Defining the minimum inhibitory concentration of 22 rifamycins in iron limited, physiologic medium against Acinetobacter baumannii, Escherichia coli, and Klebsiella pneumoniae clinical isolates

Recently, we reported rifabutin hyper-activity against Acinetobacter baumannii. We sought to characterize if any additional rifamycins (n = 22) would also display hyper-activity when tested in iron-limited media against A. baumannii, K. pneumoniae, and E. coli. MICs were determined against representative clinical isolates using the iron-limited media RPMI-1640. Only rifabutin was hyperactive against A. baumannii.


Introduction
We previously conducted a modified compound screen assay against Acinetobacter baumannii using RPMI-1640 broth, a medium that better models the in vivo physiologic blood environment as compared to the rich mediumMueller Hinton II broth, cation-adjusted (MHII)). We found that rifabutin (RBT) possesses previously unrecognized hyper-activity against A. baumannii, but only in physiological media, and not in rich media [1][2][3]. RBT is able to rapidly enter the cell through the FhuE protein, an iron transport protein that is downregulated when the bacteria are replete with iron [1,4]. We have previously tested rifabutin, rifampin, rifaximin, rifapentine, FCE-22250, rifalazil, and rifamycin sv against A. baumannii and only rifabutin was able to traffic through the FhuE protein in iron limited conditions [1]. Here, we sought to determine if any additional, commercially available rifamycins would have increased activity in iron limited conditions when tested against representative isolates of A. baumannii, Klebsiella pneumoniae, and Escherichia coli.

CGP-4832
Previous studies have shown that the rifamycin CGP-4832 is able to traffic through the E. coli iron transport protein FhuA [7][8][9]. CGP-4832 MICs against E. coli MG1655 were 8 mg/L and 0.25 mg/L in MHII and RPMI-1640 respectively. This 32-fold shift in MIC is consistent with the described mechanism of action for CGP-4832 against E. coli. As this mechanism is conceptually similar to the mechanism described for RBT entry, we tested if A. baumannii is hyper susceptible to CGP-4832. However, no benefit was observed for A. baumannii. CGP-4832 MICs against A. baumannii HUMC1 and LAC-4 were 2 mg/L and 4 mg/L respectively in MHII; and 8 mg/L and 2 mg/L respectively in RPMI-1640. It was previously estimated that about 33% of clinical isolates were hypersusceptible to CGP-4832 [7]. However, this number was determined using only MHII for susceptibility testing because it was not yet known that transport of CGP-4832 through the FhuA protein depended on an iron-limited environment. CGP-4832 MICs were determined against a panel of 45 E. coli clinical isolates, obtained from the CDC and FDA AR Isolate Bank, in MHII or RPMI-1640 to test if a greater proportion of clinical isolates would be hypersusceptible to CGP-4832 if MICs were determined in iron-limited media. There was a significant difference in CGP-4832 and RBT MICs determined in MHII (p = 0.001, Mann-Whitney), but no significant difference when MICs were determined in RPMI-1640. Additionally, there was not a statistically significant difference between the distribution of GCP-4832 MICs determined in the iron-rich MHII as compared to the iron-limited RPMI-1640 media (Fig 2, S1 Table). This result was somewhat surprising because sideromycin antibiotics such as albomycin, also dependent on transport by FhuA, displayed increased activity in iron-limited media [10].

Conclusions
The use of culture media that mimics the in vivo physiologic environment during infection is a valuable tool for characterizing antibiotic activity in vitro. Rifabutin and CGP-4832 are the only rifamycin compounds that are able to traffic through the FhuE protein of A. baumanii or the FhuA protein of E. coli respectively. We have previously shown that rifabutin is not able to enter E. coli through FhuA of E. coli [2]. The library of compounds tested represents broad chemical diversity and will further guide the development of additional derivatives capable of being transported more broadly by both FhuA and FhuE proteins of diverse bacterial species.

MIC assay
Unless otherwise indicated, the standard broth microdilution method following CLSI methodology was used to determine MICs [5]. The media used for the minimum inhibitory concentration (MIC) assays performed in this study were MHII or RPMI-1640 as previously described [1][2][3]6,11,12]. Briefly, 100 μL of media was added to the wells in columns 2-10. Column 11 served as a positive growth control and contained only bacteria and media. Column 12 served as the

PLOS ONE
sterility control and contained only culture media without bacteria. Next, 200 μl of a 2X antibiotic working solution was added to the wells in column 1. Two-fold serial dilutions of the antibiotic were performed through column 10. Next, 100 μl of a 1×10 6 CFU/mL working solution of bacteria were added to each of the wells in columns 1-11. The inoculum concentration was confirmed by plating serial dilutions on TSA plates. MIC plates were incubated at 35±2˚C and results were recorded at 24 hours.
Supporting information S1