Neonatal Deletion of Hand1 and Hand2 within Murine Cardiac Conduction System Reveals a Novel Role for HAND2 in Rhythm Homeostasis

The cardiac conduction system, a network of specialized cells, is required for the functioning of the heart. The basic helix loop helix factors Hand1 and Hand2 are required for cardiac morphogenesis and have been implicated in cardiac conduction system development and maintenance. Here we use embryonic and post-natal specific Cre lines to interrogate the role of Hand1 and Hand2 in the function of the murine cardiac conduction system. Results demonstrate that loss of HAND1 in the post-natal conduction system does not result in any change in electrocardiogram parameters or within the ventricular conduction system as determined by optical voltage mapping. Deletion of Hand2 within the post-natal conduction system results in sex-dependent reduction in PR interval duration in these mice, suggesting a novel role for HAND2 in regulating the atrioventricular conduction. Surprisingly, results show that loss of both HAND factors within the post-natal conduction system does not cause any consistent changes in cardiac conduction system function. Deletion of Hand2 in the embryonic left ventricle results in inconsistent prolongation of PR interval and susceptibility to atrial arrhythmias. Thus, these results suggest a novel role for HAND2 in homeostasis of the murine cardiac conduction system and that HAND1 loss potentially rescues the shortened HAND2 PR phenotype.


Introduction
The pumping of the mammalian heart is a result of synchronized contractions of atrial and ventricular cardiomyocytes that are coordinated via electrical impulse transmitted through a specialized cell network, referred to as the cardiac conduction system (CCS). Starting in the right atrium, pacemaker cells within the sinoatrial node (SAN), initiate the cardiac electrical impulse [1]. This electrical impulse moves across both atria before its velocity is reduced at the atrioventricular node (AVN), allowing for ventricular diastole. The delayed electrical signal then passes to the bundle of HIS and right and left bundle branches (BBs) releasing their trigger for ventricular cardiomyocyte contraction at the apex of each ventricle through epicardial break through points [2,3]. The electrical signal that triggers the wave of cardiomyocyte contraction through the ventricles is via terminal network of specialized cardiomyocytes termed Purkinje fibers (PF) which compose the ventricular conduction system (VCS) [4]. PF activation and subsequent epicardial breakthrough of the electrical signal results in ventricular systole. Improper development and/or maintenance of any of the CCS components are associated with causation of cardiac arrhythmia and sudden cardiac death [5].
ECG phenotype, suggesting loss of HAND1 rescues loss of HAND2. Deletion of Hand2 in the embryonic LV results in inconsistent increases in PR interval at some time points, increased susceptibility to atrial arrhythmias, and increased dyssynchrony between left and right ventricular activation in hearts from male mice. Taken together these findings support the idea that HAND1 function within the CCS appears to be purely congenital, that HAND2 loss alters atrial arrhythmogeneity as well as AV and intraventricular conduction, and that removing HAND1 can restore normal PR interval due to HAND2 loss-of-function.

Mouse Lines and Breeding
Mouse lines Cntn2 3 UTR-IRES-Cre-EGFP (Cntn2 Cre-EGFP ) [18], Hand1 LV -Cre [17], Hand1 tm2Eno (Hand1 fx/fx ) [12], Hand2 tm1Cse (Hand2 fx/fx ) [19] were genotyped by PCR and Southern blotting as previously described [15,20]. Complete list of alleles used for experiments is listed in Table S1 (Supplementary Materials). Mice were randomly assigned numbers at birth and selected for experiments without blinding based on genotype. ARRIVE guidelines 2.0 were used for study design (Table S8). Mice were weaned at 4 weeks and consequently used for ECG analysis between 5 and 25 weeks of age as indicated. All repeated measurements were conducted under standardized conditions.

