The Immunosuppressive Drug Mycophenolic Acid Does Not Bind Iron(II) Under Conditions Mimicking the Upper Gastrointestinal Environment

Abstract

Anemia is common in renal transplant patients, especially pediatric recipients, which makes iron supplementation usual in therapeutic regimens (1). The potential of iron to impede absorption of the immunosuppressive drug mycophenolic acid (MPA) has been noted. A trial by Morii et al. (2) showed a remarkable 89.7% reduction in bioavailable MPA when the prodrug mycophenolate mofetil (MMF) was co-administered with oral iron supplements in healthy adult volunteers. As a result of the Morii trial (2), a number of studies in both healthy subjects and renal transplant patients were undertaken, yet all have failed to show iron influencing MPA availability (3–5). The rapid hydrolysis of MMF converts the ester to the carboxylic acid, MPA, which is capable of metal binding. Indeed, we have shown that MPA can form a stable complex with copper (6). Morii et al. (2) hypothesized that reduction in MPA availability was due to iron binding to MMF, however as our results with copper binding indicated, it is more likely that MPA itself is the potential iron ligand. The later trials tend to suggest MPA is either not coordinating iron or, if it does bind iron, then a stable complex is not formed. In this letter we have investigated if MPA-iron complexes are likely to occur in the gastrointestinal tract after concomitant administration of MMF and iron supplements. We used ultraviolet/visible spectroscopy to investigate the binding of both iron(II) and iron(III) to MPA when in a simple solvent (methanol) and compared that to the situation when in simulated gastric acid (sGA) at pH 6.1 (7). Figure 1A shows the spectrum obtained after the addition of Fe2+ to MPA (both 1 mM) dissolved in sGA. In pure methanol, a small peak was observed near 500 nm (Fig. 1A), however in sGA no such peak was observed and the solution remained colorless. The peak at 500 nm in methanol is suggestive of some Fe2+ binding to MPA (the peak is due to ligand-to-metal charge-transfer or d-d transitions). Analysis by mass spectrometry provided evidence for the formation of a small amount of [Fe(MPA)2]+, where two MPA molecules act as ligands for the Fe2+ ion (data not shown). The low intensity of the peak associated with this complex suggested that the complex is not formed to any great extent or is not stable. Figure 1B shows that the addition of Fe3+ to MPA (both 1 mM) in methanol results in an intense 590-nm peak, suggesting that MPA is able to coordinate Fe3+. This intense peak is not observed when Fe3+ and MPA are in sGA, only a very weak absorbance is apparent (the addition of Fe3+ to sGA alone had no effect). Thus there is only very limited coordination of Fe3+ by MPA in sGA.FIGURE 1.: Binding of iron to MPA in simulated gastric acid (sGA). (A) MPA (1 mM) and Fe2+ (1 mM) were mixed in sGA and the UV/Vis spectrum obtained. For comparison the spectrum obtained from MPA (1 mM) and Fe2+ (1 mM) in methanol (MeOH) is shown. (B) MPA (1 mM) and Fe3+ (1 mM) were mixed in sGA and the UV/Vis spectrum obtained. The spectrum obtained for MPA (1 mM) and Fe3+ (1 mM) in methanol is shown (black line). The dashed line shows the addition of Fe3+ to sGA in the absence of MPA.In conclusion, although we observed evidence for iron/MPA complexes when prepared in methanol, we were unable to see similar effects when we combined MPA and iron in sGA. The lack of MPA-iron(II) complexes in sGA supports the trials that show no effect on MPA availability when MMF and iron supplements are coadministered (3–5). Alison C. Badrick Christopher E. Jones Centre for Metals in Biology School of Molecular and Microbial Sciences The University of Queensland St. Lucia, Queensland