Molecular Approaches to the Side Effects of Immunosuppressive Drugs

Mycophenolate 1mofetil (MMF) is an important therapeutic immunosuppressive agent in clinical practice (1). It has been shown to decrease the rate of acute rejection in renal and liver transplantation compared with Azathioprine. Also it may have a role in decreasing the progression and graft loss due to chronic allograft nephropathy. Similar to the calcineurin inhibitors it is recognized as a drug with a narrow therapeutic window. However, in contrast to the calcineurin inhibitors, therapeutic drug monitoring is not in routine use, despite some evidence that dose adjustment following estimation of drug levels can be associated with improved clinical outcomes (2). The active component off MMF is mycophenolic acid (MPA), which is derived rapidly from MMF by plasma and tissue esterases. MPA is metabolized in the liver by glucuronidation. The major metabolic product is the acyl glucuronide acyl phenolic glucuronide (MPAG). There are also four minor metabolites. It is unclear whether MPAG plays a role in the side effect profile of MPA, but that is possible, as it is excreted into the bile and subsequently undergoes enterohepatic circulation. The immunosuppressive effect induced by MMF is achieved by an inhibition of lymphocyte proliferation via the inhibition of inosine monophosphate dehydrogenase (IMPDH), the source of de novo guanosine synthesis from inosine and adenosine. T and B cells are particularly prone to this inhibition as they are unable to significantly generate guanosine from the scavenger pathway. This pathway is present in other cells and thus MMF was expected to be “lymphoselective” in its antiproliferative effects. However, MPA does have other effects. It is know to inhibit fibroblast/smooth muscle cell proliferation, an effect that may be important in a long-term role in the modulation of chronic allograft rejection. It also has an inhibitory effect on cellular glycosylation pathways and this leads to a reduction in adhesion molecule expression, another pathway for immunosuppression. The side effect profile of MPA includes leukopenia (expected via its mode of action), diarrhea in up to 20% of patients, anemia, and mild hepatotoxicity. These later effects could be simply explained by overdosing and the failure of the salvage pathway to keep with the supply of guanosine in these tissues. However, the article by Shipkova et al. in this issue suggests it may not be that simple (3). In doing so the authors provide insights into new mechanisms of MMF toxicity and show us that novel approaches to the problem of drug toxicity should be encouraged. The article by Shipkova et al. uses a rat model of MMF toxicity. They clearly show effects on the intestine, particularly the colon, anemia, and minor effects on the liver. The authors approach to this model was to use gene microarrays to identify altered gene expression in the livers of animals with toxicity versus controls. They then selected genes and examined expression by real-time polymerase chain reaction (PCR) in the liver, jejunum, ileum, and colon. Using this approach they identify four interesting genes that could be related to anemia, intestinal, and hepatic side effects. These findings include down-regulation of hemoglobin α, which may explain anemia (the presence of hemoglobin α transcripts is explained by hematological tissue contamination in liver and intestinal specimens). More interesting is the finding of down-regulation of the poly immunoglobulin (Ig) receptor, catalase, and CCAAT/enhancer protein alpha (C/EBP-α). The poly Ig receptor is an important molecule involved in the transport of IgA and IgM to the mucosal surface and its decreased expression may have lead to decreased IgA/IgM in the bowel lumen (although this was not measured in the article). The authors speculate that such a decrease could lead to the predisposition to mucosal gut damage via various mechanisms. The decrease in catalase, an antioxidative stress molecule, may also predispose the gut to oxidative stress and subsequent injury. C/EBP-α is an important, highly expressed transcription factor in the gut and liver. It is important in the control of cellular proliferation in epithelial tissues. Thus, a decrease in expression may once again predispose to MMF intestinal toxicity. Finally, the most significant differential expression in the microarrays was the master gene regulator HNF4. This molecule unfortunately was not selected for further study by the authors. HNF4 has been recently shown to be master regulator of hepatic and pancreatic gene expression and to be essential for the development of these organs (4). The observation that this is the most down-regulated gene (18-fold) by MMF is intriguing. The authors go on to use this gene expression data, together with real-time PCR confirmation, to provide some correlations with biochemical markers of toxicity. These data are however not that convincing and underline that this study lacks causative conclusions. Nevertheless, the approach of the authors to the examination of drug toxicity is novel and they are to be congratulated and encouraged. The insights they provide clearly are not proven but point the way to further study. One of the spinoffs of gene microarray studies is that it provides the researcher with unexpected (or nonhypothesis driven) data. In the experiments by Shipkova et al. they were looking for, and interpreting data, in terms of drug toxicity. However, two molecules that were observed to be down-regulated (syndecan and vascular endothelial growth factor [VEGF]) may be important in the potential beneficial response to MMF. As mentioned earlier MMF may inhibit fibroblast/smooth muscle cell proliferation and decrease extracellular matrix production. Both syndecan and VEGF are up-regulated in the fibrotic response, and VEGF has been recently shown to play a key role in the hepatic fibrotic response (5). These effects may be important in the potentially inhibitory effect of MMF on chronic allograft rejection. Before finishing a discussion of the molecular approach to drug toxicity it is pertinent to mention the emerging field of pharmacogenomics (2). MPA is metabolized by the uridine diphosphate glucuronyltransferase (UDP-GT) system, which has many polymorphisms and its main target of inhibition IMPDH has two isoforms. It has been shown that there are polymorphisms in UDGPT 1A8, one of the enzymes responsible for MMF metabolism, and it is known that blood concentrations of MPA vary widely amongst individuals. This raises the question whether drug toxicity (and efficacy) can be predicted in individuals using molecular genotyping. Such approaches, together with those outlined in the article by Shipkova et al., have the potential to better understand and predict the fine line between drug toxicities and efficacy at the molecular level.