Author + information
- Maggie P.Y. Lam, PhD,
- Vidya Venkatraman, MSc,
- Quan Cao, MD, PhD,
- Ding Wang, PhD,
- T. Umut Dincer, MSc,
- Edward Lau, PhD,
- Andrew I. Su, PhD,
- Yi Xing, PhD,
- Junbo Ge, MD,
- Peipei Ping, PhD∗ ( and )
- Jennifer E. Van Eyk, PhD† ()
- ↵∗National Institutes of Health BD2K Center of Excellence for Biomedical Computing at UCLA, National Heart, Lung, and Blood Institute Proteomics Center at UCLA, David Geffen School of Medicine at UCLA, University of California, Los Angeles, 675 Charles East Young Drive, MRL Building, Suite 1-619, Los Angeles, California 90095
- ↵†Advanced Clinical Biosystems Research Institute, Cedars-Sinai Medical Center, Advanced Health Sciences Pavilion, 127 South San Vicente Boulevard, Los Angeles, California 90048
The human genome consists of more than 20,000 genes; about one-half have been found in microarray experiments to be expressed in the heart. We asked which of the ∼10,000 cardiac proteins are most studied in and are thus essential to cardiac research, for which, perhaps surprisingly, no objective measures yet existed. This question is relevant to clinical investigations, basic research, and biomarker discovery, because high-quality quantification assays are available for relatively few proteins. Selected reaction monitoring (SRM) is a targeted mass spectrometry technique that allows the quantification of virtually any user-specified proteins with high sensitivity (∼20 amol) and specificity (1). SRM assays have important advantages over conventional immunoblots or enzyme-linked immunosorbent assays in their precision, throughput, and ability to target post-translational modifications, but adoption rates in basic/clinical research remain poor. It is thought that challenges for method developers to strategize development to prioritize important and popular proteins contribute to a bottleneck in technology dissemination (2).
To assess protein importance objectively, we performed a large-scale bibliometric analysis of the 24 million research papers curated on PubMed. We used the search terms “heart [MeSH term/All fields] or cardiac [All fields]” to retrieve cardiac-related articles, then tallied the occurrences of each protein being referenced to the retrieved papers on NCBI Gene using a new software tool (BD2KPubMed). In total, we retrieved ∼1.4 million cardiac-related papers referenced to 8,325 distinct human proteins. Table 1 lists the top 15 cardiac proteins with the most publications.
The data underline an intense focus of the majority of cardiac research on relatively few proteins. The 50 most-studied proteins accounted for 19% of all referenced cardiac publications, averaging 150 studies each. Publications declined precipitously after the 50th protein, with the next 50 proteins having one-third as many publications, whereas 84% of the investigated proteins had ≤5 publications. The most-studied human protein is natriuretic peptide B (BNP/NPPB), a routinely utilized clinical marker for evaluation and risk stratification of heart failure patients, which alone accounts for 1.3% of all publications referenced to cardiac proteins. Following BNP are angiotensin-converting enzyme and NaV1.5 (sodium channel protein type 5), accounting for 1.0% and 0.9% of referenced publications, respectively.
Intriguingly, there exists a marked discrepancy between the most studied proteins in humans and in mice. Among the top 50 mouse and human proteins, only 17 are shared. BNP ranks 61st in mouse studies, which instead featured homeobox protein NKX2-5/tinman as the top protein. NKX2-5 is a master regulator of cardiac differentiation/development that accounts for 1% of publications in mice but ranks 29th in humans. We suggest that this discrepancy reflects different priorities between basic and clinical investigations. Many human proteins have clinical values prior to further reports on mechanisms of action. The top 20 human proteins alone contained 9 secreted proteins that are known disease biomarkers, versus 4 in mice. The popularity of BNP increased following its adoption as a clinical biomarker of heart failure circa 2003. In contrast, top mouse proteins include signal transducers in developmental pathways that have not been sufficiently translated to humans. Gene ontology analysis corroborates functional differences: the top 50 human proteins are significantly enriched in regulation of transport (p < 4.7 × 10−21) and contractility (p < 8.3 × 10−18), whereas the top 50 mouse proteins are enriched in heart morphogenesis (p < 2.1 × 10−35) and cardiac muscle development (p < 4.2 × 10−29).
The measurement of protein abundance is essential to biomedical research, where the availability of high-quality protein quantification tools can dictate the pace of discovery (3). There are currently few experimentally verified SRM assays, and most available assays await further validation in relevant cohorts. We present here a resource to comprehend research trends, and a list of high-priority proteins that merit expedited assay development and clinical translation. These most-studied proteins provide opportunities for developers to target community interests in cardiovascular research, but complementary metrics may also be employed (e.g., hub proteins in biological networks). Developmental priority may further be influenced by the scale required in clinical applications (50 to 5,000+ samples), which often lies beyond the domain of basic research (10 to 100 samples). To accelerate translation, we suggest that more concerted basic clinical collaborations to optimize sample and data sharing pipelines are in order.
Users interested in identifying high-impact proteins in other topics can download BD2KPubMed and additional methodology information (4).
Please note: This work was supported by National Institutes of Health grants NIH U54-GM114833-01 and HHSN268201000035C (to Dr. Ping) and HHSN268201000032C (to Dr. Van Eyk). Dr. Su has served as a consultant for Avera McKennan. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation