Cardiac Functional and Structural Abnormalities in a Mouse Model of CDKL5 Deficiency Disorder

CDKL5 (cyclin-dependent kinase-like 5) deficiency disorder (CDD) is a severe neurodevelopmental disease that mostly affects girls, who are heterozygous for mutations in the X-linked CDKL5 gene. Mutations in the CDKL5 gene lead to a lack of CDKL5 protein expression or function and cause numerous clinical features, including early-onset seizures, marked hypotonia, autistic features, gastrointestinal problems, and severe neurodevelopmental impairment. Mouse models of CDD recapitulate several aspects of CDD symptomology, including cognitive impairments, motor deficits, and autistic-like features, and have been useful to dissect the role of CDKL5 in brain development and function. However, our current knowledge of the function of CDKL5 in other organs/tissues besides the brain is still quite limited, reducing the possibility of broad-spectrum interventions. Here, for the first time, we report the presence of cardiac function/structure alterations in heterozygous Cdkl5 +/− female mice. We found a prolonged QT interval (corrected for the heart rate, QTc) and increased heart rate in Cdkl5 +/− mice. These changes correlate with a marked decrease in parasympathetic activity to the heart and in the expression of the Scn5a and Hcn4 voltage-gated channels. Interestingly, Cdkl5 +/− hearts showed increased fibrosis, altered gap junction organization and connexin-43 expression, mitochondrial dysfunction, and increased ROS production. Together, these findings not only contribute to our understanding of the role of CDKL5 in heart structure/function but also document a novel preclinical phenotype for future therapeutic investigation.


Introduction
CDKL5 (cyclin-dependent kinase-like 5) deficiency disorder (CDD; OMIM 300203) is a rare and severe X-linked neurodevelopmental disease caused by mutations in the CDKL5 gene, which lead to a lack of CDKL5 protein expression or function [1][2][3]. CDD represents one of the most common genetic causes of epilepsy in infants [4], affecting females, who are heterozygous for CDKL5 deficiency due to random X-chromosome inactivation four times more often than males [5]. Due to the numerous clinical features that overlap with the more well-characterized Rett syndrome (RTT), it was initially termed as an "early seizure variant" or "Hanefeld variant" of RTT [6]. Similarities between RTT and CDD include severe neurodevelopmental and motor impairment, intellectual disability, and, in some cases, respiratory dysregulation. Nevertheless, the improvement of the clinical overview of CDD in the past few years has defined a more detailed phenotypic spectrum, including very common alterations of peripheral organ and tissue function, such

Prolonged QTc Interval and Elevated Heart Rate in Cdkl5 +/− Mice
We found that Cdkl5 is expressed in the mouse heart ( Figure 1A), albeit at lower levels than in the brain. Cdkl5 expression was reduced to 39 ± 5.6% in heterozygous (Cdkl5 +/−) female mice in comparison with wild-type (Cdkl5 +/+) mice ( Figure 1B) and was absent in Cdkl5 −/− mice ( Figure 1B).
Cdkl5 As expected, tachycardia also produced significant reductions in P wave duration and in QRS and JT intervals in Cdkl5 +/− mice compared to Cdkl5 +/+ mice (Supplementary Figure S1). Using validated mouse indices computed in the time domain on spontaneous RR fluctuations (pNN8 and MRSSD), we explored the vagal (parasympathetic) contribution to cardiac modulation ( Figure 1E,F and Table 1). Both indices were significantly reduced in Cdkl5 +/− and Cdkl5 −/− mice compared to Cdkl5 +/+ mice. No significant difference was found between Cdkl5 +/− and Cdkl5 −/− mice for any of these parameters. Since CDD female patients are all heterozygous for CDKL5 deficiency, subsequent analyses were carried out in heterozygous Cdkl5 +/− female mice. In order to investigate the influences of the autonomic nervous system on heart activity and to dissect the relative contribution of each autonomic arm in Cdkl5 +/− mice, we evaluated heart responses to selective pharmacological blockades of the parasympathetic and sympathetic systems elicited by IP injection of atropine or atenolol. As expected, the constant IP infusion of atenolol in Cdkl5 +/+ mice produced a significant increase in RR intervals, while atropine induced a significant decrease compared to saline infusion ( Figure 1G and Supplementary Figure S1E,F). Accordingly, in Cdkl5 +/+ mice, these drugs produced a significant increase (atropine) or decrease (atenolol) in QTc intervals compared to saline ( Figure 1H and Supplementary Figure S1E,F). Interestingly, while an effect on RR and QTc intervals was elicited in Cdkl5 +/− mice by atenolol infusion ( Figure 1G,H), atropine did not produce any significant variation of RR or QTc intervals compared to saline in Cdkl5 +/− mice ( Figure 1G,H).
To investigate whether the rate variability in Cdkl5 +/− mice is associated with an altered vagal/sympathetic component, we analyzed the expression of muscarinic 2 (Chrm2) and adrenergic beta 1 (Adrb1) receptors. We found that the mRNA levels of Chrm2 were significantly lower in the hearts of Cdkl5 +/− female mice than in those of wild-type (Cdkl5 +/+) mice ( Figure 1I), while no difference in the expression of the Adrb1 receptor was observed between Cdkl5 +/+ and Cdkl5 +/− female hearts ( Figure 1J).

Alterations of Voltage-Gated Channel Expression in the Hearts of Cdkl5 +/− Mice
To investigate the molecular alterations underlying the prolonged QTc interval and elevated heart rate in Cdkl5 +/− mice, we analyzed the expression of different voltage-gated channels involved in the action potential of cardiac myocytes. In particular, we focused our attention on the genes encoding for the voltage-gated potassium (Kcnq1, Kcnh2) and sodium (Scn5a) channels. Mutations in these genes are the most common causes of inherited long QT interval syndrome (LQTS), a condition characterized by a prolongation of the QT interval on the ECG, an increase in heart rate, and the risk of ventricular arrhythmias [29].
RT-qPCR analyses revealed no differences in the expression of Kcnq1, Kcnh2, or Kcnj2 channels between Cdkl5 +/+ and Cdkl5 +/− female hearts (Table 2); however, we did find that the mRNA levels of Scn5a were significantly lower in the hearts of Cdkl5 +/− female mice than in those of wild-type (Cdkl5 +/+) mice (Table 2).
To rule out any potential abnormalities in the sinoatrial node as possible contributing factors to the observed sinus tachycardia, Hcn4 expression was examined. Interestingly, we found that Hcn4 mRNA levels were significantly lower in Cdkl5 +/− female mice than in wild-type mice ( Table 2). Table 2. Comparative expression of ion channel cardiac genes between Cdkl5 +/+ and Cdkl5 +/− mice. mRNA expression levels of ion channel genes in the ventricle of 3-month-old Cdkl5 +/+ and Cdkl5 +/− mice. mRNA levels in tissues obtained from Cdkl5 +/+ mice were set to 1, and relative expression levels in Cdkl5 +/− mice are represented as mean ± SEM. Relative quantification was performed using the ∆∆Ct method. The means of two stable reference genes (Actin and Gapdh) were used as normalization factors. * p < 0.05; n.s., not significant (two-tailed Student's t-test).

Increased Cardiac Fibrosis in Cdkl5 +/− Mice
To determine whether the absence of Cdkl5 could impact the heart morphological structure in Cdkl5 KO mice, dissected hearts were weighed. Since the weight of the heart positively correlates with body weight and growth, each animal's body weight was measured before sacrifice, and the ration of heart weight to body weight (HW/BW) was calculated. No significant differences in heart weight, body weight, or HW/BW were found between Cdkl5 +/− and Cdkl5 +/+ mice (Table 3). Similarly, atrioventricular distance (Table 3) and the gross morphology of hearts (data not shown) revealed no differences in the chamber diameters of wild-type (Cdkl5 +/+) and Cdkl5 +/− mouse hearts.
Heart sections stained with hematoxylin and eosin showed normal architecture of cardiac myocytes (Figure 2A). However, Masson's trichrome-stained sections showed increased collagen deposition in the hearts of Cdkl5 +/− mice compared with wild-type (Cdkl5 +/+) mice ( Figure 2B). Increased fibrosis was accompanied by increased expression of fibrotic genes such as the α-2 chain of collagen type 1 (Col1a2) and the α-3 chain of collagen type 2 (Col3a1) ( Figure 2C). Table 3. Effect of genotype on heart weight and dimension. Heart weight, body weight, heart over body weight ratio, and atrioventricular distance of 3-4-month-old Cdkl5 +/+ (n = 18) and Cdkl5 +/− (n = 27) mice. Values are represented as mean ± SEM; n.s., not significant (two-tailed Student's t-test).  Given that cardiac fibrosis is mainly the result of proliferation and trans-differentiation of cardiac fibroblasts, we measured the content of vimentin, a marker of cardiac fibroblasts [30]. Western blot analysis showed significantly higher vimentin levels in the hearts of Cdkl5 +/− mice in comparison with wild-type (Cdkl5 +/+) mice ( Figure 2D).

Altered Gap Junction Organization and Connexin Expression in the Hearts of Cdkl5 +/− Mice
The electrical conduction in cardiac muscle relies on efficient gap-junction-mediated intercellular communication between cardiomyocytes, which involves rapid anisotropic impulse propagation through connexin (Cx)-containing channels, namely Cx43, the most abundant Cx in the heart.
Western blotting analysis revealed significantly diminished expression of Cx43 in Cdkl5 +/− mice in comparison with Cdkl5 +/+ mice ( Figure 3A,B). The mean density of the Cx43 band in Western blotting was reduced by 24% in Cdkl5 +/− mice compared to the Cdkl5 +/+ group ( Figure 3B). In parallel, the mRNA levels of Cx43 did not differ between Cdkl5 +/− and Cdkl5 +/+ mice ( Figure 3C), suggesting a post-transcriptional mechanism of Cx43 downregulation. Immunofluorescence labelling showed Cx43 expression to be mainly located at the intercalated discs ( Figure 3D). As compared to Cdkl5 +/+ mice, the Cx43 fluorescence in Cdkl5 +/− mice was slightly less intensive ( Figure 3D). It is interesting to note that some degree of lateralization of the Cx43 expression on the cardiomyocytes of Cdkl5 +/− mice was also present ( Figure 3D).
Interestingly, the expression of β-catenin protein, a component of the adherens junction of the intercalated disc, was increased in the myocardium of Cdkl5 +/− mice ( Figure 3E-H). Western blot analysis revealed that β-catenin levels were increased by 71.9% in Cdkl5 +/− mice in comparison with Cdkl5 +/+ mice ( Figure 3E,F), which correlated with the much stronger immunoreactivity for β-catenin at the intercalated discs in the hearts of Cdkl5 +/− mice ( Figure 3G

Increased AKT Activation in the Hearts of Cdkl5 +/− Mice
Recent evidence has described a potential role of AKT in aging-associated organ deterioration including cardiac hypertrophy and fibrosis [31]. In particular, aging enhances AKT phosphorylation in the mouse heart, while GSK-3β phosphorylation levels are unaffected by aging or AKT overexpression [31]. Evaluation of the phosphorylation levels of the AKT/GSK-3β pathway showed higher phosphorylation levels of AKT in the hearts of Cdkl5 +/− mice in comparison with Cdkl5 +/− mice ( Figure 4A,B) but no difference in phosphorylated GSK-3β levels between Cdkl5 +/+ and Cdkl5 +/− mice ( Figure

Increased AKT Activation in the Hearts of Cdkl5 +/− Mice
Recent evidence has described a potential role of AKT in aging-associated organ deterioration including cardiac hypertrophy and fibrosis [31]. In particular, aging enhances AKT phosphorylation in the mouse heart, while GSK-3β phosphorylation levels are unaffected by aging or AKT overexpression [31]. Evaluation of the phosphorylation levels of the AKT/GSK-3β pathway showed higher phosphorylation levels of AKT in the hearts of Cdkl5 +/− mice in comparison with Cdkl5 +/− mice ( Figure 4A,B) but no difference in phosphorylated GSK-3β levels between Cdkl5 +/+ and Cdkl5 +/− mice ( Figure 4C,D). Similarly, there was no difference in phosphorylated extracellular regulated kinases (ERK) 1 and 2 in the hearts of Cdkl5 +/− mice in comparison with Cdkl5 +/+ mice ( Figure 4C,D).  and total AKT levels normalized to GAPDH levels (right histogram) in ventricular extracts from Cdkl5 +/+ (n = 5) and Cdkl5 +/− (n = 5) mice. (B) Examples of immunoblots for P-AKT, AKT, and GAPDH for three mice from each experiential group. (C) Western blot analysis of P-GSK-3β and P-ERK1/2 levels in ventricular extracts of mice as in (A). Histograms on the left show P-GSK-3β (upper) and P-ERK1/2 (lower) levels normalized to corresponding total protein content. Histograms on the right show ERK1/2 and GSK-3β protein levels normalized to GAPDH levels. (D) Examples of immunoblots for P-GSK-3β, GSK-3β, P-ERK1/2, ERK, and GAPDH of three animals from each experimental group. Data are expressed as a percentage of expression in Cdkl5 +/+ mice. Values represent mean ± SEM. ** p < 0.01 (two-tailed Student's t-test).

Mitochondrial Dysfunction in the Hearts of Cdkl5 +/− Mice
There is growing evidence that mitochondrial dysfunction contributes to the development and progression of cardiac fibrosis [32]. To address whether mitochondrial function is affected in the Cdkl5-deficient heart, we first measured the ATP and ADP content in Cdkl5 +/− mouse heart homogenates using HPLC. This analysis detected a 35% decline in total ATP content in heart tissue from Cdkl5 +/− mice compared to controls ( Figure 5A).
Moreover, in Cdkl5 +/− mice, we found a significant decrease in the ATP/ADP ratio compared to wild-type animals ( Figure 5B), suggesting a diminished cell energy status.
Since mitochondria are the major site of ATP production within the cell, we evaluated whether there were alterations in the process of oxidative phosphorylation (OXPHOS) in Cdkl5 +/− mice. For this purpose, we isolated intact mitochondria from mouse hearts and evaluated the oxygen consumption in the presence of glutamate-malate or succinate under non-phosphorylating conditions (state 4) and in the presence of ADP (state 3, phosphorylating condition). Glutamate-malate and succinate donate electrons to the respiratory chain at complex I and complex II, respectively. State 3 of respiration, in the presence of glutamatemalate or succinate, was significantly decreased in cardiac mitochondria from Cdkl5 +/− compared to wild-type mice ( Figure 5C). In accordance with the oxygen consumption data, the spectrophotometric analysis of the single respiratory enzyme activity in isolated cardiac mitochondria provided evidence supporting reduced activity of succinate CoQ reductase (Complex II) in Cdkl5 +/− mice in comparison with Cdkl5 +/+ mice ( Figure 5D). The complex I activity in Cdkl5 +/− heart mitochondria showed a trend of reduction compared to controls, although it did not reach statistical significance ( Figure 5D).
The decrease in the activity of mitochondrial respiratory chain enzymes prompted us to assess whether overall mitochondrial content, size, and distribution were altered in the hearts of Cdkl5 +/− mice. We found no significant difference in the density of mitochondria in the myocytes (data not shown), nor was there any difference in the overall size of individual mitochondria ( Figure 5E). However, crystalline-like inclusions in the intracristal space were found in mitochondria of Cdkl5 +/− mice ( Figure 5E), indicating mitochondrial morphological abnormalities in the Cdkl5 +/− condition.

Increased ROS Production in the Hearts of Cdkl5 +/− Mice
Mitochondrial dysfunction is often associated with increased oxidative stress [33]. The level of reactive oxygen species (ROS)production determined using the fluorogenic probe DFCDA was significantly higher in Cdkl5 +/− mitochondria energized with succinate in comparison with the Cdkl5 +/+ condition ( Figure 6A). Moreover, ROS production in Cdkl5 +/− heart mitochondria treated with antimycin A, a specific inhibitor that induces reactive oxygen species production from complex III, was significantly higher in comparison with the antimycin-A-treated Cdkl5 +/+ condition ( Figure 6A). The mitochondria isolated from Cdkl5 +/− mouse hearts also showed higher levels of malondialdehyde (MDA), an end product of lipid peroxidation, consistent with a condition of oxidative stress ( Figure 6B). bars: low magnifications = 2 μm, high magnifications = 200 nm. The results in (A-D) are presented as means ± SEM; in (A), ( * ) p = 0.055 (Fisher's LSD test after two-way ANOVA); in (B-D), * p < 0.05, ** p < 0.01 (two-tailed Student's t-test).

