Loss of A1BG and Sex-Differential Cardiac Effects on the Left Ventricle
Female mice homozygous for a cardiac muscle-specific conditional loss-of-function mutation in A1BG (A1bgCM/CM) have been reported to exhibit cardiac dysfunction, characterized by a failure of the left ventricle wall to interact with the intraventricular septum properly [33]. To verify and expand upon these findings, we conducted a histological examination, which revealed that the left ventricles in the hearts of female mice increased in size, an observation not seen in their male counterparts (Fig. 1A-D). Echocardiogram analyses corroborated this result (Fig. 1E, F). Additionally, these analyses indicated that in association with a dilated left ventricle, female but not male A1bgCM/CM mice had a reduction in the left ventricular posterior wall thickness during both diastole and systole in females but not in males (Fig. 1E, F). These findings suggest that the absence of A1BG in female CMs, in contrast to males, induces significant cardiac remodeling and a pathology that is associated with DCM.
Female hearts, but not male hearts, regulate the expression of genes related to cardiac metabolism and DCM
To understand the molecular underpinnings of A1BG sex-differential effects, transcriptional profiling via RNA sequencing (RNA-seq) was performed on wild-type and A1bgCM/CM male and female hearts at four weeks of age. This analysis in female hearts identified 122 differentially expressed genes (DEGs; adjusted P-value ≤ 0.05 and log2 fold change ≥ ± 1), of which 61 genes were significantly upregulated, and 61 genes were significantly downregulated compared with controls (Fig. 2A). Though male hearts do not appear to require A1BG, we find that analysis in male hearts identified 311 differentially expressed genes, of which 162 genes were significantly upregulated and 149 genes were significantly downregulated in A1bgCM/CM male hearts compared with controls (Fig. 2B). Consistent with phenotypic analysis, we find very few genes that are co-regulated in male and female A1bgCM/CM hearts: 10/200 downregulated, and 7/216 were upregulated (Fig. 2C). Therefore, supporting a sex-specific requirement for A1BG in CMs.
Pathway analyses of control versus A1bgCM/CM female hearts revealed a potential role for A1BG in Acetyl-CoA and glucose-6-phosphate metabolism, i.e., “monocarboxylic acids metabolic process” (Fig. 2D). Genes dysregulated in A1bgCM/CM female hearts include Acsl6, Adpgk, Gck, Ankrd23, Aldob, Fah, Acsf2, and Acsm5. None of these genes was dysregulated in male A1bgCM/CM hearts. (Fig. 2E). These findings are significant because CMs in DCM have a higher dependence on glucose oxidation [42–47]. In addition to these genes, we identified 4 genes that were downregulated in the A1bgCM/CM heart, causing DCM: Chrm2, Nebl, Tcap, and Zbtb17. Overall, these findings imply a role for A1BG in female hearts for acetyl-CoA and glucose-6-phosphate metabolism, with dysregulated genes pointing to a critical function in preventing DCM.
A1BG is required in females to form the cardiac intercalated disc
DCM is a condition characterized by the enlargement and weakened contraction of the left ventricle or both ventricles. This condition is often linked to changes in the connections between heart muscle cells [21–23]. Intercalated discs mediate CM cell-cell connections [48–52]. Mutations in genes related to the intercalated discs in the heart, such as desmoplakin, plakophilin-2, and plakoglobin, which are involved in the formation and function of desmosomes, can cause DCM [48, 50, 53, 54].
To test the hypothesis that loss of A1BG in female but not male hearts led to alteration of the cardiac intercalated discs, we used transmission electron microscopy to compare intercalated discs in A1bgCM/CM mice and control mice to determine whether A1BG expression affects intercalated disc structure. High-resolution (50000x) images uncovered that female A1bgCM/CM intercalated disc morphology was altered compared to control females (Fig. 3A-D).
To quantitatively evaluate the shape of intercalated discs, we measured the ratio of the total length of intercalation to the straight-line length of the cell boundary, as explained in [29] (Fig. S1). This allowed us to identify discs with higher values, indicating greater intercalation and, as a result, a larger surface area available for cell adhesion and ion transport. Our analysis showed that female mice had more intercalation than male control mice (Fig. 3E). Female A1bgCM/CM had significantly less intercalation than female controls. In fact, the level of intercalation in female A1bgCM/CM was initially similar to that of male mice. Conversely, male A1bgCM/CM did not differ in intercalated disc shape relative to controls. These findings suggest that there are inherent sex differences in cardiac intercalated disc structure and highlight a sex-differential requirement for the role of A1BG in forming intercalated discs.