Surface ECG Recording and Analysis
Male and female mice used in ECG studies were assayed at 5-week, 10-week, 15-week, 20-week, and 25-week time points. The same cohorts of mice for each time point were maintained. Mice were anesthetized with isoflurane administered as 2.5% volume vapor and placed on a heating pad in a supine position with continuous body temperature monitoring using a rectal probe to maintain 37-38 • C. Mice were under anesthesia via mask for 90-120 s and depth of narcosis was assessed by gently pressing on paws and monitoring heart rate. Needle electrodes were placed subcutaneously in the right and left axilla and in the left groin. Bipolar leads I and II signals were simultaneously recorded for a minimum of 1 min using the PowerLab 26T (ADI Instruments, Colorado Springs, CO, USA) with a sampling rate of 2 k/s at 20 mV range. The heart rate and ECG signal from Lead I and Lead II were monitored in real time using the LabChart Pro software (ADI Instruments, Colorado Springs, CO, USA) and recorded.
Analysis of ECG data was carried out using LabChart Pro software (ADI Instruments, Colorado Springs, CO, USA). Lead III data was not analyzed in this study due to inaccuracies identified within the algorithm used to calculate ECG values by the LabChart software. For Lead I and Lead II, ECG signals were averaged by alignment to QRS maximum using a minimum of 100 beats. For analysis, PR interval was measured from the beginning of P-wave (as determined by deflection from isoelectric line) to the beginning of QRS complex. QRS duration was measured from the first deflection of the Q-wave to the nadir of the Swave (defined as the point of minimum voltage in the terminal portion of the QRS complex; method I, QRS1), and to the onset of the J wave (method II, QRS2) [21]. The QT interval was defined as the interval from the beginning of the QRS complex to the end of the T-wave (defined as the point where the T-wave merges with the isoelectric line). This QT value is corrected (QTc) by the Mitchell formula QT/(sqrt(RRx10)) [22]. The R-R interval was obtained as the average R-R interval over the sampling period. ECG recordings of good quality (low baseline drift, low noise, minimum number of beats to average) were used for analysis for each genotype/time point.

Optical Mapping
High-resolution optical mapping experiments were performed on hearts isolated from 30-to 50-week-old Cntn2 Cre-EGFP/+ ; Hand1 fx/fx , Hand1 LV -Cre; Hand2 fx/fx mice and control littermates (Hand1 fx/fx and Hand2 fx/fx mice) as described previously [13,23]. Hearts were isolated and retrogradely perfused in Langendorff mode with temperature-controlled (37 • C) Krebs-Henseleit solution (pH 7.4 when gassed with a mixture of 95% O 2 and 5% CO 2 ) at an aortic pressure of 70 cm H 2 O. A volume-conducted ECG was monitored continuously throughout the experiment. After 10 min of stabilization, the hearts were stained with the voltage-sensitive dye Di-4-ANEPPS (2 µL of a 2-mmol/L stock solution). The heart was then washed with dye-free solution for 5 min followed by the addition of (±)-blebbistatin to uncouple contraction from excitation (10 µmol/L; Tocris Bioscience, Minneapolis, MN). The stained hearts were illuminated with a laser at a 532 nm wavelength and the fluorescence was collected by a MiCAMUltima-L CMOS camera (SciMedia, Costa Mesa, CA, USA) through a 715-nm long-pass filter. The fluorescence was recorded at a 1 ms/frame rate in a 100 × 100-pixel grid with a spatial resolution of 0.35 × 0.35 mm 2 per pixel. Optical signals were processed with both spatial (3 × 3 pixels Gaussian filter) and temporal (3 frames moving average) filtering. Hearts were paced at the right atrium at a cycle length of 120 ms. Two 1 s recordings were captured sequentially while the right atrium was paced. Finally, two 1 s recordings were acquired while the hearts were paced from the ventricular apex at a cycle lengths of 120 ms [13,23].

Statistical Analysis
Differences between data sets were first analyzed for normal distribution by applying the Shapiro-Wilks' test for normality. Student's t-test calculator for independent means was used for all normal data sets. Data sets that failed the test for normality were analyzed using the Mann-Whitney U test (* p < 0.05, # p < 0.01). The one-way repeated measures ANOVA test was used to examine trends within groups. Groups that showed significant differences were further analyzed with pairwise t-test comparisons. SigmaPlot version 13 (Inpixon, Palo Alto, CA, USA) was used to conduct statistical analysis.

