We appreciate the comments from Dr. Alisi and colleagues about our recent published article.1 We agree that genomic and proteomic technologies are extremely useful in studying a complex disease such as nonalcoholic fatty liver disease (NAFLD) and its progressive subtype of nonalcoholic steatohepatitis.2–6 Although we have previously used gene expression of liver tissue as well as omental adipose tissue to study NAFLD, our recent approach has been to use proteomic technology to study this important disease.1, 3, 5, 6 Although using proteomic technology such as surface-enhanced laser desorption/ionization (SELDI) provides valuable data, it does require additional purification and identification of protein peaks,4 and cannot directly detect the low-abundance cell signaling proteins. By using a new protein microarray technology, we can now effectively explore the intracellular signaling pathways contained within select tissue specimens from patients with NAFLD.1
Cellular signal transduction is largely a posttranslational, phosphorylation-driven event, underpinned by rapid enymatically catalyzed events of kinases and their specific substrates. Because these kinases constitute most of the drug targets for targeted therapies, their characterization is of intense scientific interest. Importantly, because these cellular networks are protein-based, and because past studies have shown little to no concordance of protein expression with gene expression,7, 8 analysis of cellular signaling cannot be inferred from gene expression studies and requires new approaches. This type of analysis requires a direct functional protein-based assay, and with the advent of phosphospecific antibodies, a direct nonenzymatic assay for protein phosphorylation has become routine. In our study of NAFLD, we employed a new type of protein microarray technology, called the reverse-phase protein microarray (RPPA), which has been invented for global multiplexed protein network profiling.1 Other types of protein and phosphoprotein measurement technologies, such as immunohistochemistry/tissue microarray or suspension bead arrays (for example, Luminex), are either nonquantitative and limited to measuring only a few analytes at a time (the former) or limited by sensitivity and a paucity of well-performing 2-site antibody pairs (the latter). RPPAs have been developed such that from as little as a few thousand cells, one can quantify hundreds of proteins at once and the test requires only one specfic antibody to work. An inherent quality of the RPPAs is the excellent intra-assay and interassay reproducibility, precision, and accuracy of the method. Routinely, between-run coefficients of variation of less than 7% are achieved from as few as 500 cells, with less than a cell equivalent of protein printed per spot.9, 10 A potential limitation of the RPPA is the inability to provide cellular localization context (for example, nuclear staining). However, insofar as cell signaling analysis is concerned, cellular localization such as membrane or nuclear localization shows excellent correlation with the phosphorylation of the analyte, thus this information can largely be gleaned by phosphorylation quantitation.
Of course, protein signaling analysis does not provide information about which genes are expressed, and when. Thus, a full understanding of the complexity of the biology of NAFLD requires technologies and approaches which can illuminate the entire cascade of molecular events that govern cellular function. Although the DNA is the information archive, it is the proteins that do the work of the cell, and the utility and promise of these new proteomic technologies in NAFLD are that they can directly measure the activity of drug targets and potentially provide both diagnostic and therapeutic information at the same time. However, these and any other results need to be extensively validated in larger clinical study sets of patients with NAFLD before clinical implementation. These studies are ongoing.