Proteomics reveal the breadth and limits of model systems inferences. Focus on "proteomic analysis of V-ATPase-rich cells harvested from the kidney and epididymis by fluorescence-activated cell sorting".

PHYSIOLOGISTS OF MANY GENERATIONS have employed tractable model systems to make observations and then have extrapolated those observations to other more complex or less amenable systems. Such an approach is built on initial obvious similarities of the systems in form or function. The success of this philosophical approach has been mixed. Depending on the tissue, species, and especially depending on the function being assessed, there has been a variety in the levels of success. Thus, one can easily be suspect of such comparisons. Over the past two decades many reports have highlighted the similarities or hypothesized similarities between intercalated cells of the nephron and narrow or clear cells of the epididymis (2). The rationale for the comparison is the knowledge that both organs are derived from the same embryonic tissue, that both cell types are known to contain a substantial number of mitochondria, and that both cells types have a major role in secreting H into their respective ducts. The nephron regulates systemic pH by secreting H or base equivalents, and the epididymis generates an acid luminal environment that promotes sperm maturation and allows for sperm storage. Routine techniques have revealed substantial overlap between these forkheadrelated (FORE) cells, a class of cells that also includes interdental cells of the inner ear, in which Foxi1 is required for the expression of specific vacuolar (V) H -ATPase subunits A1, B1, E2, and a4 (7). Each of these cell types reportedly expresses, in a regulated manner, V-type H -ATPase in its apical membrane, where it is thought to be required for H secretion. Other well-documented similarities between these cells include the expression of NBC3 and carbonic anhydrase 2 (4, 6). Surprisingly, however, interdental cells appear to secrete little acid (Miyazaki H, Marcus DC, and Wangemann P, personal communication). Nonetheless, based on the success of identifying similarities between these cell types thus far, a number of laboratories are continuing with lines of investigation built on the supposition that numerous additional parallels exist between these cell types in the kidney and the male reproductive duct. The cell isolation and proteomic approach taken by Da Silva et al. (3) seems to break new ground in the correlative study arena. The tack of hypothesizing functional correlates and then searching for individual proteins or regulatory cascades is an inefficient strategy to define commonalities and differences between cells. The new approach described by the authors is philosophically simple, but technically challenging—isolate pure populations of FORE and non-FORE cells from both the kidney and the epididymis followed by proteomic analysis of those cells. Expression of enhanced green fluorescent protein driven by the ATPV1B1 promoter was used to “tag” FORE cells for sorting (5). (Interdental cells in the mouse inner ear were also “tagged” by this approach, although the cells were not analyzed in the current report.) A simple 2 2 factorial design allows for comparisons of proteins identified in FORE and non-FORE cells within a tissue and for comparisons between FORE cells across tissues. The tacit assumption is that this proteomic approach will reveal areas of broad similarity and of mechanistic identity between these cell populations. This approach has produced a bonanza of data for which the cell biology community should be thankful. Perhaps the most important outcome is the identification of 1,564 and 2,297 proteins expressed in epididymal and renal FORE cells, respectively (full lists are available at the National Heart, Lung, and Blood Institute Laboratory of Kidney and Electrolyte Metabolism website). The authors provide an excellent initial analysis of the components, functions, and processes represented by the proteins that were identified. As expected, this analysis reveals substantial overlap in the characteristics of proteins identified in these two cell types. However, a closer inspection and parsing of the data reveal quite illuminating insights (Fig. 1). First, 875 of the identified proteins are present in FORE cells in both tissues, which accounts for roughly one half to one third of the proteins identified. Thus, the complement of proteins that makes each of these cell types unique, one half to two thirds, is quite substantial. Clearly, a sizable repository of data is available to be mined or dissected to reveal unique capabilities and unique regulation of the FORE cells derived from each tissue. A second goal addressed by the authors was to determine the identity of proteins that were present in greater