Neonatal Deletion of Hand1 Does Not Observably alter CCS Function
To investigate the role of HAND1 in the post-natal CCS, we employed the Cntn2 Cre-EGFP knock-in allele crossed to Hand1 fx/fx mice to generate Cntn2 Cre-EGFP ; Hand1 fx/fx mice (H1CKOs). Male and female littermates wherein the Cre allele was absent (Hand1 fx/fx , H1control) were used as controls. We measured heart rate, PR interval, QRS1, QRS2, and QTc (Table S2). Results reveal a statistically significant reduction in heart rate in female H1CKOs at 5-week time point (371 bpm ± 28 vs. 422 bpm ± 9 in controls, p = 0.002 by Student's t-test), an increase in PR interval duration in female H1CKOs at 25 weeks in Lead I (44.93 ms ± 3.35 vs. 39.48 ms ± 2.35 in controls, p = 0.0096 by Student's t-test), an increase in PR interval (44.30 ms ± 4.49 vs. 38.42 ms ± 3.29 in controls, p = 0.022 by Student's t-test) and decrease in QTc duration (37.18 ms ± 6.05 vs. 47.74 ms ± 5.83 in controls, p = 0.0082 by Student's t-test) in female H1CKOs at 25 weeks in Lead II ( Figure S1, Table S2). However, none of these changes are consistently observed across time points or tested parameters. Therefore, we conclude that in contrast to the observations reported from Hand1 congenital deletion mice [15], there are no significant electrocardiographic correlations of CCS dysfunction in the parameters tested in H1CKOs to suggest that HAND1 is playing a role in maintaining cardiac conduction beyond the development of the CCS ( Figure S1).
Hand1 is robustly expressed in the myocytes of the embryonic LV when the terminal VCS is patterned [10]. Additionally, it is also established that the lack of an electrocardiographic phenotype does not rule out the possibility of a ventricular activation defect [21]. Indeed, we previously observed VCS activation defects in Hand1 36;75/36;75 SNP mutants that exhibited no statistically significant variations in ECG parameters [15]. Therefore, to determine ventricular activation patterns in H1CKOs, we performed epicardial optical voltage mapping of isolated perfused, atrially paced hearts from five male H1CKOs, three male H1controls, three female H1CKOs, and two female H1controls between 30 and 50 weeks of age ( Figure 1). Four male and two female H1CKO hearts as well as one male and both female controls showed epicardial breakthroughs occurring synchronously over each ventricular apex, which is expected for mature mouse hearts ( Figure 1A). One male and one female H1CKO heart exhibited a single LV epicardial breakthrough indicative of right BBB (labeled L in Figure 1B). One female H1CKO heart displayed multiple sites of activation throughout the ventricles ( Figure 1C, R1-R2, L1-L4), resulting in wavefront collisions within the ventricular myocardium. Two male H1controls also presented with right BBB or delayed (3 ms) RV breakthrough ( Figure 1D). During pacing at the apex of the LV (cycle length, 120 ms), activation occurred within 10-16 ms in all 13 hearts. This range of ventricular activation times is in excellent agreement with previous reports for isolated-perfused wildtype mouse hearts [24]. Since activation maps during ventricular pacing did not reveal areas of slowed or blocked conduction, we did not perform detailed analyses of local conduction velocities. Additionally, since no robust changes in the QT c interval across time were found, we did not evaluate ventricular action potential duration. Taken together, these findings do not reveal consistent differences in ventricular myocardial conduction between H1CKO mutants and H1controls, further suggesting that Hand1 function is not required for post-natal VCS homeostasis.