Increased ROS Production in the Hearts of Cdkl5 +/− Mice
Mitochondrial dysfunction is often associated with increased oxidative stress [33]. The level of reactive oxygen species (ROS)production determined using the fluorogenic probe DFCDA was significantly higher in Cdkl5 +/− mitochondria energized with succinate in comparison with the Cdkl5 +/+ condition ( Figure 6A). Moreover, ROS production in Cdkl5 +/− heart mitochondria treated with antimycin A, a specific inhibitor that induces reactive oxygen species production from complex III, was significantly higher in comparison with the antimycin-A-treated Cdkl5 +/+ condition ( Figure 6A). The mitochondria isolated from Cdkl5 +/− mouse hearts also showed higher levels of malondialdehyde (MDA), an end product of lipid peroxidation, consistent with a condition of oxidative stress (Figure 6B). production in isolated heart mitochondria from Cdkl5 +/+ and Cdkl5 +/− mice measured using the fluorescent probe DCFDA. The mitochondria were energized with succinate, and ROS production was determined by following the oxidation of H 2 DCF to its fluorescent form DCF. The specific inhibitor antimycin A was used to promote ROS production from mitochondrial complex III (n = 6). Data are expressed as fold change relative to controls and normalized to protein content; ** p < 0.01; *** p < 0.001 (Fisher's LSD test after two-way ANOVA). (B) Measurement of the levels of lipid peroxidation biomarker malondialdehyde (MDA) in isolated heart mitochondria from Cdkl5 +/+ and Cdkl5 +/− mice (n = 3). Data are expressed as fold change relative to controls and normalized to protein content; ** p < 0.01 (two-tailed Student's t-test). (C,D) Western blot analysis of poly(ADPribose) polymerase 1 (PARP1) levels in extracts of ventricular heart tissue from Cdkl5 +/+ (n = 5) and Interestingly, two proteins, the Poly(ADP-ribose) polymerase 1 (PARP1) and the nuclear factor erythroid 2-related factor 2 (Nrf2), that are known to be activated due to oxidative stress showed increased levels in the hearts of Cdkl5 +/− mice compared to Cdkl5 +/+ mice ( Figure 6C-F). Similarly, we found an increase in the LC3-II/LC3-I ratio ( Figure 6G,H), which is indicative of increased autophagosome formation, following oxidative stress.

