Erbb3 and Hspa2, two novel predictors of heart failure in diabetic patients

This editorial refers to ‘The diabetic myocardial transcriptome reveals Erbb3 and Hspa2 as a novel biomarkers of incident heart failure’, by M.S. Conning-Rowland et al., https://doi.org/10.1093/cvr/cvae181. Heart failure (HF) is a major health problem worldwide, with a very high morbidity and mortality and a remarkable impact on national health systems. The incidence of HF dramatically increases in patients with diabetes, as compared to the general population.1,2 Diabetes is also associated with a worse HF-related prognosis, further aggravating the already high socio-economic burden associated with this disease. Therefore, there is an urgent need to better understand the molecular mechanisms associated with HF development and progression, particularly in patients with diabetes, in order to discover new prognostic predictors for improving risk stratification. Recent advancements in omics technologies and the availability of large biobank cohorts represent a great help in filling the gaps in knowledge. In this issue of Cardiovascular Research, Conning-Rowland et al.3 provide novel insights into the characterization of novel and potential biomarkers of incident HF associated with diabetes (Figure 1). By analysing RNA sequencing data of the right atrial (RA) and left ventricle (LV) tissues from donors of the Genotype-Tissue Expression (GTEx) project, the authors discovered 64 differentially expressed genes (DEGs) associated with diabetes. Among DEGs, the differential expression of epidermal growth factor receptor B3 (Erbb3) and Hspa2 found in LV but not in RA was validated by proteomics in the plasma of more than 50 000 participants from the UK Biobank (UKB) cohort. HSPA2 shows a higher expression in diabetic patients as compared to non-diabetic patients, whereas ERBB3 has a lower expression. Of interest, the expression of Erbb3 and Hspa2 is not altered in non-cardiac tissues, suggesting the cardiac origin of these circulating biomarkers. In order to explore the clinical relevance of these findings with cardiovascular outcomes, the authors further demonstrate that low plasma levels of ERBB3 or high plasma levels of HSPA2 correlate with increased HF incidence, cardiovascular mortality, and myocardial dysfunction. These correlations remain associated also after adjustment for confounding factors, such as age, sex, body mass index (BMI), systolic blood pressure (SBP), and diabetes mellitus (DM), with the exception of NT-pro B-type natriuretic peptide (BNP) levels. Among DEGs associated with diabetes in the RA and LV, HSPA2 and ERBB3 were validated in the plasma of UKB cohort. Higher HSPA2 or lower ERBB3 plasma levels are independent predictors of incident HF, cardiovascular death, and cardiac dysfunction. Elements were modified from Servier Medical Art, licensed under a Creative Commons Attribution 3.0 Generic License (https://smart.servier.com/; https://creativecommons.org/licenses/by/3.0/). Moreover, by dividing UKB participants into quartiles, according to different plasma levels of ERBB3 and HSPA2, the authors found that the lowest plasma levels of ERBB3 are associated with older age, male sex, lower BMI, higher prevalence of DM, prior myocardial infarction (MI) and HF, and with a large use of medications for cardiovascular diseases and RBB3 diabetes. On the other hand, the highest levels of HSPA2 levels are associated with aging, female sex, higher BMI and prevalence of DM, prior MI and HF, and with use of medications for cardiovascular diseases and diabetes. By using cardiovascular magnetic resonance imaging data from UKB, the authors further observed that lower levels of ERBB3 or higher levels of HSPA2 are associated with impaired LV contractility and higher LV mass. Finally, lower ERBB3 or higher HSPA2 circulating levels are associated with clinically relevant circulating biomarkers of cardiometabolic status such as higher NT-proBNP, troponin I, SBP, urinary microalbumin, and HbA1c. Overall, Conning-Rowland et al.3 discovered new biomarkers for the management of patients with or at high risk of developing HF. The large population that was analysed in this work and the combined unbiased approach with transcriptomics and proteomics represents the strength of the analyses, which provided robust and conceptually novel data. However, there are several important issues that remain to be addressed. Among the 64 DEGs associated with diabetes, plasma proteomic analysis from UKB was available only for the 12 of them. Future studies with a similar unbiased approach are needed to explore the biological significance of the other DEGs discovered and not validated by proteomics in this study. In order to improve the diagnostic/prognostic value of Erbb3 and Hspa2 for the development of precision medicine strategies, a further stratification between type 1 and type 2 diabetes, as well as between HF with reduced ejection fraction and HF with preserved ejection fraction, should be performed. When adjusted for NT-proBNP, the association between HF incidence with plasma levels of ERBB3 and HSPA2 is partially lost. In addition, ERBB3 and HSPA2 levels do not provide additive prediction value towards HF incidence with respect to the prediction ability of NT-proBNP alone. However, ERBB3 and HSPA2 levels were still found to able to predict