The generation of MHC-TRAF2_LC mice was previously described (9). Briefly, animals were generated with α-myosin heavy chain promoter targeting murine TRAF2 and characterization of the mice was done using histological analysis at 12 weeks (9). We used 8 male MHC-TRAF2_LC at 2–3 months of age. The diets of both groups of mice consisted of standard commercial chow (2,920× Harlan Teklad, Indianapolis, IN, USA) with free access to food and water. All animal protocols were approved by the Institutional Animal Care and Use Committee of Baylor College of Medicine in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Anesthesia was initially induced in mice with 2.5% isoflurane before transfer to a heated (37 ± 1°C) electrocardiography board (MouseMonitor S, Indus Instruments, Webster, TX) with isoflurane maintained at 1.5%. The mouse paws were attached to electrodes to measure electrocardiogram (ECG). The experimental setup and the workflow is shown in Figure 1.

Cardiac Doppler aortic outflow velocity and mitral inflow signals were measured with 20 MHz Doppler probe with all signals (including blood pressure and ECG) simultaneously acquired using Doppler Flow Velocity System (DFVS; Indus Instruments, Webster, TX). From the cardiac Doppler signals, we calculated peak and mean aortic velocities, stroke distance (Sd), aortic ejection time (ET), peak and mean aortic accelerations, Mitral early peak (E) and atrial peak (A) velocities, E/A ratio, E-deceleration time (DT), isovolumic contraction (IVCT) & relaxation (IVRT) times, and myocardial performance index (also known as Tei index = (IVCT + IVRT)/ET).

Blood pressure measurements were made as previously described (22, 23). Briefly, the right carotid artery was isolated and cannulated with a pressure catheter (SPR-1,000: Millar Instruments, Inc., Houston, TX) and advanced into the ascending aorta to measure aortic pressure (along with simultaneous measurement of aortic blood flow velocity as described below). The aortic pressure and velocity signals were displayed simultaneously, and 2 s segments of both signals were recorded using DFVS system. The catheter was then advanced into the LV. Again, 2 s segments of LV pressure, and systolic (SBP), diastolic (DBP), mean (MBP), pulse pressures (PP), and rate-pressure product (RPP) were calculated from aortic pressure. Peak LV pressure (P_LVP), indices of contractility (+dP/dt_max) and relaxation (−dP/dt_max), relaxation time constant (tau), and LV end-diastolic pressure (LVEDP) were calculated from LV pressure.

Aortic input impedance was determined using previously reported methods (22–25). Briefly, impedance was determined using the simultaneously measured ascending aortic pressure and velocity signals. The pressure and velocity waveforms were processed using a discrete Fourier transformation to determine aortic input impedance (Zi). Peripheral vascular resistance (Zp—impedance at zero frequency), the impedance at first harmonic (Z₁), and characteristic impedance (Zc—an average of 2nd to 10th harmonic) were extracted from the modulus of aortic impedance (|Zi|). Pulse wave velocity was calculated from |Zi| as Zc/ρ (ρ—density of blood).

Arterial elastance (Ea) was calculated as ESP/SV (stroke volume, SV = Sd * aortic cross-sectional area), end-systolic elastance (Ees) was calculated as ESP/ESV, ventricular-vascular coupling (VVC) was calculated as Ea/Ees, and stroke work (SW) was calculated as ESP*SV (17).

All the data are presented as mean ± standard error of the mean (SEM). Statistical analyses were performed via analysis using Student's T-test on Prism (GraphPad Software; La Jolla, USA). Non-parametric tests (Mann-Whitney, in lieu of t-test) were utilized for data that were not normally distributed, as determined by the Kolmogorov-Smirnov, Shapiro-Wilk, Anderson-Darling, and D'Agostino & Pearson tests. Normalcy is indicated in Supplementary Figure S1 as QQ plots with non-normally distributed data marked by *.

We found that the cardiac systolic function, as determined by peak and mean aortic flow velocities, peak, and mean aortic accelerations (Figure 3) and + dP/dt_max (Figure 4), is diminished in TRAF2 mice compared to the CTR mice, which may indicate a mild systolic dysfunction or enhanced systolic reserve. The LV pressure indices of contractility (+dP/dt_max) and relaxation (−dP/dt_max) were significantly lower in the TRAF2 mice and followed a similar trend that was observed in the aortic blood pressure and velocity, and in agreement with the decrease in %FS reported by Burchfield et al. (9) Again, these diminished contractile and relaxation functions in TRAF2 mice are similar to what we observed in the young dwarf Little mice (26). We have not challenged these mice to determine systolic reserve capacity, but our group showed that left ventricular developed pressure recovered better in the TRAF2 mice after IR injury and observed that the cytoprotective effects of TRAF2 were mediated by NF-κB activation (9). However, others have reported that suppression of NF-kB with oligonucleotide transcription factor decoys diminishes IR myocardial injury (14) and that cardiac function improved after myocardial infarction with the absence of NF-kB p50 unit (13). Our group reiterated the cytoprotective effects of TRAF2 may have been afforded via crosstalk between canonical and noncanonical NF-kB signaling pathways (27). Other groups have also shown that silencing of TRAF2 results in the dysregulation of NF-κB marked by overactivation (28). Conversely, TRAF2 overexpression has been linked to the activation of IκB kinase, responsible for the nuclear translocation of NF-κB (29). Given the relevance of NF-κB in developing an anti-inflammation drug, future studies must further expound upon the changes in NF-κB upon TRAF2 expression, as activation of the NF-κB offers a potential pathway by which TRAF2 overexpression mediates cardiac function.