Discussion
CDD is a very severe and debilitating neurodevelopmental infantile disorder with harsh neurological symptoms such as intractable seizures, neurodevelopmental delay, and autistic-like features. Nevertheless, the improvement of the clinical overview of CDD in the past few years has defined a more detailed phenotypic spectrum; this includes very common alterations in peripheral organ and tissue function, such as gastrointestinal problems, irregular breathing, hypotonia, and scoliosis [32], suggesting that CDKL5 deficiency compromises not only CNS function but also that of other organs/tissues. Here, we report, for the first time, that a mouse model of CDD, the heterozygous Cdkl5 KO (Cdkl5 +/−) female mouse, exhibits cardiac functional and structural abnormalities. Cdkl5 +/− mice exhibited QTc prolongation and increased heart rate accompanied by impaired cardiac autonomic control. Moreover, the Cdkl5 +/− heart shows typical signs of heart aging, including increased fibrosis, mitochondrial dysfunctions, and increased ROS production.
ECG measurements from Cdkl5 +/− mice showed statistically significant increases in heart rate and in rate-corrected QTc intervals, which are indicative of delayed ventricular depolarization and repolarization. This is in line with a recent finding showing that in a cohort of individuals with CDD, the incidence of prolonged QTc or other abnormalities including sinus tachycardia was higher than the prevalence seen within the general population [26]. However, as mentioned by the authors, this first patient study has some important limitations: it was retrospective in nature, and the sample size was small [26]. Here, using an experimental approach that allows for continuous ECG recording during sleep in a validated mouse model of CDD, we overcame these limitations and, importantly, deeply reduce environmental confounders besides the genetic confounder. We found that, similarly to heterozygous Cdkl5 +/− mice, homozygous Cdkl5 −/− female mice exhibited tachycardia and a longer QTc interval in comparison with Cdkl5 +/+ mice. Interestingly, the magnitude of the cardiac defects was similar in the two Cdkl5-deficient conditions. This is not surprising, since it has been shown that heterozygous Cdkl5 +/− mice develop some behavioral abnormalities that are comparable to defects identified in homozygous Cdkl5 −/− females [33]. The putative skewing toward the mutated X chromosome found in the heart of heterozygous females (levels of Cdkl5 decreased to 39%) could explain the similar cardiac phenotypic outcome show by heterozygous and homozygous Cdkl5 KO female mice.
It is well known that neurological dysfunction may affect the control of cardiac rate and rhythm [34]. In particular, it has previously been described that autonomic neuropathies prolong QTc intervals in patients with CNS disease, including Rett syndrome [27,[35][36][37][38]. Therefore, we tested the hypothesis that cardiac autonomic control is impaired in Cdkl5 +/− mice. By applying validated indices of spontaneous RR variability in mice, we assessed cardiac vagal modulation during sleep under baseline conditions and following pharmacological manipulations. Our findings that in Cdkl5 +/− mice, pNN8 and RMSSD indices were lower than in control mice indicate an impairment in the vagal modulation of the heart period in Cdkl5 +/− mice. Compared to saline infusion, treatment with muscarinic antagonist (atropine) did not significantly modify the heart rate or QTc interval in Cdkl5 +/− mice, confirming the reduction in parasympathetic tone in the heart. A growing body of evidence indicates that reduced cardiac parasympathetic activity is a common alteration in brain disorders; it is present in children with autism and Rett syndrome [39,40], as well as in neurological conditions characterized by neuronal degeneration [41][42][43]. The mechanism underlying this autonomic alteration is partly unknown [44,45]; however, reduced cardiac vagal tone is thought to be prodromic for lethal arrythmias and sudden death. The observed reduction in cardiac M 2 receptor expression in the heart of Cdkl5 +/− mice may be one of the possible mechanisms underlying the parasympathetic dysfunction in the mouse model of CDD. However, since, in many situations, loss of Cdkl5 function appears to lead to impaired neuronal activity [16,18,46], we cannot exclude that an overall reduction in activity in regions of the brainstem that are important for cardiorespiratory function may underlie this autonomic dysfunction. Regarding the sympathetic regulation of the heart, similarly to wild-type mice, Cdkl5 +/− hearts responded to atenolol infusion by increasing RR and decreasing QTc intervals, indicating the presence of a preserved sympathetic modulation in the Cdkl5 +/− heart. Interestingly, after atenolol infusion, RR and QTc intervals no longer differed between genotypes (Supplementary Figure S1E,F). Since no indices of sympathetic cardiac control have been validated for mouse heart rhythm variability to, at the moment, we cannot exclude an impairment of this autonomic branch in modulating heart rhythm in Cdkl5 +/− mice.
The discovery that two cardiac-specific genes (Scn5a and Hcn4) are altered in their expression in Cdkl5 +/− female hearts indicates that CDKL5, either directly or indirectly, regulates the expression of genes that play a role in the cardiac conduction systems at the cardiomyocyte level. Loss-of-function mutations in the SCN5A gene, which encodes the α subunit of the cardiac voltage-gated Na+ channel NaV1.5, underlie cardiac disorders, including long QT syndrome [47]. Notably, long QT has been described in almost 20% of Rett syndrome patients, and alterations in the expression of Scn5a genes have been reported in Mecp2-null mice [48]. The lower expression of the cardiac pacemaker-specific channel Hcn4 found in Cdkl5 +/− hearts could also contribute to the altered heart rate in Cdkl5 +/− mice. Our results suggest that Cdkl5 deficiency selectively affects the expression of voltage-gated channel genes in the heart and that abnormal cardiac gene expression may be an arrhythmogenic substrate in Cdkl5 +/− mice.
Immunostaining for Cx43 in Cdkl5 +/− ventricular tissues revealed a significant reduction in Cx43 levels at cell-cell junctions, as confirmed by Western blot analysis. Deregulation of β-catenin levels was also present in the intercalated discs of Cdkl5 +/− cardiomyocytes. Both Cx43, as a major component of gap junctions [49], and β-catenin, as a component of the gap junctions and intercalated discs, contribute to the regulation of the transmission of electrical signals through cardiac myocytes. Therefore, the present findings strongly suggest an impairment of intercellular communication in Cdkl5 +/− cardiomyocytes, which could further be considered as an arrhythmogenic substrate in Cdkl5 +/− hearts. Indeed, heterogeneous reduction in Cx43 expression and altered patterns of gap junction distribution are features of human ventricular disease and correlate with electrophysiologically identified arrhythmic changes in animal models [49,50]. It is of note that age-related disorganization of intercalated discs, including increased β-catenin expression, which may be responsible for the slower conduction of the depolarization wave within the heart, have been recently described [51], suggesting an accelerated cardiac senescence in Cdkl5 +/− mice.