A1BG leads to alterations in female cardiac electrophysiology
DCM is associated with alterations in the heart's electrical properties and conduction pathways [21–23, 55, 56]. Intercalated discs are crucial in coordinating the heart's contractions by facilitating mechanical and electrical connections between CMs [48–51]. Disruptions in the structure or function of these intercalated discs can significantly affect the heart's electrical properties, leading to impaired cardiac function [51, 55, 57–60].
Given the intercalation phenotype in female A1BGCM/CM mice, we investigated the electrophysiological consequences of A1BG in both sexes. A1BG is more highly expressed in cardiac atria [33]; therefore, we expected to observe alterations in the electrocardiogram (EKG) PR interval. The PR interval is the time from atrial to ventricular depolarization, indicating the time for electrical impulses to be transmitted through the atria to the AV node (Fig. 3F).
Sex differences exist in human atrial conduction; females have a shorter PR interval than males [61]. This difference was conserved in adult mice, as shown in this study and others (Figure [62, 63]). A1BGCM/CM female mice had significantly longer PR intervals than female control mice, indicating a longer time needed for atrial depolarization (Fig. 3F). The female A1BGCM/CM mouse PR interval was similar to the male baseline PR interval (Fig. 4B). As expected, the PR interval was inversely correlated with CM intercalation, with greater intercalation corresponding to shorter PR, affirming the sex-differential role for A1BG in the heart.
A1BG in females and males interacts with a distinct set of cardiac proteins
There have been limited studies on the function of A1BG. It has been found that the loss of A1BG causes defects in cardiac function that resemble DCM in females but not in males. This observation does not clarify the function(s) of A1BG or explain why there is a different requirement for it based on sex. Additionally, it has been reported that A1BG is one of the most differentially expressed cardiac proteins between males and females in mice at E9.5 and in adults, with higher expression in females than in males [64].
To better understand why females specifically require A1BG, we conducted a predicted structural analysis of mouse and human A1BG using Alphafold [65]. Our analyses suggest that the first two exons of the mouse and human A1BG transcript are predicted to encode a signal peptide, and the following five exons encode repeating IgG-like domains (Fig. 4A-C). The structural prediction of mouse and human A1BG suggests that the protein is secreted or associated with the outer cell membrane. We tested this hypothesis by co-immunostaining adult heart tissue in mice with an A1BG and CM (tropomyosin) antibody (Fig. 4D). Our results demonstrate that A1BG is associated with the outer surface of atrial CMs.
The observation that A1BG is a CM cell surface protein, which is required for the proper formation of intercalated discs and cardiac conduction in females but not males, as well as the structural prediction of A1BG, has led us to characterize cardiac A1BG interactomes in females and males. Researchers have not found a highly specific, high-affinity antibody against mouse A1BG that can function in immune-affinity purification. To address these issues, we generated an inducible A1BG allele by knocking an epitope-tagged version of A1BG (A1BG-3xHA) flanked by flox-stop-flox cassettes into the ROSA26 locus to create A1BG3XHA. To induce expression of A1BG-3xHA in CMs we crossed the A1BG3XHA to cTnt-Cre mice, CM-A1BG3XHA. F1 and F2 heterozygous and homozygous CM- A1BG3XHA mice were viable and fertile and had no observable phenotypic abnormalities and expression in the F2 was confirmed by immunoblot with anti-HA antibodies (Fig. 5A, Fig. S1).
To deduce the function of A1BG in cardiac tissue and to further explore the sex difference requirements for A1BG, we defined the A1BG endogenous cardiac interactome by performing mass spectrometry (MS) analysis of immuno-affinity purified (IP-MS/MS) female A1BG3HA CM complexes [40, 66, 67] (Fig. 4B). The complexes (N = 3) were obtained under physiological conditions from CMs derived from the hearts of female and male CM-A1BG3XHA mice at 4 weeks of age (Fig. 4B) in the presence of RNAse and DNAse. Results demonstrate that we could recover A1BG3XHA at 73%, the theoretical maximum with a trypsin digest (Fig. S2).