weeks of age ( Figure 1). Four male and two female H1CKO hearts as well as one male a both female controls showed epicardial breakthroughs occurring synchronously o each ventricular apex, which is expected for mature mouse hearts ( Figure 1A). One m and one female H1CKO heart exhibited a single LV epicardial breakthrough indicative right BBB (labeled L in Figure 1B). One female H1CKO heart displayed multiple sites activation throughout the ventricles ( Figure 1C, R1-R2, L1-L4), resulting in wavefront c lisions within the ventricular myocardium. Two male H1controls also presented with ri BBB or delayed (3 ms) RV breakthrough ( Figure 1D). During pacing at the apex of the (cycle length, 120 ms), activation occurred within 10-16 ms in all 13 hearts. This range ventricular activation times is in excellent agreement with previous reports for isolat perfused wildtype mouse hearts [24]. Since activation maps during ventricular pacing not reveal areas of slowed or blocked conduction, we did not perform detailed analy of local conduction velocities. Additionally, since no robust changes in the QTc inter across time were found, we did not evaluate ventricular action potential duration. Tak together, these findings do not reveal consistent differences in ventricular myocardial c duction between H1CKO mutants and H1controls, further suggesting that Hand1 funct is not required for post-natal VCS homeostasis.

Neonatal Deletion of Hand2 Results in Sex-Dependent Effects on CCS Function
To investigate the role of HAND2 in the post-natal CCS, we bred the Cntn2 Cre-EGFP/+ knockin Cre line to Hand2 fx/fx mice to generate Cntn2 Cre-EGFP/+ ; Hand2 fx/fx mice (H2CKOs) and compared these to littermate Hand2 fx/fx (H2control) mice. We observe statistically significant shortening in Lead I PR interval in H2CKOs males at 5-, 10-, 15-, 20-, 25-week time points (Figure 2A,B, Table S3). In Lead II, we observe statistically significant shortening in PR interval at 5-20-week time points in H2CKOs males compared to H2control males ( Figure S2). No changes were observed in QRS1, QRS2, and QTc in male H2CKOs in Lead I and Lead II ( Figure S2, Table S3). Heart rates in H2CKOs are decreased at 10 weeks (517 bpm ± 28, p = 0.021 by Student's t-test) and 20 weeks (521 bpm ± 23, p = 0.045 by Student's t-test) compared to H2controls (550 bpm ± 27 at 10 weeks and 550 bpm ± 35 at 20 weeks); however, at the three other remaining time points examined, heart rates between H2CKOs and H2controls exhibit no statistically significant differences, suggesting that the consistent shortening of PR-interval phenotype is independent of changes in heart rate.
It is noteworthy that, similar to our observation in males, the PR interval in female H2CKOs is shorter at all time points analyzed in both Lead I and Lead II (Table S3). To examine the role that sex might play in the reduction of PR interval, we compared male and female Lead I PR values from H2CKOs ( Figure 2D, Table S3). At 5-week, 10-week and 20-week time points, there is no significant difference in PR interval; however, at 15-and 25-week time points, PR interval in H2CKO males are shorter ( Figure 2D). We compared the morphology of the P-wave in these recordings and do not observe a variation over time points (data not shown). To determine trends within groups and individual mice, we performed a one-way repeated ANOVA on the data (Table S6). Results show that there does not appear to be a simple progressive decrease in the PR interval within individual H2CKOs. However, pairwise multiple comparisons using the Student-Newman-Keul method indicate statistically significant changes between the 5-week and 25-week time points, and between the 20-week and 25-week time points ( Figure 2B, Table S6), suggesting a worsening of the PR interval shortening phenotype over time. This would suggest a homeostatic role for Hand2 in the regulation of atrioventricular conduction. Thus, loss of Hand2 in post-natal conduction system leads to shortened PR interval in males, and females exhibit similar phenotype albeit less severe, suggesting defective atrioventricular conduction.