One of the main risk factors for cardiovascular diseases is aging [52]. Previously, we showed that Cdkl5 KO mice are characterized by an increased rate of apoptotic cell death during brain aging due to accelerated neuronal senescence, a factor that causes a consequent age-related cognitive and motor decline [18]. Here, we observed a similar accelerated senescence in the heart of Cdkl5 +/− mice. We found signs of increased cardiac fibrosis and increased extracellular matrix deposition of collagen and vimentin expression. Premature senescence develops through various external and internal stress signals, including energetic dysfunction, giving rise to free radical reactive oxygen species (ROS) that cause damage to cellular macromolecules; accumulation of this damage leads to the physiological compromise seen in aging [53]. Current evidence suggests that mitochondrial dysregulation is the cause and primary target of energetic dysfunction and free radical production [54].
In this study, we found a decrease in mitochondrial oxygen consumption rate in isolated mitochondria from Cdkl5 +/− hearts when ADP was supplied as a substrate for the ATP synthase (state 3 respiration) and diminished specific activity of complexes I and II. These data are consistent with the decreased level of ATP and the decreased ATP/ADP ratio measured in heart tissue homogenates from Cdkl5 +/− mice compared with wildtype animals. In addition, mitochondria isolated from the hearts of Cdkl5 +/− mice showed increased ROS and lipid peroxidation biomarker (MDA) production compared with controls, suggesting a link between mitochondrial dysfunction and the onset of oxidative and energetic stress.
These findings are in line with recent evidence that mitochondrial dysfunction and oxidative stress occur in CDD [55,56]. Cytokine dysregulation, inflammatory status, oxidative stress marker 4HNE-Pas, and redox imbalance were evidenced in plasma from CDD patients [56][57][58]. Additionally, studies in Cdkl5 KO mice identified brain mitochondrial functional abnormalities, including reduced activity of mitochondrial respiratory chain complexes and impairment in mitochondrial ATP production rate [59,60], as well as alterations in patient-derived iPSCs [61,62]. It is worth noting that we found mitochondrial structural changes that were consistent with the development of intramitochondrial crystalline-like inclusions in the Cdkl5 +/− heart. The presence of mitochondrial paracrystalline inclusions in a clinically characterized group of patients with genetically defined mitochondrial disease [63,64] suggests that this ultrastructural alteration could contribute to mitochondrial function. Therefore, we can speculate that the crystalline inclusions found in the inner membrane may underlie the bioenergetic defects found in Cdkl5 +/− mitochondria. However, we cannot exclude the possibility that the crystalline-like inclusions represent compensatory responses to mitochondrial stress.
Recent evidence has described a potential role of AKT and autophagy, molecular mechanisms of cellular senescence, in aging-associated organ deterioration. When cardiomyocytes incur oxidative stress, they activate the PI3K/AKT signaling pathway [65], suggesting that oxidative stress is an upstream event related to the activation of the AKT pathway in cardiomyocytes. Our finding that AKT pathway activation is increased in the hearts of Cdkl5 +/− female mice could be a consequence of an increase in the generation of ROS due to aberrant mitochondrial function. Similarly, increased LC3-II levels, which are associated with either enhanced autophagosome synthesis or reduced autophagosome turnover [66], may be associated with increased ROS production and the resulting oxidative cell stress that occurs in many disease states [67]. We found that the ratio of LC3-II/LC3-I was enhanced in the Cdkl5 +/− heart, suggesting an increasing number of autophagosomes.
The principal growth-promoting intracellular signaling pathways that are activated by ROS in cardiac myocytes include not only the AKT pathway but also the mitogen-activated protein kinase cascades (ERK1/2 pathway) [65]. In association with Cdkl5 +/− cardiac myocyte dysfunctions, we found alterations in AKT activity but not in ERK1/2 signaling. Activation of ERK1/2 is generally associated with cell growth and survival, and studies of transgenic mice have shown that selective activation of the ERK1/2 cascade in the myocardium induces adaptive cardiac hypertrophy [68]. Further evidence linking ERK1/2 with hypertrophy was reviewed in [68]. Since no signs of hypertrophy are evident in the Cdkl5 +/− heart, this might explain the lack of ERK1/2 dysregulation in these mice.
Considering the protective molecular pathways that might be triggered by oxidative stress [69] and autophagy [70], we found increased levels of Nrf2, a transcription factor known to activate multiple enzymes with antioxidant properties [71], as well as increased levels of PARP1, a DNA damage sensor that facilities base excision repair [72] in the Cdkl5 +/− heart. We hypothesized that increased levels of Nrf2 are triggered by ROS production and that this increase is most likely a compensatory effort to increase antioxidant defenses in the Cdkl5 +/− context. This should correlate with a reduction in oxidative stress in Cdkl5 +/− hearts; however, this is not the case, and the issue requires further investigation. It is worth noting that PARP activation has been demonstrated to impair mitochondrial function [73] and promote autophagy in cardiomyocytes [74]. Therefore, increased PARP1 levels may have contributed to the pathologic signs observed in the Cdkl5 +/ − heart.
At present, we have no evidence for which of the observed Cdkl5-related alterations is the primary cause underlying the structural and functional abnormalities in the Cdkl5 +/− heart. However, increasing evidence suggests that ROS production is associated with cardiac arrhythmias; in particular, elevated cellular ROS can cause alterations of the cardiac ion channels, changes in mitochondrial function, and gap junction remodeling, leading to arrhythmic conditions [75][76][77]. Therefore, oxidative stress, a common pathophysiological factor in cardiac disease [78][79][80], may be the main defect underlying cardiac alterations in the Cdkl5-deficient heart. Future studies aimed at rescuing the cardiac phenotype in Cdkl5 KO mice, targeting one of the observed defects, could help to shed light on the primary causes that drive cardiac abnormalities in the absence of Cdkl5.