The analysis of CM-A1BG3XHA complexes utilized an unbiased gene ontology-based bioinformatics classification to scrutinize the functions of proteins linked with A1BG. Functional network analyses clearly showed that A1BG interacts with a group of 15 proteins enriched in females and 19 enriched in males (Fig. 5C, D). Upon conducting gene ontology analysis, it was apparent that the female interactome is enriched with proteins involved in generating precursor metabolites and energy, while the male interactome is enriched in in extracellular matrix (ECM)-receptor interaction and cell adhesion proteins (Fig. 5C, D, Table 1). Among the 15 proteins found to be enriched in females, 7 have not undergone a study in the context of the heart, while the remaining 8 have been linked to cardiac disease, including DCM (Table 1). None of the female A1BG interacting proteins were identified in the male A1BG cardiac interactome (Fig. 5C, D, Table 1, 2). Instead, the male A1BG cardiac interactome comprises proteins involved in protein degradation. These proteins were absent in the female cardiac interactome (Fig. 5C, D, Table 1, 2). Thus, the specific set of interacting proteins differed significantly from that in females. Female interactomes are enriched in proteins related to energy metabolism and are associated with DCM pathologies. Our findings suggest a sex-specific requirement for A1BG in cardiac health and imply that A1BG interactions may underlie the sex-specific requirements for A1BG in cardiac function.
Table 1
Proteins enriched in Female A1BG IP
Protein name | Description | Connection to cardiac physiology |
Slc25a12 | Calcium-binding mitochondrial carrier protein Aralar1 | NA |
Atp5f1a | ATP synthase subunit alpha, mitochondrial | NA |
Myo1c | Unconventional myosin-Ic | NA |
Eef2 | Elongation factor 2 | Pathological hypertrophy (Varma et al., 2023[71]) |
Idh2 | Isocitrate dehydrogenase [NADP], mitochondrial | Cardiac hypertrophy (Wu et al., 2022[72], Ku et al., 2015[73]) |
Fbxo6 | F-box only protein 6 | NA |
Taf2 | Transcription initiation factor TFIID subunit 2 | NA |
Ppp1r3a | Protein phosphatase 1 regulatory subunit 3A | Atrial fibrillation (Alzina et al., 2019[74]), Heart failure Cordero et al., 2019[75]) |
Gsn | Gelsolin | Myocardial infarction (Li et al., 2009[76]), Atrial fibrillation (Schrickel et al., 2009[77]) |
Atad3 | ATPase family AAA domain-containing protein 3 | Perinatal cardiomyopathy (Frazier et al., 2021[78]) |
Hrg | Histidine-rich glycoprotein | NA |
Hsp90b1 | Endoplasmin | Kawasaki disease (Mingguo et al., 2020[79]) |
Hnrnpf | Heterogeneous nuclear ribonucleoprotein F | NA |
Mdh1 | Malate dehydrogenase, cytoplasmic | Acute myocardial infarction (Pan et al., 2020[80]) |
Vdac2 | Voltage-dependent anion-selective channel protein 2 | Dilated cardiomyopathy (Shankar et al., 2021[81]) |
Table 2
Proteins enriched in male A1BG IP
Protein name | Description | Connection to cardiac physiology |
Gja1 | Gap junction alpha-1 protein | Arrhythmogenic cardiomyopathy (Palatinus 2023[82]) |
Nid1 | Nidogen-1 | NA |
Col6a1 | Collagen alpha-1(VI) chain | Trisomy 21 congenital heart disease (Davies et al., 1995[83]) |
Flnc | Cluster of Filamin-C | Hypertrophic & Dilated cardiomyopathy (Verdonscot et al., 2020[84]) |
Tln2 | Talin-2 | Atrial septal defect (Teekakirikul et al., 2022[85]) |
Thbs1 | Cluster of Thrombospondin-1 | NA |
Myl7 | Myosin regulatory light chain 2, atrial isoform | NA |
Macroh2a1 | Core histone macro-H2A.1 | NA |
Sorbs1 | Sorbin and SH3 domain-containing protein 1 | NA |
Dcn | Decorin | NA |
Samm50 | Sorting and assembly machinery component 50 homolog | Promotes hypertrophy (Xu et al., 2021[86]) |
Trim7 | E3 ubiquitin-protein ligase TRIM7 | NA |
Agrn | Agrin | Catecholaminergic polymorphic ventricular tachycardiac (Jaouadi et al., 2022[87]) |
Mb | Myoglobin | Myoglobinopathy (Olive et al., 2019[88]) |
Serpinh1 | Serpin H1 | NA |
Emilin1 | EMILIN-1 | Aortic valve disease (Munjal et al., 2014[89]) |
Spta1 | Spectrin alpha chain, erythrocytic 1 | NA |
Tgfbi | Transforming growth factor-beta-induced protein ig-h3 | Atrial fibrillation (Guan et al., 2022[90]) |
Obscn | Obscurin | Hypertrophic cardiomyopathy (Wu et al., 2021[91]), Arrhythmogenic right ventricular cardiomyopathy (Chen et al., 2020[92]) |
rps27a | ribosomal protein 27a | NA |