Pharmacokinetics of the Individual Major Components of Polymyxin B

Jan 12, 2017 - Pharmacokinetics of the Individual Major Components of Polymyxin B and Colistin in Rats. Sivashangarie ... Citation data is made availa...
0 downloads 4 Views 875KB Size
Note pubs.acs.org/jnp

Pharmacokinetics of the Individual Major Components of Polymyxin B and Colistin in Rats Sivashangarie Sivanesan,† Kade Roberts,†,§ Jiping Wang,†,§ Soon-Ee Chea,† Philip E. Thompson,‡ Jian Li,§ Roger L. Nation,† and Tony Velkov*,† †

Drug Delivery, Disposition and Dynamics and ‡Medicinal Chemistry, Monash Institute of Pharmaceutical Sciences, Monash University, 381 Royal Parade, Parkville, Victoria 3052, Australia § Monash Biomedicine Discovery Institute, Department of Microbiology, Monash University, Clayton, Victoria 3800, Australia ABSTRACT: The pharmacokinetics of polymyxin B1, polymyxin B2, colistin A, and colistin B were investigated in a rat model following intravenous administration (0.8 mg/kg) of each individual component. Plasma and urine concentrations were determined by LC-MS/MS, and plasma protein binding was measured by ultracentrifugation. Total and unbound pharmacokinetic parameters for each component were calculated using noncompartmental analysis. All of the polymyxin components had a similar clearance, volume of distribution, elimination half-life, and urinary recovery. The area under the concentration−time curve for polymyxins B1 and B2 was greater than those of colistins A and B. Colistin A (56.6 ± 9.25%) and colistin B (41.7 ± 12.4%) displayed lower plasma protein binding in rat plasma compared to polymyxin B1 (82.3 ± 4.30%) and polymyxin B2 (68.4 ± 3.50%). These differences in plasma protein binding potentially equate to significant differences in unbound pharmacokinetics, highlighting the need for more stringent standardization of the composition of commercial products currently available for clinical use.

I

nonyl). This structural difference is mirrored between colistins A and B. The individual polymyxin components have been shown to have different levels of plasma protein binding, so it follows that compositional variations between batches could potentially impact pharmacokinetics, pharmacodynamics, and toxicodynamics, especially given that it is the unbound (free) drug that is responsible for antibacterial activity.2,8,9 What is cause for concern in polymyxin formulations is that even within the limits imposed by the European and British Pharmacopeias the proportion of different components of both polymyxins often fluctuates between different manufacturers and between different batches.6,7,10 The main goal of this study was to compare the unbound pharmacokinetics of isolated pure polymyxins B1 and B2 and colistins A and B that were for the first time individually administered intravenously in a rat model. The presented findings have important implications for the use of polymyxin B and colistin methanesulfonate in the clinic. Polymyxin B1, polymyxin B2, colistin A, and colistin B display closely related chemical structures (Figure 1). Minor changes in chemical structure can potentially affect the pharmacological properties of these lipopeptide components. Tam et al.11 previously reported the pharmacokinetics of polymyxins B1, B2, B1-Ile,7 and B3 in rats following the administration of a single intravenous dose of 4 mg/kg polymyxin B sulfate. The first

n recent times the world has seen the emergence of Gramnegative bacteria that are resistant to almost all available antibiotics.1 Faced with a lack of therapeutic options, this has led to the rebirth of polymyxins as the last line of defense against infections caused by extremely drug resistant (XDR) superbugs.2,3 Currently, only polymyxin B and colistin methanesulfonate (CMS), the inactive prodrug of colistin, also known as polymyxin E, are used clinically.4 As polymyxins are prepared via fermentation, the pharmaceutical preparations of polymyxin B and colistin are mixtures of closely related lipopeptides, namely, polymyxin B1 and B2 and colistin A and B. The pharmacopoeia limits for batches of polymyxin B are ≥80% for the sum of polymyxins B1, B2, B1−Ile,7 and B3, with the last two components representing no more than 6% and 15%, respectively. For colistin, the pharmacopoeia limits for batches are ≥77% colistins A and B, together with three minor components (polymyxin E3, polymyxin E7 (7-methyloctanyl), E1-Ile7), which must each be 95%), isolated polymyxin B1, polymyxin B2, colistin A, and colistin B that were administered individually. The pharmacokinetics data revealed that concentrations of all of the polymyxin components declined to ∼0.2 mg/L over 360 226