Neonatal Deletion of Hand1 and Hand2 Results in an Intermediate CCS Phenotype
Considering the variable phenotypes observed when Hand1 or Hand2 are deleted in the post-natal conduction system, we proceeded to test whether loss of both HAND factors resulted in changes to CCS. We intercrossed Cntn2 Cre-EGFP/+ and Hand1 fx/fx Hand2 fx/fx mice to generate Cntn2 Cre-EGFP/+ ; Hand1 fx/fx Hand2 fx/fx mice (H1H2DKOs) and littermate Hand1 fx/fx Hand2 fx/fx (H1H2controls). We observe a statistically significant decrease in male PR interval duration in Lead I at 10-week and 20-week time points ( Figure 3A, Table S4). PR interval in H1H2DKO is shortened in Lead I at other time points compared to H1H2controls, although it does not reach significance (Table S4, Figure S3). We also observe that the shortening of PR interval in H1H2DKO is not as severe as observed in H2CKOs (Tables S3 and S4). In Lead I, H1H2DKOs QRS1 is significantly lengthened compared to controls in 25-week males. In Lead II, H1H2DKOs show increased QRS1 duration at 15 weeks compared to H1H2controls. However, these QRS defects observed in H1H2DKOs are not consistently seen across all time points, and thus, might reflect an experimental artifact.
to generate Cntn2 Cre-EGFP/+ ; Hand1 fx/fx Hand2 fx/fx mice (H1H2DKOs) and litterm Hand1 fx/fx Hand2 fx/fx (H1H2controls). We observe a statistically significant decrease in m PR interval duration in Lead I at 10-week and 20-week time points ( Figure 3A, Table PR interval in H1H2DKO is shortened in Lead I at other time points compared to H1H2 trols, although it does not reach significance (Table S4, Figure S3). We also observe the shortening of PR interval in H1H2DKO is not as severe as observed in H2CKOs (Ta S3 and S4). In Lead I, H1H2DKOs QRS1 is significantly lengthened compared to con in 25-week males. In Lead II, H1H2DKOs show increased QRS1 duration at 15 weeks c pared to H1H2controls. However, these QRS defects observed in H1H2DKOs are not sistently seen across all time points, and thus, might reflect an experimental artifact.
In female H1H2DKOs, the PR interval in Lead I and Lead II trends shorter at all points and is significantly shortened at the 25-week time point ( Figure 3B, Table S4). L II measurements of females also shows significantly longer QRS2 and QTc interval a week compared to H1H2controls, although this trend is not maintained across all points (Table S4). Repeated one-way ANOVA testing did not demonstrate any prog sive changes in PR interval (data not shown). Thus, loss of both Hand genes within post-natal CCS results in some significant conduction parameters changes, but there no consistent trends in these data.  In female H1H2DKOs, the PR interval in Lead I and Lead II trends shorter at all time points and is significantly shortened at the 25-week time point ( Figure 3B, Table S4). Lead II measurements of females also shows significantly longer QRS2 and QTc interval at 15-week compared to H1H2controls, although this trend is not maintained across all time points (Table S4). Repeated one-way ANOVA testing did not demonstrate any progressive changes in PR interval (data not shown). Thus, loss of both Hand genes within the post-natal CCS results in some significant conduction parameters changes, but there are no consistent trends in these data.

Embryonic Deletion of Hand2 within the LV Results in Increase in PR Interval and Arrhythmias
Although HAND2 is predominantly expressed in the endocardium [25] of the developing heart, myocytes of the embryonic RV and LV also express Hand2 and LV deletion of Hand2 shows dramatic phenotypic influence when Hand1 is similarly deleted [17,25]. To test whether loss of Hand2 in the developing embryonic LV myocardium leads to VCS defects, we intercrossed Hand1 LV -Cre with Hand2 fx/fx to generate Hand1 LV -Cre; Hand2 fx/fx (H2LVCreCKOs) and littermate H2 fx/fx (H2controls). In contrast to neonatal CCS Hand2 deletion, H2LVCreCKOs males exhibit statistically significantly longer PR intervals at 10 weeks and 15 weeks compared to H2controls in Lead I (Table S5). At week 10, H2LVCreCKO males in Lead I also reveal a significantly longer QRS2 duration (Table S5). At 25 weeks, H2LVCreCKOs males exhibit a statistically significant reduction in heart rate. Female H2LVCreCKOs have statistically significant longer QTc interval at week 10 in Lead I and at week 5 in Lead II; however, this phenotype was not consistent at other time points ( Figure S4, Table S5).
