Characterization of a Polyethylene Glycol-Amphotericin B Conjugate Loaded with Free AMB for Improved Antifungal Efficacy

Abstract

Amphotericin B (AMB) is a highly hydrophobic antifungal, whose use is limited by its toxicity and poor solubility. To improve its solubility, AMB was reacted with a functionalized polyethylene glycol (PEG), yielding soluble complex AmB-PEG formulations that theoretically comprise of chemically conjugated AMB-PEG and free AMB that is physically associated with the conjugate. Reverse-phase chromatography and size exclusion chromatography methods using HPLC were developed to separate conjugated AMB-PEG and free AmB, enabling the further characterization of these formulations. Using HPLC and dynamic light scattering analyses, it was observed that the AMB-PEG 2 formulation, having a higher molar ratio of 2 AMB: 1 PEG, possesses more free AMB and has relatively larger particle diameters compared to the AMB-PEG 1 formulation, that consists of 1 AMB: 1 PEG. The identity of the conjugate was also verified using mass spectrometry. AMB-PEG 2 demonstrates improved antifungal efficacy relative to AMB-PEG 1, without a concurrent increase in in vitro toxicity to mammalian cells, implying that the additional loading of free AMB in the AMB-PEG formulation can potentially increase its therapeutic index. Compared to unconjugated AMB, AMB-PEG formulations are less toxic to mammalian cells in vitro, even though their MIC50 values are comparatively higher in a variety of fungal strains tested. Our in vitro results suggest that AMB-PEG 2 formulations are two times less toxic than unconjugated AMB with antifungal efficacy on Candida albicans and Cryptococcus neoformans.

Fig 1

(A) AMB-PEG reaction scheme. The NHS ester on MS(PEG)₄ reacts with the primary amine (–NH2) group on AMB at a 1:1 molar ratio, generating conjugated AMB-PEG via amide bond formation. (B) Solubility of AMB-PEG at varying AMB:PEG molar ratios. AMB-PEG mixtures and unreacted AMB were prepared in DMSO, incubated for 2 hours and then (i) dispersed to a final concentration of 2 mM in PBS-EDTA, containing 10% DMSO by volume. (ii) The mixtures were then centrifuged to facilitate the observation of insoluble precipitates. Higher molar ratios yielded a suspension of yellow particles that precipitated upon centrifugation, similar to what was observed with unconjugated AMB.

Fig 2. Aqueous solubility of AMB-PEG 1 and 2.

Increasing amounts of AMB-PEG formulations were added to PBS-EDTA, and post-centrifugation, the concentration of soluble AMB-PEG in the supernatants quantified based on their absorbance at 365 nm. Error bars denote the standard deviation between 3 independent formulations. Dotted line represents the theoretical condition where AMB formulation is completely soluble (i.e. concentration of total AMB = concentration of soluble AMB).

Fig 3. Representative absorbance spectra of AMB-PEG 2 and unconjugated AMB formulations prepared in buffers with varying hydrophobicity.

20 mM AMB-PEG 2 and unconjugated AMB formulations were prepared in DMSO and resuspended in 20% and 48% ACN buffers containing 4.3% acetic acid, as well as PBS-EDTA, to a final concentration of 2 mM AMB. As buffer hydrophobicity increases with higher ACN concentrations, the A₃₄₈/A₄₀₉ ratio decreases, implying that AMB-PEG is increasingly in its monomeric form. As AMB-PEG 1 and 2 have similar UV-visible absorption profiles, with identical peak height ratios in all buffers tested, data for AMB-PEG 1 is not shown. AMB-PEG formulations that have been subjected to buffer exchange to PBS-EDTA through a 10 kDa centrifugal filter have the same UV-visible absorption spectra as the initial formulation of AMB-PEG in PBS-EDTA, which contains 10% DMSO.

Fig 4

(A) Reverse phase chromatogram of AMB-PEG and unconjugated AMB, with eluted peaks detected at 406 nm. 50 μL of AMB-PEG 1 and 2, and 10 μL of unconjugated AMB dispersed in PBS-EDTA and 48% ACN respectively were injected into a C18 reverse phase column and eluted isocratically in a 48% ACN buffer at a flow rate of 0.5 ml/min for 40 minutes. Peaks were detected at 406 nm. The AMB-PEG conjugate has a shorter retention time, implying that it is more hydrophilic. From the AMB-PEG samples, AMB-PEG conjugate and free AMB (based on the retention time of unconjugated AMB) fractions were collected for further analysis via size exclusion chromatography. (B) Size exclusion chromatogram of AMB-PEG 2 and unconjugated AMB, as well as the relevant fractions collected from RPC, in a 20% ACN mobile phase. 20 mM of AMB-PEG 2 formulations and unconjugated AMB in DMSO were prepared and resuspended at 2 mM in a 20% ACN buffer. 50 μL of these samples (unconjugated AMB diluted tenfold prior to analysis) and the peak fractions collected previously from RPC were passed through a size exclusion column and eluted peaks were detected at 406 nm. Unconjugated AMB was eluted later compared to the AMB-PEG conjugate, implying that it has a smaller hydrodynamic volume compared to AMB-PEG in 20% ACN. The previously collected AMB-PEG conjugate and free AMB peak fractions had similar retention times as the AMB-PEG 2 formulation and unconjugated AMB samples respectively, thus verifying their respective peak identities. AMB-PEG 1 and 2 have identical elution profiles. (C) Size exclusion chromatogram of AMB-PEG and unconjugated AMB dispersed in PBS-EDTA. 3 μL of 2 mM AMB-PEG formulations that have been retained by 10 kDa centrifugal filters (Millipore) and 50 μL of the supernatant obtained from unconjugated AMB were analysed using a Superdex 75 size exclusion column and eluted peaks detected at 406 nm. In a PBS-EDTA mobile phase, unconjugated AMB has a shorter retention time of 20 minutes compared to AMB-PEG at 40 minutes, implying that AMB-PEG has a smaller hydrodynamic volume under these experimental conditions.

Fig 5. MALDI-TOF mass spectrum of the AMB-PEG fraction from RPC.

The major mass peaks observed had masses corresponding to that of the AMB-PEG conjugate, thereby verifying the identity of that peak fraction. AMB-PEG mass peaks were absent from the collected unconjugated AMB fraction.

Fig 6. LIVE/DEAD staining of HEK293 and IMR-90 cells after exposure to AMB-PEG 1, 2 and unconjugated AMB for 24 hours.

Live cells are stained green and dead cells stained red. AMB-PEG did not cause cell death at concentrations of 139 μM in HEK293 cells and 277 μM in IMR-90 cells. The molar ratio of AMB to PEG did not have any visible effect on cell toxicity. Conversely, unconjugated AMB caused extensive cell death at concentrations above 4.33 μM in both cell lines. Experiment was performed twice, each time with three independently prepared AMB-PEG formulations.