU/ml; (+) indicates ~1-2 × 10⁸ CFU/ml; ^dBovine Serum Albumin fraction IV; ^eIndicates combination of indicated supplements, e.g. A-C = A, B, C; ^fMedium used for anaerobic growth of M. catarrhalis for all subsequent experiments; ^gNot Determined.

The MIC and MBC for antimicrobials used in the treatment of M. catarrhalis otitis media, pneumonia and sinusitis were measured using a microdilution method [14] [20] . Each drug was tested in sextuplicate and repeated once (M. catarrhalis, n = 8; 35˚C, 48 h; 5% CO₂ in air; normal atmospheric conditions; or anaerobic conditions) [19] .

The effect sub-MICs have on M. catarrhalis biofilm formation was determined, as previously described, by the simple expedient of emptying, washing and staining the 96 well plate wells after growth in the presence and absence of antibiotics (PBS, 3x wash; crystal violet, stain; absolute ethanol elutant) [21] . Biofilm levels were determined as a measure of crystal violet staining (Beckman EIA reader; Abs₅₄₀). The measurements were compared against a control 96 well plate that contained only drug (negative control) or only organisms (positive control). Significance testing was determined by ANOVA, with Tukey post hoc test (GraphPad Software, San Diego, CA).

Previous studies report the growth of laboratory-adapted M. catarrhalis in brain heart infusion medium (BHI) with 10 mM NaNO₃ [22] [23] . This medium did not support the growth of any M. catarrhalis clinical isolates tested (n = 14). This was probably because clinical isolates can exhibit growth characteristics that are different from laboratory-adapted strains [24] . Therefore, a requisite first step focused on development of a medium that supported the anaerobic growth of M. catarrhalis clinical isolates (n = 14) to levels similar to that measured for the isolates’ aerobic growth in Mueller Hinton broth, the standard medium used for antibiotic testing (Table 1). Of the media tested, only BHI with supplements A-E (BHI-S) or A-G consistently supported growth of all isolates; the level of anaerobic and aerobic growth was similar to growth measured in MH incubated aerobically (1 - 3 × 10⁹ CFU/ml). Middlebrook medium with supplements A-E or A-G also supported the anaerobic growth of M. catarrhalis, but to a lesser extent (~8% - 13% less) than that measured for BHI-S. BHI-S was used for all subsequent determinations of the effects aerobic, capnophilic and anaerobic growth conditions have on antimicrobial susceptibility and biofilm formation.

The anaerobic MIC for all antimicrobials tested ranged from significantly less to within 2 fold dilution of that measured for aerobic and capnophilic conditions (Table 2). Anaerobically grown M. catarrhalis’ MICs in the presence of clarithromycin, and the quinolones nalidixic acid and ciprofloxacin were the most sensitive as compared to aerobic/capnophilic growth (31-250; 508-1016; 8-31 fold less MIC, respectively). However, the MBC range for these drugs was within 2 fold dilution for all growth conditions. This pattern of the MBC ranges overlapping regardless of growth conditions vs. the MIC range registering significantly less under anaerobic growth occurred for all compounds with the exception of ceftriaxone where the MBC was 4 fold less than that measured under atmospheric conditions and 2 fold below capnophilic conditions. No discernible pattern of susceptibility was noted for the individual isolates with the exception of M. catarrhalis isolate 6 which exhibited the highest MIC and MBC to macrolide-azide drugs clarithromycin and azithromycin under all growth conditions.

Biofilm formation in positive controls (organism alone) was unaffected by the lack of oxygen or increased concentration of CO₂ (Table 2). Of the antimicrobials tested, sub-MIC levels of azithromycin and the quinolones, with the exception of levofloxacin, significantly (p < 0.05) affected biofilm formation. Nalidixic acid exhibited a bimodal effect on biofilm formation, inhibiting biofilm formation under anaerobic growth conditions while pro- moting biofilm levels under capnophilic and aerobic conditions. Norfloxacin and ofloxacin increased the biofilm levels under aerobic and capnophilic conditions only. Ciprofloxacin only affected biofilm levels under anaerobic growth (1.8 fold above positive growth control). Azithromycin promoted biofilm formation for one isolate (iso-

¹Growth conditions: Aero = aerobic-atmospheric oxygen; CO₂ = 5% CO₂ in air; Anaero = anaerobic. ²Peak biofilm Formation-Maximum level of biofilm measured across sub-MIC antibiotic concentrations tested. ³Maximum [Drug] affecting Biofilm: The maximum concentration of antimicrobial causing a significant (p < 0.05) alteration in biofilm formation. ⁴NE = No effect. ⁵Maximum level of biofilm measured across sub-MIC antibiotic concentrations tested (Abs₅₄₀)/Maximum level of biofilm measured for positive antibiotic-free control (Abs₅₄₀) where biofilm levels were significantly (p < 0.05) affected. ⁶Penicillin class. ⁷Cephalosporin class. ⁸Macrolide class. ⁹Quinolone class.

late 3; 2.21 fold increase above positive control) to the greatest extent under 5% CO₂ in air. The response of the isolates was highly variable with none responding to more than one of the compounds; isolate 9 biofilm forma- tion was unaffected, regardless of growth condition.