Next, we functionally evaluated the CCS of H2LVCreCKOs by optically mapping 41 H2LVCreCKO hearts (14 male hearts) and 16 H2control hearts (seven male hearts; Figure 4). Maps were obtained during atrial or ventricular pacing at a cycle length of 120 ms. In each heart, we measured the interval between right atrial depolarization and first ventricular breakthrough during atrial pacing, the number of right and left ventricular breakthrough sites, the prevalence of BBB, the delay between the right and left ventricular breakthroughs (in hearts without bundle branch block), and the ventricular activation time during pacing at the LV apex. Exemplary voltage maps are illustrated in Figure 4A-D, and the results are summarized in Table S6. While the prevalence of BBB was not significantly different between female (10 out of 36 hearts) and male hearts (4 out of 21 hearts; p > 0.05 by Fisher Exact test), female hearts exhibited a higher prevalence of right BBB (p = 0.04 by Fisher Exact test). Numbers of breakthrough sites were similarly distributed between sexes and genotypes, with the majority of hearts exhibiting a single breakthrough on either or each ventricular apex. In male H2LVCreCKO hearts without BBB, there was a significant delay between occurrences of epicardial breakthroughs compared to H2control hearts (Table S6). During ventricular pacing at the LV apex (cycle length, 120 ms), activation occurred within 10 to 16 ms in all 57 hearts with no local conduction slowing or block being apparent on epicardial voltage maps. The average time interval from onset of right atrial depolarization to the first ventricular epicardial breakthrough in hearts without BBB was significantly longer in H2LVCreCKO hearts of either sex compared to their respective control hearts (see Table S6). Since isochrone activation maps and surface electrocardiograms did not reveal ventricular conduction or repolarization abnormalities, respectively, we did not perform detailed analyses of local conduction velocities or action potential duration. Four H2LVCreCKO mice of either sex, but none of H2controls, displayed arrhythmias, manifesting as supraventricular tachycardia with Wenckebach-type or 1:1 AV conduction (examples 1 and 4, respectively, in Figure 4E), junctional rhythm (example 2 in Figure 4E), or varying P wave morphologies (example 3 in Figure 4E). tachycardia with Wenckebach-type AV conduction blo (left; first arrow denotes sinus beat whereas the following three arrows indicate atrial prematu beats exhibiting a P wave morphology that is distinctly different from that during sinus rhyth The third premature beat is not conducted to the ventricles). Right: A sinus beat (first arrow) followed by an atrial premature beat (second arrow) initiating a sustained atrial tachycardia. No difference in P wave morphology before from that during tachycardia. Example 2: progressi shortening of PR interval until the P wave is eventually embedded in the QRS complex and und tectable on the ECG. The sinus node functions as a pacemaker for the atria but slowed conducti across the AV node allows a junctional pacemaker located between the AV node and the origin the His bundle branches to emerge and drive the associated QRS complex, i.e., junctional rhyth Left lower panel shows zoom-in views of the correspondingly numbered cycles in the low-reso tion traces. Example 3: sudden change in P wave morphology, indicative of a competing ectop atrial pacemaker. Example 4: sustained supraventricular tachycardia with 1:1 AV conduction. A terisks (*) denote P waves. The tachycardia is initiated by an atrial premature contraction (arrow Lower trace shows zoom-in views of 3 consecutive cycles framed by the dotted box.