A more recent study noted that TRAF2 maintains myocardial homeostasis through the facilitation of cardiac myocyte mitophagy thus preventing inflammation and cell death (16). We also found that aortic ejection times were significantly prolonged in the TRAF2 mice (Figure 3D). Obata et al. (30) reported that longer aortic ejection time is associated with lower blood pressures (30), and perhaps the low blood pressure in TRAF2 mice in our study may have caused longer ejection times. Also, longer ET results from longer action potential duration in old animals and may be associated with cellular changes in the mechanics of excitation–contraction coupling (31). Similar cellular changes may be present in TRAF2 that could have caused prolonged ejection time.

The cardiac diastolic function, as defined by E/A ratio, was higher in TRAF2 mice than CTR mice (Figure 4C) indicating improved diastolic function. However, E >> A (or E/A > 2) and shorter E-deceleration time is associated with a moderate decrease in compliance (32). The TRAF2 mice had E/A > 2 and shorter E-deceleration time (Figure 4D) perhaps indicating a decline in compliance which agrees with the observation by Burchfield et al. (9), that TRAF2 mice had mildly hypertrophic hearts (9). Isovolumic contraction time is the index of LV contractility and isovolumic relaxation time is an index of LV relaxation, but both are dependent on the heart rate (33) and other factors. Prolonged IVRT indicates impaired relaxation and prolonged IVCT is associated with reduced EF, both increasing the risk for heart failure (34). Since IVCT and IVRT are both associated with LV ejection time, the myocardial performance index (MPI or Tei Index) provides an overall cardiac performance (33). We found that TRAF2 overexpression resulted in a small but significant increase in Tei Index (Figure 4G), which may indicate impaired cardiac function like that reported in patients (35).

While TRAF2 overexpression confers cardioprotective effects, the small but significant changes in the systolic and diastolic parameters indicate a modest decrease in baseline function. It remains to be seen if this diminished function deteriorates with age or remains stable over their life span like that in the dwarf Little mice, which maintain a diminished but stable cardiac function over their life span (26).

TRAF2 mice had significantly lower systolic, mean, and pulse pressures compared to CTR mice. The lower blood pressures combined with lower blood velocities in TRAF2 mice make LV afterload appear normal. Previous studies reported that every 20 or 10 mmHg increase in SBP or DBP, respectively, beyond traditionally healthy values (36) doubles the chance of major long-term consequences, such as heart failure or stroke (37). Since SBP remains a stronger indicator of cardiovascular health than DBP at ages over 50, our findings suggest that the cardioprotective effects of TRAF2 potentially may be conferred across aging (38). Pulse pressure that is ≤25% of systolic pressure is considered too low and typically occurs in aortic stenosis, cardiac tamponade, aortic stenosis, or with blood loss (39). Pulse pressure in TRAF2 mice was ≈25% of systolic pressure, perhaps indicating modest dysfunction. Given that we measured blood pressure in these mice at a young age, studies at older ages may reveal if the cardioprotective effects of TRAF2 overexpression are maintained across the lifespan. The rate pressure product (RPP = SBP × heart rate), which is an indirect measurement of the heart oxygen consumption (40), was significantly lower in TRAF2 mice than in CTR mice, indicating a lower heart workload.

A measure of myocardial performance that includes interaction with aortic function, the ventricular-vascular coupling (VVC) was significantly higher in the TRAF2 mice indicating a diminished global cardiovascular efficiency. While aortic elastance was unchanged, the higher VVC in TRAF2 mice was mainly due to significantly lower end-systolic LV elastance. Similar observations were made in patients with major adverse cardiovascular events (17). Also, it has been shown that decreased LV stroke work index is associated with diminished LV systolic and diastolic function in cardiac intensive care unit patients (41). We found that stroke work is significantly lower in the in the TRAF2 mice, perhaps indicating a diminished cardiac function.

Defined as the ratio of modulus of pressure to modulus of velocity at several harmonics (0–10) in the frequency domain, aortic impedance provides a comprehensive description of the afterload experienced by the left ventricle. This includes the pulsatile and the steady components of the hydraulic load compared with when pressure and velocity are considered individually (23, 24, 42). We used blood flow velocity instead of volume flow because both blood flow velocity and pressure are independent of body weight or size (24). We found that the peripheral vascular resistance (Zp, impedance modulus at zero frequency), the strength of wave reflections from the periphery (Z1, impedance modulus at the first harmonic), and characteristic impedance (Zc, average of 2–10 harmonics of impedance modulus), and pulse wave velocity (PWVZ; derived from Zc; aortic stiffness index) were not different in TRAF2 mice when compared to CTR mice (Figure 8). The unchanged Zc & PWV_Z seems to conflict with our finding that pulse pressure was lower in TRAF2 mice. While PWV is a direct measure of aortic stiffness, pulse pressure is used as a surrogate stiffness index that also reflects stroke volume (which was also lower in TRAF2 mice) and both measures (stiffness & pulse pressure) are mildly correlated (43). In another study anti-hypertensive treatments in patients resulted in no changes in aortic PWV but had a significant decrease in brachial pulse pressure (44).