Animal Husbandry
The mice used in this work were derived from the Cdkl5 −/Y strain in the C57BL/6N background developed in [8] and backcrossed in C57BL/6J for three generations. Heterozygous Cdkl5 +/− and homozygous Cdkl5 −/− females were produced and genotyped as previously described [8], and age-matched wild-type Cdkl5 +/+ littermate controls were used for all experiments. The day of birth was designated as postnatal day (P) zero, and animals of 24 h of age were considered as 1-day-old animals (P1). After weaning (P21-23), mice were housed three to five per cage with a 12 h light/dark cycle in a temperature-and humidity-controlled environment with food and water provided ad libitum. The animals' health and comfort were controlled by the veterinary service. Experiments were carried out on a total of 121 adult (3-4-month-old) Cdkl5 KO mice (Cdkl5 +/+ n = 50; Cdkl5 +/− n = 62; Cdkl5 −/− n = 9). The study protocols complied with EU Directive 2010/63/EU and with Italian law (DL 26, 4 March 2014) and were approved by the Italian Ministry of Health (protocol n • 535/2022-PR). All efforts were made to minimize animal suffering and to keep the number of animals used to a minimum.

Surgical Procedure, In Vivo Recording, and Data Analysis
Eight Cdkl5 +/+, 10 Cdkl5 +/−, and 8 Cdkl5 −/− female mice were instrumented with electroencephalographic (EEG), electromyographic (EMG), and electrocardiographic (ECG) electrodes to characterize the electrical activity of the heart during sleep under baseline conditions. A second group of mice (Cdkl5 +/+ n = 10, Cdkl5 +/− n = 10) instrumented with the same surgical protocol also underwent surgery to implant an intraperitoneal (IP) catheter for the continuous infusion of autonomic blockers with the purpose of testing the autonomic modulation of heart rhythm.
All mice were deeply anesthetized with isoflurane (1.8-2.4% in O 2 , inhalation route) and treated with intraoperative analgesia (carprofen 4 mg/kg subcutaneously, Pfizer, Italy) and postoperative antibiotic prophylaxis (benzylpenicillin benzathine, 12,500 IU/kg, and dihydrostreptomycin sulphate, 5 mg/kg, subcutaneously). For the recording of the EEG signal, mice were implanted with two miniature stainless-steel screws (2.4 mm length, Plas-ticsOne, Roanoke, VA, USA) in contact with the dura mater (frontoparietal derivation); for the recording of the EMG signal, mice were implanted with two multistranded PFA-coated stainless-steel wires (KF Technology srl, Roma, Italy) inserted into the nuchal muscles. For ECG recording, two PFA-coated stainless-steel wires were inserted subcutaneously; one was put in contact with muscles in the right-upper quadrant of the thorax, while the second touched the abdominal muscles on the left flank. All electrode wires were then collected in a socket placed over the mouse's head and fixed with dental cement (RelyX Unicem, 3M ESPE, Pioltello, (MI), Italy) and dental acrylic (Respal NF, SPD, Mulazzano (LO), Italy). The second batch of mice was also implanted with a silicone catheter [81,82] with the tip inserted into the abdominal cavity and the other extremity tunneled to the mouse head and fixed with the abovementioned socket. After 12 days of postoperative recovery and habituation to the recording apparatus (ambient temperature set at 25 • C), EEG, EMG, and ECG signals of mice included in the baseline protocol were continuously recorded for 24 h. Signals were transmitted via a cable connected to a rotating electrical commutator (SL2 + 2C/SB, Plastics One, USA) and to a balanced cable suspensor, allowing the mice to make unhindered movements [83].
Mice included in the protocol for testing of the autonomic control of heart rhythm were allowed to recover from surgery and to habituate to the recording settings for 14 days. Each mouse underwent 3 recording sessions (7 h each starting at light on) with acquisition of EEG, EMG, and ECG signals while being continuously infused with either saline, with the muscarinic receptor antagonist atropine methyl nitrate (Vinci-Biochem, Italy, 0.5 mg/mL in saline) to block parasympathetic activity to the heart, or with the selective β1-adrenergic receptor antagonist atenolol (Sigma-Aldrich, Milano, Italy, 0.25 mg/mL in saline) to block the sympathetic activity to the heart. Each mouse first received saline infusion and was then randomly subjected to atropine and atenolol infusions. Each infusion was performed at least 48 h after the preceding infusion. The IP catheter was connected to a remote infusion pump (model 22 multiple syringe pump, Harvard Apparatus, Cambridge, MA, USA) by an external tube prefilled with saline solution or drug solution, as previously described [81,82]. A rapid infusion was performed before the start of each recording session at a rate of 30 µL/min for 5 min to fill the IP catheter with saline or either of the drug solutions; then, the infusion rate was set at 100 µL/h for 7 h. At the end of the recording session, mice were sacrificed under deep anesthesia (isoflurane 4% in O 2 ).
The EEG signal was cut with a band pass filter between 0.3 and 100 Hz and stored at 128 Hz. The EMG signal was filtered between 100 and 1000 Hz and stored at 128 Hz, while the ECG was filtered between 10 and 1000 Hz and stored at 2048 Hz. Data acquisition was performed with LabVIEW 8.0 software (National Instruments, Austin, TX, USA).
The EEG and EMG signals were imported and analyzed to automatically score the wake-sleep states (wakefulness, rapid-eye-movement sleep (REMS) and non-rapid-eyemovement sleep (NREMS)). On the contrary, the ECG signal was imported in Labchart 8.0 (ADInstruments, Colorado Springs, CO, USA), a specific notch (50 Hz) filter was digitally applied to exclude electrical noise, and the signal was then analyzed with the ECG analysis module to automatically detect QRS complexes and their intervals. For each mouse, ECG analysis was restricted to 8 episodes of NREMS and 8 episodes of REMS (longer than 30 s) homogeneously distributed throughout the whole recording. For each session, we calculated the following ECG parameters: P-wave duration, QRS, RR, PR, JT, and QT intervals. To compensate for the elevated mouse heart rate, we corrected the QT interval (QTc) with Hodges' formula (the most effective for rodents, [84]).
The vagal (parasympathetic) contribution to cardiac modulation was examined using validated mouse indices computed in the time domain on spontaneous RR fluctuations [85]: pNN8, % of RR values that differ from the following values by >8 ms [86]; RMSSD, root mean square of the successive RR differences [87]. To date, no index of spontaneous RR fluctuation in the time domain has been validated for the sympathetic contribution to cardiac modulation.

Heart Dissection, Measurement, and Collection
Adult Cdkl5 −/−, Cdkl5 +/−, and Cdkl5 +/+ female mice aged 3-4 months were weighed and put under deep anesthesia through inhalation of 2% isoflurane in pure oxygen and sacrificed through cervical dislocation. Hearts were quickly removed, cleaned from the surrounding structures, and thoroughly washed in PBS to remove all blood, then weighed. The ratio of heart weight to body weight (HW/BW) was then calculated by dividing the weight of the heart by the weight of the whole animal. The atrioventricular distance was measured. All measurements were performed by the same person with the same precision scales. Hearts were quickly frozen in isopentane, cooled in liquid nitrogen, and stored at −80 • C until used for RT-qPCR, immunohistochemistry, and Western blot analyses.

RNA Isolation and RT-qPCR
RNA isolation and RT-qPCR were conducted on frozen hearts of Cdkl5 −/−, Cdkl5 +/−, and Cdkl5 +/+ female mice and on frozen cortices of Cdkl5 +/+ female mice. Total RNA was isolated using the TRI reagent method (Sigma-Aldrich, St. Louis, MO, USA), and cDNA synthesis was achieved with 5 µg of total RNA using an iScript™ advanced cDNA synthesis kit (Bio-Rad, Hercules, CA, USA) according to the manufacturer's instructions. Real-time PCR was performed using SsoAdvanced Universal SYBR Green Supermix (Bio-Rad) in a CFX real-time PCR detection system (Bio-Rad). We used primer pairs (Supplementary Table S1) that provided an efficiency close to 100%. Each biological replicate was run in triplicate. The mean of the two more stable reference genes (Actb and Gapdh) were used as a normalization factor in the RT-qPCR analysis, and relative quantification was performed using the ∆∆Ct method.

Image Acquisition and Measurements
Fluorescence images were taken with an Eclipse TE 2000-S microscope equipped with a DS-Qi2 digital SLR camera (Nikon Instruments Inc.). A light microscope (Leica Mycrosystems) equipped with a motorized stage and focus control system and a color digital camera (Coolsnap-Pro, Media Cybernetics) were used to take brightfield images of Masson's trichrome and hematoxylin eosin-stained sections.

Quantification of β-Catenin Staining Intensity and Areas
Starting from 20× magnification images of β-catenin-stained ventricular slices, the area of β-catenin staining in intercalated discs was manually drawn using the Image Pro Plus measurement function and expressed in µm 2 . The intensity of β-catenin staining within each area was then quantified by determining the sum intensity of all positive (bright) pixels within the area. Approximately 100 intercalated discs were analyzed from each sample.

Western Blotting
For the preparation of protein extracts, ventricles were homogenized in RIPA buffer and quantified using the Bradford method as previously described [88].
Equivalent amounts (50 µg) of protein were subjected to electrophoresis on a 4-12% Mini PROTEAN ® TGX™ Gel (Bio-Rad) and transferred to a Hybond ECL nitrocellulose membrane (GE Healthcare Bio-Science). The primary and secondary antibodies used are listed in Supplementary Table S2. The densitometric analysis of digitized Western blot images was performed using Chemidoc XRS Imaging Systems and Image Lab TM software (Bio-Rad), which automatically highlights any saturated pixels of the Western blot images in red. Images acquired with exposition times that generated protein signals outside of a linear range were not considered for quantification.