Discussion
Anomalies in cardiac rhythm when encountered and untreated are known to lead sudden death [5]. The CCS specifies and differentiates from existing cardiomyocytes v the interactions of several transcription factors within the embryonic heart [7]. We recent discovered that HAND1 plays an important congenital role in the formation and functi of the VCS [15] and the goal of this study was to interrogate HAND1 function after t CCS is formed and to determine if HAND2 plays a similar or unique role in CCS functio We can conclude from Hand1 neonatal deletion Cntn2 Cre-EGFP ; Hand1 fx/fx (H1CKOs) th HAND1 does not appear to be influencing CCS homeostasis in any significant way. A hough we do observe a longer PR interval duration on ECG at 25-week in H1CKO femal short run of atrial tachycardia with Wenckebach-type AV conduction block (left; first arrow denotes sinus beat whereas the following three arrows indicate atrial premature beats exhibiting a P wave morphology that is distinctly different from that during sinus rhythm. The third premature beat is not conducted to the ventricles). Right: A sinus beat (first arrow) is followed by an atrial premature beat (second arrow) initiating a sustained atrial tachycardia. Note difference in P wave morphology before from that during tachycardia. Example 2: progressive shortening of PR interval until the P wave is eventually embedded in the QRS complex and undetectable on the ECG. The sinus node functions as a pacemaker for the atria but slowed conduction across the AV node allows a junctional pacemaker located between the AV node and the origin of the His bundle branches to emerge and drive the associated QRS complex, i.e., junctional rhythm. Left lower panel shows zoom-in views of the correspondingly numbered cycles in the low-resolution traces. Example 3: sudden change in P wave morphology, indicative of a competing ectopic atrial pacemaker. Example 4: sustained supraventricular tachycardia with 1:1 AV conduction. Asterisks (*) denote P waves. The tachycardia is initiated by an atrial premature contraction (arrow). Lower trace shows zoom-in views of 3 consecutive cycles framed by the dotted box.
Overall, our findings indicate that embryonic loss of cardiac Hand2 expression predisposes to arrhythmias and causes abnormal conduction across the CCS.

Discussion
Anomalies in cardiac rhythm when encountered and untreated are known to lead to sudden death [5]. The CCS specifies and differentiates from existing cardiomyocytes via the interactions of several transcription factors within the embryonic heart [7]. We recently discovered that HAND1 plays an important congenital role in the formation and function of the VCS [15] and the goal of this study was to interrogate HAND1 function after the CCS is formed and to determine if HAND2 plays a similar or unique role in CCS function. We can conclude from Hand1 neonatal deletion Cntn2 Cre-EGFP ; Hand1 fx/fx (H1CKOs) that HAND1 does not appear to be influencing CCS homeostasis in any significant way. Although we do observe a longer PR interval duration on ECG at 25-week in H1CKO females and BBB in some males and females when optically mapped, these phenotypes are not consistent over the evaluation period making these mice indistinguishable from controls ( Figure 1 and Table S1). It is possible that Hand1 fx/fx mice could also present with CCS phenotypes that could be confounding our analysis. However, when compared to Nkx2.5 Cre/+ ;Hand1 fx/fx [13], Hand1 ∆LV/∆LV or Hand1 36;75/36;75 [15], significant phenotypes are statistically evident when utilizing Hand1 fx/fx mice for control comparisons. Based on these cumulative data, we conclude that under normal non-pathological conditions, the neonatal loss of HAND1 does not pathologically alter murine CCS function.
Interestingly, neonatal deletion of Hand2, Cntn2 Cre-EGFP ; Hand2 fx/fx (H2CKOs) results in a shorter PR interval compared to controls (Figure 2). Earlier studies revealed that HAND2 influenced the formation of pacemaker like cells from fibroblast differentiation analysis [14]. Shortening of PR interval can occur due to a change in location of the atrial pacemaker such that the time from impulse generation to AVN excitation is shortened as a result of the smaller distance between the pacemaker and AVN. However, this is an unlikely mechanism here, as the neonatal deletion of Hand2 occurs well after CCS morphogenesis is complete. Another possible cause for shortening of PR interval is an acceleration of conduction across the bundle of His. Pre-excitation syndromes such as Wolf-Parkinson-White can also cause a shortening of PR-interval [26]. However, we can rule out the existence of accessory atrioventricular conduction pathways as underlying mechanism because neither atrial nor ventricular pacing at incremental rates in our optical mapping experiments revealed evidence for propagation across AVN-bypassing tracts. It is interesting that shortening of the PR interval is also observed in Hand1 ∆LV/∆LV mice [15] as well as in Gata4 +/mice [27]. Evidence suggests GATA4 is an upstream transcriptional regulator of HAND1 (binding to two evolutionarily conserved GATA cis-elements within the Hand1 LV-enhancer) [15], as well as regulating Hand2 RV expression [28]. Embryonic LV deletion of both Hand1 and Hand2 exhibit single ventricle like phenotypes with compromised cardiac function, but the extensive morphological phenotypes would confound any clear conclusions from measure of conduction parameters [17].