DOI: 10.1021/acs.jnatprod.6b01176 J. Nat. Prod. 2017, 80, 225−229

Journal of Natural Products

Note

the clearance values determined in the current study [colistin A (2.75 ± 0.54 mL/min/kg) and colistin B (3.05 ± 0.62 mL/ min/kg)]. Pharmacokinetic parameters reported in the aforementioned studies highlight the disparity in results when polymyxin B and colistin are administered as mixtures as compared to individually administered polymyxin components, implicating potential interactions between the major components in vivo. A low urinary recovery for each component was observed, as expected, given that both polymyxin B and colistin are cleared via nonrenal routes of elimination and undergo extensive tubular reabsorption.12,13 Interestingly, when comparing the major components of polymyxin B and colistin with each other, there is no significant difference between most pharmacokinetic parameters (Table 1). However, there is a significant difference (P = 0.04) in the volume of distribution between the major components of polymyxin B and colistin. The volume of distribution is affected by the extent of plasma protein binding; notably the most dramatic differences between the polymyxin components were seen with the plasma protein binding. Colistin A (56.6 ± 9.25%) and colistin B (41.7 ± 12.4%) displayed lower plasma protein binding in rat plasma compared to polymyxin B1 (82.3 ± 4.30%) and polymyxin B2 (68.4 ± 3.50%). This finding could be construed in terms of the only structural difference between these components, namely, the D-phenylalanine residue at position 6 of polymyxin B1 and polymyxin B2 versus the D-leucine residue at position 6 of colistins A and B (Figure 1). First, although both of these amino acids display hydrophobic side chains, their structures are very distinct; phenylalanine possesses a benzyl side chain, whereas leucine displays an aliphatic isobutyl side chain. Second, besides the aromaticity, the higher ranking of phenylalanine on the amino acid hydrophobicity scale potentially confers the polymyxin B scaffold with higher plasma protein binding properties.14 The second notable aspect of the plasma protein binding data was that polymyxin B1 and colistin A displayed significantly higher plasma protein binding compared to polymyxin B2 and colistin B, respectively (Table 1 and Figure 2). Again this difference could be afforded by the increased hydrophobicity at the N-terminus of polymyxin B1 and colistin A (S-6-methyloctanoyl) due to the extra methylene group, compared to the one-carbon-shorter N-terminus of polymyxin B2 and colistin B (6-methylheptanonyl) (Figure 1). These results are in contrast to the rat protein binding data for polymyxin B by Tam et al.,15 which was reported to be as high as 92%. Here ultrafiltration was employed, which can result in nonspecific binding of the lipopeptides to the membrane of the apparatus.16 The protein binding results of colistins A and B herein are comparable with the protein binding values reported by Li et al.9 Although the major components of polymyxin B

min (Figure 2). The pharmacokinetic parameters for polymyxin B1 polymyxin B2, colistin A, and colistin B are documented in

Figure 2. Plasma concentration vs time profile after a single intravenous dose of 0.8 mg/kg of (A) polymyxins B1 and B2 and (B) colistins A and B. Data points are the mean ± SD (n = 4).

Table 1. There was no significant (P > 0.05) difference between the major components of polymyxin B across the pharmacokinetic parameters of total clearance, volume of distribution and the elimination half-life percent, area under the concentration− time curve (AUC), and the urinary recovery (Table 1). In the Tam et al.11 study, the reported clearance of polymyxin B1 (1.65 ± 0.62 mL/min/kg) and the combined clearance of polymyxins B2 and B3 (1.69 ± 0.52 mL/min/kg) were somewhat lower than the clearance values determined in this study [polymyxin B1 (2.39 ± 0.62 mL/min/kg) and polymyxin B2 (1.96 ± 0.29 mL/min/kg)]. There were no significant differences in pharmacokinetic parameters with the major components of colistin (Table 1). In the study by Li et al.,9 study clearances for colistin A (5.1 ± 0.5 mL/min/kg) and colistin B (5.3 ± 0.5 mL/min/kg) were almost twice as high as

Table 1. Pharmacokinetic Parameters for Total Plasma Concentrations of Polymyxin B1, Polymyxin B2, Colistin A, and Colistin B Following a Single i.v. Bolus Dose of 0.8 mg/kg (n = 4 per Group) individual polymyxin component pharmacokinetic parameter Clearance (mL/min/kg) Volume of distribution (L/kg) Elimination half-life (min) AUC0‑∞ (mg·min/L) Urinary recovery of unchanged dose (%) Plasma protein binding (%)

polymyxin B1 2.39 0.21 79.5 365 0.13

± ± ± ± ±

0.75 0.06 10.4 103 0.10

82.3 ± 4.30

polymyxin B2 1.96 0.23 94.9 416 0.09

± ± ± ± ±

0.29 0.04 7.3 63.0 0.04

68.4 ± 3.50

colistin A 2.75 0.27 82.0 301 0.14

± ± ± ± ±

0.54 0.04 30.9 56.0 0.07

56.6 ± 9.25 227

colistin B

P value B1 and B2

P value colistin A and B

P value (all)