cardiovascular mortality in patients with high baseline NT-proBNP. For these reasons, further studies in different populations are warranted to understand whether ERBB3 and HSPA2 may be useful for cardiovascular risk stratification and for the prediction of HF in the clinical setting, independently of the currently used biomarkers. The exact cut-off values of ERBB3 and HSPA2 should be further characterized to improve their diagnostic and prognostic value. The possible direct role of ERBB3 and HSPA2 in the development of heart abnormalities should also be characterized in future experimental studies, particularly in the context of diabetes. The molecular mechanisms regulated in cardiac cells by ERBB3 and HSPA2 should also be elucidated with the help of preclinical experiments performed in models of diabetes and HF with gain or loss of function studies for Erbb3 and Hspa2. The latter approach would help to find therapeutic strategies able to modulate these signalling in patients at high risk of developing HF, especially in the presence of diabetes. ERBB3 belongs to the ERBB family of receptor tyrosine kinases. Previous studies demonstrated that Erbb3 is expressed in cardiac myocytes and cardiac endothelial cells and plays a fundamental role in cardiac valve development.4 Several studies also demonstrated that low serum levels of neuroregulin (NRG)-1β, which stimulate ERBB receptors, correlate with the worst prognosis in patients with chronic HF.5 In line with this evidence, the administration of NRG-1 was reported to rescue cardiac function in animal models of HF and patients with HF.4,6 The protective effects of NRG-1 are attributable to the improvement of multiple mechanisms involved in cardiovascular biology, such as cardiomyocyte survival, adaptive hypertrophy, angiogenesis, and vasodilatation.4,7 Unfortunately, Conning-Rowland et al.3 did not analyse the plasma concentration of NRG-1, since this marker is not available in the UKB cohort. Interestingly, more data are available regarding the role of ERBB2, another ERBB isoform, in the heart. In particular, ERBB2 was previously found to protect the heart in response to stress and to promote cardiac regeneration, partly through the activation of the co-transcription factor YAP, a major effector of the Hippo pathway.8 It will be interesting to check whether ERBB3 also acts through similar mechanisms. ERBB signalling is also a target of cancer therapies used for the treatment of different forms of cancer. Unfortunately, these therapies usually exert cardiotoxic effects. For example, trastuzumab, an anti-ERBB2 antibody, given alone or in combination with anthracyclines, was reported to cause HF and severe LV systolic dysfunction in patients with breast cancer.9 It would be interesting to evaluate whether the assessment of circulating levels of ERBB3 may represent a tool to monitor cardiotoxicity induced by agents targeting ERBB signalling. Of note, previous work showed that the administration of an engineered bivalent NRG reduces doxorubicin-induced cardiotoxicity in a preclinical mouse model, without exerting the pro-neoplastic effects of NRG.10 These results, together with the findings of Conning-Rowland et al., may suggest that in addition to patients undergoing chemotherapy, the activation of NRG/ERBB signalling may be a potentially useful future approach for the reduction of incident HF in diabetic patients. Little is known about the role of HSPA2 in cardiomyocytes, although another study found its upregulation in failing human hearts, as compared to non-failing hearts.11 HSPA2 belongs to the heat shock proteins of the A family of 70 kDa, well-known chaperones that usually act as stress-responsive elements and activate protein quality control mechanisms. The inhibition of HSP70, another heat shock protein, was reported to attenuate cardiac hypertrophy in mice,12 through the reduction of the activity of histone deacetylase 2 (HDAC2). HSPA2 plays well-defined effects during spermatogenesis,13 but its specific role in somatic cells remains unknown. Some evidence demonstrated that HSPA2 mediates lipid accumulation in vitro and polymorphisms in the Hspa2 gene were found to be associated with obesity, DM, and risk for carotid plaque rupture and cerebral ischaemia in DM.14–16 Future studies are warranted to assess the role of HSPA2 in preclinical models of HF and the underlying molecular mechanisms involved in its upregulation in the presence of diabetes. The interaction between HSPA2 and other proteins in the heart should also be investigated by interactomics analysis. Despite these open questions, the study of Conning-Rowland et al.11 provides important translational perspectives. Additional studies should be performed to characterize the potential role of ERBB3 and HSPA2 in the myocardium, particularly in the setting of diabetes, for the development of a therapeutic approach that would be able to restore plasma levels of ERBB3 and HSPA2 in patients at higher cardiovascular risk and see whether HF incidence is effectively reduced. This work was partially supported by a grant from the Italian Ministry of Health (RF-2021-12375256) and by a grant from the European Union—Next Generation EU—PNRR M6C2—Investment 2.1 Enhancement and strengthening of biomedical research in the NHS (PNRR-MAD-2022-12376632 and F63C22000830006) to S.S. Not applicable.