Given the significant variations in ECG phenotypes observed in H1CKOs and H2CKOs, we generated conditional knockouts of both HAND factors with the Cntn2 Cre-EGFP/+ ; Hand1 fx/fx Hand2 fx/fx (H1H2CKOs) allele to test if loss of both factors in the post-natal CCS had an additive or reductive effect on CCS function ( Figure 3, Table S3). We observe significant reduction in PR interval duration in males and females at the 10-week and 20-week time point. However, this phenotype is not as severe or as consistent as observed in H2CKOs, suggesting some rescue of the H2CKO phenotype from the loss of the Hand1 allele.
Surprisingly, embryonic Hand2 deletion using the Hand1 LV -Cre revealed a longer PR interval on ECG (Table S5), as well as longer delays between onset of right atrial depolarization and first epicardial breakthrough in optical voltage maps (Table S7). At first glance this seems odd; however, it is possible that the loss of HAND2 impacts CCS morphogenesis such that distance between the atrial pacemaker and AVN is increased. We currently do not have evidence of this possibility, but it would explain the variation from the neonatal results obtained from Cntn2 Cre-EGFP/+ mediated deletion. Notably; however, some adult H2LVCreCKO mice have varying P-wave morphology and atrial tachycardia (Figure 4), suggesting that embryonic loss of cardiac Hand2 expression alters atrial electrical properties, increasing their arrhythmia susceptibility. Atrial arrhythmias result from increased automaticity, triggered activity arising from afterdepolarizations, re-entry electrical activity, or a combination of these. Since afterdepolarizations were never observed in atrial voltage maps, they are unlikely to underlie arrhythmia in H2LVCreCKO mice. Whether the other mechanisms contributed to atrial arrhythmogenesis remains to be determined. HAND1 and HAND2 are robustly expressed in the sympathetic nervous system (SNS), derived from migrating cardiac neural crest cells, and HAND2 is required for SNS development and function [29][30][31][32]. Functional changes in the AV conduction system can be caused by the action of the autonomic nervous system [33]. It is possible that loss of HAND factors within the post-natal cardiac autonomous nervous system could cause the observed time-point specific variations in the CCS system.
The bHLH transcription factor MYOR is expressed within the AVN during development and represses GATA4-dependent activation of the Cx30.2 enhancer [34]. HAND2 and GATA are established to physically interact to co-regulate transcription [35]. Given that Cx30.2 knockout mice also exhibit a shortened PR interval, by accelerating suprahissian impulse propagation [36], an interesting mechanism of action could be that HAND2 and GATA4 act at the Cx30.2 enhancer where loss of GATA4 or HAND2 results in a reduction in PR interval.
Another interesting observation is that we observe a sex-dependent effect on loss of HAND factors on the CCS, with a more severe phenotype being observed in male mutants compared to females ( Figure 2D, Tables S3 and S7, Figure S5). Although interesting, it is currently unclear what role sex plays in HAND-factor-mediated developments/maintenance of the CCS.
Overall, these experiments suggest that although HAND1 plays a critical role in CCS formation and morphogenesis, its deletion from the CCS after it is formed does not appear to significantly alter function. In contrast, HAND2 does appear to play a homeostatic role in maintaining the CCS, as HAND2 loss-of-function within the post-natal CCS results in abnormalities in AV conduction. Loss of both HAND alleles in the post-natal CCS results in an intermediate phenotype. This is an intriguing result as we would predict that loss of both factors would lead to a worse phenotype, suggesting that loss of HAND1 rescues the shortened PR-interval HAND2 phenotype. Further, our results support a role of HAND2 expression in the embryonic heart for CCS formation and development as H2LVCreCKO mice exhibit abnormalities in conduction across the cardiac conduction system.