Need for new approachestechnology

The application of high-throughput global genomic and proteomic techniques has resulted in the identification of large numbers of potential targets but few sufficiently validated leads worthy of further development. Apart from the technical problems described in the last section, other related factors negatively affect the ultimate success of the global approach. We are now in the ironic position of going from the relatively piecemeal process of discovery that slowly generated too little data to having the opposite problem with so much more data being generated than can be rapidly and accurately validated. A decade ago, a typical scientist may spend his whole career discovering and studying a single molecule. Although genome-based technologies can provide a catalog of expressed genes (i.e., identify all mRNAs present in a given cell), this information is not always helpful in determining the level of actual protein expression, localization, and accessibility ofpotential targets. The human genome is estimated to encode 40,000 genes, but the pool of actual targets is far greater in number. At the protein level, this pool expands to several hundred thousand possible targets because of extensive RNA processing, posttranslational modification, and the formation of functional protein complexes (34). This greatly complicates new research efforts using proteomic analysis to identify differentially expressed proteins that might have utility as drug targets.

This overwhelming abundance of potential targets also relates, in part, to the molecular complexity of solid tumors, which limits the ability of global genomic and proteomic analysis to discover key targets in vivo. Often, the tissue samples used as the starting material for analysis are highly heterogeneous, consisting of many different cell populations. For example, solid tumors consist of not only the tumor cells themselves but also stromal cells, perivascular and endothelial cells, inflammatory cells, and, in some cases, adjacent normal tissue cells. Unfortunately, the clinical utility of most of the potential targets on any of these tumor constituents is limited because they reside in compartments that are not readily accessible to the circulating blood. Thus, global survey techniques applied to whole tumors might not be the best approach to identifying useful targets. It could instead be more promising to focus on meaningful subsets of cell types or parts of cells to enrich the starting samples in bona fide targetable molecules prior to applying high-throughput, global identification strategies.

Genomic and proteomic analysis of normal and diseased tissue and cell samples have already yielded thousands of genes and gene products as candidate targets for diagnostic and tissue assignment as well as potential therapeutic targeting (5-8,34,35). The development bottleneck occurs when each of these thousands of potential targets must undergo a laborious validation process to evaluate its specificity for the cell type in question and, importantly, but often overlooked at least initially, the accessibility of the target to the targeting drug (biologic composite such as immunotoxin). It is becoming clear that the shear number of candidates discovered via high-throughput, global analytical techniques is overwhelming the required but time-consuming in vivo validation process. These issues have led some to question the ultimate impact these global genomic- and proteomic-based analytical technologies will have on speeding up target discovery (6,7,34).

0 0

Post a comment