Adenine Nucleotide Measurement
For the determination of adenine nucleotides, mouse hearts were quickly removed and placed in ice-cold PBS supplemented with 10 mM EDTA. The tissue was homogenized using an Ultra-Turrax (1 cycle for 10 s); then, the resuspension was transferred to a glass potter with a Teflon pestle operated at 1600 r.p.m (10 strokes). All steps were performed at 4 • C. ATP and ADP were extracted and detected as previously described [89] using HPLC (Agilent 1100 series system) using a Kinetex C18 column (250 × 4.6 mm, 100 Å, 5 µm; Phenomenex). Nucleotide peaks were identified at λ = 260 nm by comparison and coelution with the standards. Different nucleotides were quantified through peak area measurement compared with standard curves.

Mitochondrial Oxygen Consumption
Intact mitochondria from mice hearts were isolated as previously described [90]. Briefly, freshly prepared mitochondria were assayed for oxygen consumption at 30 • C by means of a thermostatically controlled oxygraph apparatus (Instech Mod. 203, Plymouth Meeting, PA, USA) equipped with a Clark electrode and a rapid mixing device. Mitochondria (0.2 mg protein) were incubated in the respiration medium (10 mM Tris/HCl, 5 mM MgCl 2 , 2 mM Pi, 20 µM EGTA and 0.25 M sucrose, pH 7.4) in a final volume of 1600 µL. Respiratory substrates (5 mM glutamate-malate or 13 mM succinate) were added after signal stabilization. State 4 respiration was recorded for 2 min, and state 3 respiration was induced by the addition of 300 µM ADP. The respiratory rates were expressed in nmol oxygen/min/mg of protein.

Measurement of Lipid Peroxidation
Lipid peroxidation in isolated mitochondria from Cdkl5 +/− and Cdkl5 +/+ female mice was assessed by measuring the biomarker malondialdehyde (MDA) as previously described by Reilly et al. [94]. Briefly, quantification of MDA was performed through reaction with thiobarbituric acid (TBA), and measurement of TBA-MDA adduct was carried out at 535 nm using a Jasco V-750 spectrophotometer. 1,1,3,3-tetramethoxypropane (Sigma-Aldrich, St. Louis, MO, USA) was used as a standard.

Transmission Electron Microscopy
For ultrastructural characterization, small heart specimens were fixed in 2.5% glutaraldehyde in 0.1 M cacodylate buffer and post-fixed in osmium tetroxide 1% in the same buffer. After dehydration in graded ethanol, specimens were embedded in araldite. Thin sections were counterstained with uranyl acetate and lead citrate and observed under a Philips CM100 transmission electron microscope (Philips, Amsterdam, the Netherlands).

Statistical Analysis
Statistical analysis was performed using GraphPad Prism (version 9). Values are expressed as means ± standard error (SEM). The significance of results was obtained using a two-tailed Student's t-test or one-way and two-way ANOVA, followed by Fisher's LSD post hoc test, as specified in the figure legends. For the in vivo recordings under the baseline condition, data were analyzed with 2-way repeated-measure ANOVAs (considering sleep state and genotype as factors) and corrected for multiple comparisons with Sidak's test. For mice included in the pharmacological testing of autonomic cardiac contribution, we first calculated the % modification of RR and QTc values induced by the drug compared to saline infusion. Then, for each experimental group, we performed a one-sample t-test, checking for significant deviations from the reference value (100%). A probability level of p < 0.05 was considered to statistically significant. The confidence level was taken as 95%.

Conclusions
In conclusion, the present study shows that Cdkl5 deficiency impacts heart structure and function in mice and that Cdkl5 +/ − mice could be a precious asset to accelerate the comprehension of these aspects. As our work does not provide direct causal evidence linking local CDKL5 expression in the heart or deregulation of direct CDKL5 phosphorylation targets to the electrical activity and structure of the cardiomyocyte, further in vitro and in vivo studies of the effects of specific knockout/rescuing CDKL5 expression in cardiomyocytes are needed. However, the results obtained in the present study could have a considerable impact in promoting a more in-depth investigation of cardiac function in patients for a better understanding of the CDD phenotypes. Although it is known that protein kinases can control cellular signaling by interacting with many targets, our knowledge of the CDKL5 phosphorylation targets is still very limited, hindering the investigation of direct CDKL5-dependent mechanisms underlying cardiac alterations in Cdkl5 +/− mice. However, many alterations found in the heart of Cdkl5 +/− mice are also present in the RTT mouse model, Mecp2-null mice [48,95,96], and/or in RTT patients [97][98][99], suggesting that similar molecular alterations may underlie the cardiac dysfunctions in these conditions, with several overlapping phenotypic features. Evidence that CDKL5 phosphorylates MeCP2 in vitro [100,101] suggests that MeCP2-dependent transcription regulation may be influenced by CDKL5. However, due to conflicting results [102,103], the molecular relationship between CDKL5 and MeCP2 remains to be clarified. A future in-depth gene expression analysis in CDKL5 and MeCP2-null conditions could be crucial in order to clarify the similarities and differences between CDD and RTT in terms of their cardiac phenotypes. Funding: This work was supported by a contribution from the Italian parent association "CDKL5 insieme verso la cura" to E.C.
Institutional Review Board Statement: The animal study protocol was approved by the Italian Ministry for Health and by the Bologna University Bioethical Committee (protocol code: AEDB0.26_ 535/2022-PR; date of approval: 8 September 2022).

Informed Consent Statement: Not applicable.
Data Availability Statement: The datasets analyzed during the current study are available from the corresponding author upon reasonable request.