± ± ± ± ±

0.39 0.80 0.08 0.49 0.73

0.55 0.19 0.68 0.66 0.26

0.18 0.04 0.54 0.15

3.05 0.37 91.3 276 0.24

0.62 0.11 20.5 72.7 0.06

41.7 ± 12.4 DOI: 10.1021/acs.jnatprod.6b01176 J. Nat. Prod. 2017, 80, 225−229

Journal of Natural Products

Note

isolated metabolic cages in a temperature- (21 ± 3 °C) and humiditycontrolled facility with a 12 h light−dark cycle (06:00−18:00) and acclimatized for 2 days. All animals received food and water ad libitum. Plasma Protein Binding. The plasma protein binding of the individual polymyxins was determined by ultracentrifugation with healthy drug-free rat plasma.28,29 Aliquots (200 μL) of each plasma sample were spiked with 5 mg/L of each individual polymyxin component. Plasma samples were incubated for 30 min at 37 °C. Samples were then ultracentrifuged at 215000g for 4 h at 37 °C. From the triplicate spiked plasma samples, two samples were subjected to buffer fraction analysis in which 50 μL of protein-free supernatant was removed. The final sample from the triplicate was subjected to complete plasma analysis in which 50 μL of the total resuspended content was removed. The polymyxin concentration in the supernatant and the resuspended samples was determined by LC-MS/MS.20 The percentage bound was calculated via the following equation:

and colistin have comparable total pharmacokinetic profiles, there are noticeable differences in the unbound pharmacokinetic profiles (Figure 2). The most significant differences can be seen in the unbound pharmacokinetic profile of B1 when compared to polymyxin B2. In our recent in vivo study, we did not observe significant differences in the MICs and in vivo antibacterial activity among isolated polymyxins B1 and B2 and colistins A and B.17 Notably, we also found that the commercial polymyxin B and colistin products were slightly less active than their respective individual components in vivo, suggesting that the minor lipopeptide components (>20% of the total content) may be less active in vivo.17 Nephrotoxicity remains a dose-limiting factor with intravenous polymyxin therapy.3,18−23 We reported that polymyxin B1 (0.25 mM) and colistin A (0.75 mM) produce a >3-fold greater apoptotic effect in an in vitro human kidney proximal tubular HK-2 cell culture model than polymyxin B2 (0.25 mM) and colistin B (0.75 mM), respectively.17 Furthermore, we showed that polymyxins B1 and B2 were significantly more potent inducers of apoptosis than colistin A or B. This suggests that structure nephrotoxicity relationships exist between colistin versus polymyxin B (i.e., D-leucine to D-phenylalanine substitution at position 6) and polymyxin B1/colistin A versus polymyxin B2/colistin B (S-6-methyloctanoyl versus 6-methylheptanonyl).17 Interestingly, no differences in the nephrotoxicity (in terms of histological damage) of the major components of polymyxin B and colistin (accumulated dose of 72 mg base/kg) were observed in our mouse nephrotoxicity model.17 This disparity between in vitro cell culture and animal nephrotoxicity models previously noted by our group and others may be due to kidney regeneration or potentially result from a scenario where the transporter-mediated uptake of the individual polymyxin components into the kidney proximal tubular cells differs significantly.24,25 Overall, the findings of this study indicate that the major components of polymyxin B and colistin when administered individually can behave differently compared to when administered as mixtures, as per the current clinical formulations. Another layer of complexity that we must also remain cognizant of is the fact that colistin is solely administered as the prodrug colistin methanesulfonate, which means the release of colistins A and B can be highly variable. The results of this study highlight the necessity to move toward more stringent standardization of the composition of commercial products currently available for clinical use or opt toward the development of novel, safer, single-component polymyxin lipopeptide products with well-defined pharmacokinetics.



% Plasma protein bound = [(Average concentration of total resuspension) − (Average concentration of protein − free supernatant)] /(Average concentration of total resuspension) × 100 Pharmacokinetics. Sixteen rats were cannulated in the jugular vein for polymyxin administration and in the carotid artery for blood collection. Following cannulation, animals were allowed to recover for a day. Rats were divided into four groups (n = 4, in each group) and received a 0.8 mg/kg bolus dose of a polymyxin via the jugular vein cannula. Blood (200 μL) was collected prior to the administration of the individual polymyxin components and at 10, 20, 30, 60, 90, 120, 180, 240, and 360 min thereafter and centrifuged immediately (10000g, 10 min) at 4 °C. Urine was collected in chilling chambers from the metabolic cage prior to administration of the individual components and over the intervals of 0−6 h and 6−24 h thereafter. For the plasma and urine assay, calibration curves were constructed with concentrations of individual polymyxin components ranging from 0.1 to 8.0 mg/L. The intraday accuracy and reproducibility of individual polymyxin components were