Proteomic pearl diving versus systems biology in cell physiology. Focus on “Proteomic mapping of proteins released during necrosis and apoptosis from cultured neonatal cardiac myocytes”

in the early days of the 21st century, the completion of genome sequencing projects for multiple species provided a bounty of new information for physiological investigations. The availability of comprehensive genome sequence data has also made possible new large-scale approaches to the study of biology that are particularly promising for cell physiology and, therefore, of particular interest to readers of this journal. These methods include DNA microarrays, deep sequencing techniques, and large-scale proteomics using protein mass spectrometry. Since proteins are responsible for most cellular functions, proteomics seems particularly appropriate for studies of cell physiology. Recognizing this, the Editors of AJP-Cell Physiology initiated a Call for Papers that use mass spectrometry methodology for studies of cell physiology. This call resulted in a series of excellent publications that includes metabolomics and DNA array studies in addition to proteomics (2, 3, 5, 7, 8). The latest in this series is the article by Marshall et al. (4) in this issue. This report uses protein mass spectrometry to identify proteins released from cultured neonatal cardiac myocytes in models of cell death.

While no one would debate the ability of proteomics (and other large-scale methods) to generate large amounts of data that are relevant to important physiological questions, there is considerable skepticism among cell physiologists about the utility of such data. This skepticism exists largely because of the perceived absence of adequate methods for finding relevant information in large data sets and relating it to existing knowledge. Many (if not most) cell physiologists are most comfortable exploring the function of their favorite cells one gene or protein at a time, taking advantage of methods that manipulate the abundance of a given protein in a cell or mutating one or more amino acids. Based on this reductionist perspective, when faced with a study presenting large data sets such as that of Marshall et al. (4), investigators may be inclined to search the expanse of reported data and try to identify a single protein or a few proteins that warrant further study. Or, reviewers may ask authors to “pick a particular protein from the data set and study it further.” Such an approach, viewed as proteomic “pearl diving,” is often effective when it identifies a key protein previously not recognized to play a role in the physiological process under consideration. However, doing so has a large down-side, namely, by relegating most of the results to a virtual data-dumpster, it wastes most of the information generated. Thus, the major challenge in physiological proteomics is the development and utilization of methods that consider and employ all of the information in large-scale data sets by identifying patterns in the data, relating these patterns to existing knowledge about cellular function. This objective subsumes much of what is commonly termed “systems biology.” Both the pearl diving approach and the systems biology approach are “discovery” modalities. They differ in that the former discovers interesting proteins for further study, while the latter discovers co-regulated sets of proteins that point to particular cell biological processes that may be involved in the physiological response.

The bulk of the paper by Marshall and colleagues examines models of necrosis in primary cultures of rat neonatal cardiac myocytes. They use state-of-the-art protein mass spectrometry to identify proteins released into the medium in response to H₂O₂-mediated induction of necrosis and apply both analytical approaches described above, viz. proteomic pearl diving and systems biology, to interpret their results. From their pearl diving, the authors found a number of known markers of necrosis such as lactate dehydrogenase (LDH) and the DNA-binding protein Hmgb1. In addition, a number of other interesting proteins were enriched in the media relative to controls, such as α-actinin, HSP90, and Trim72. The last is a muscle-specific protein that plays an important role in membrane repair. The proteins found are considered potential biomarkers of myocyte necrosis, information that could be of considerable use in two ways, viz. possible clinical applications and use in future studies to understand mechanisms of cardiac injury such as seen in ischemia-reperfusion injury. Thus, proteomic pearl diving proved successful. Antibodies recognizing several of the marker proteins discovered by the authors were utilized to confirm a limited set of observations from the proteomic studies. Proteomics investigators have recently suggested that antibody-based assays are inferior to targeted mass spectrometry for confirmation studies, in part because many antibodies are far less specific than is protein mass spectrometry (1). Marshall et al. implicitly provide a counterargument to this suggestion, showing that when reliable antibodies are available, they can be used for cost-effective extensions of a study. In the case of the current paper, the authors used immunoblotting to show that the same marker proteins appear in the medium after induction of necrosis by a different agent, β-lapachone instead of H₂O₂, thus supporting the idea that many of the proteins found are necrosis markers rather than redox markers.

The authors also employed a systems biology approach, identifying several pathways that contain significantly greater numbers of necrosis-associated proteins than would be expected by a chance selection from a control data set. This type of analysis has a major advantage in the physiological context in that it identifies cellular processes rather than individual proteins whose connection to the physiology may be obscure. In addition, the systems biology approach tends to be statistically more robust: False positives are less likely since the process identified depends on quantification of multiple proteins versus the individual proteins identified with pearl diving. An example from Marshall et al. is their top-ranked pathway (Table 1), viz. “aryl hydrocarbon receptor signaling.” The discovery of this pathway points to a possible role in cardiac necrosis for a set of very interesting basic helix-loop helix transcription factors that are already known to play roles in blood pressure regulation and cardiac physiology (9) as well as redox regulation (6). Although implication of aryl hydrocarbon receptor signaling is hypothetical at this point, it is easy to imagine experiments to test its role in the cellular decision process between the life and death of cardiac myocytes.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author.

AUTHOR CONTRIBUTIONS

M.A.K. drafted manuscript; edited and revised manuscript; approved final version of manuscript.