Transcription factor neuromancer/TBX20 is required for cardiac function in Drosophila with implications for human heart disease

Abstract

neuromancer/Tbx20 (nmr) genes are cardiac T-box transcription factors that are evolutionarily conserved from flies to humans. Along with other known congenital heart disease genes, including tinman/Nkx2-5, dorsocross/Tbx5/6, and pannier/Gata4/6, they are important for specification and morphogenesis of the embryonic heart. The Drosophila heart has proven to be an excellent model to study genes involved in establishing and maintaining the structural integrity of the adult heart, as well as genes involved in maintaining physiological function. Using this model, we have identified nmr as a gene required in adult fly hearts for the maintenance of both normal myofibrillar architecture and cardiac physiology. Moreover, we have discovered synergistic interactions between nmr and other cardiac transcription factors, including tinman/Nkx2-5, in regulating cardiac performance, rhythmicity, and cardiomyocyte structure, reminiscent of similar interactions in mice. This suggests a remarkably conserved role for this network of cardiac transcription factors in the genetic control of the adult heart. In addition, nmr-tinman interactions also influence the expression of potential downstream effectors, such as ion channels. Interestingly, genetic screening of patients with dilated cardiomyopathy and congenital heart disease has revealed TBX20 variants in three sporadic and two familial cases that were not found in controls. These findings suggest that the fly heart might serve as an identifier of candidate genes involved in human heart disease.

Fig. 1.

Requirement for nmr/Tbx20 in adult heart function. (A, and B) Double labeling of nmr-lacZ (nmr1, H15) with myofiber marker α-Actinin shows nuclear nmr expression in four Tinman-positive cardiomyocytes (arrows in B) and two Seven-up-positive ostia cells (asterisks in B) in each hemisegment. (C) Double labeling for nmr-lacZ and muscle-specific transcription factor Dmef2 indicates co-localization of nmr and Dmef2 in all myocardial cells including ostiae cells (arrows). (D) Double labeling of nmr-lacZ with Tinman cardiomyocyte nuclei. Notice that Tinman is absent in the ostia (arrows). (E) Loss of nmr1 or cardiac knock-down of nmr2 results in a dramatic increase in heart arrest/fibrillation rate (percentage of flies whose hearts failed when paced, see Materials and Methods). χ² analysis compared to wild type (w) and controls (nmr1⁶¹⁴/+ and Df(2L)Exel612/+): *, P < 0.01, **, P < 0.001, n = 50–100. (F) TARGET-mediated knock-down of nmr2 in the adult myocardium after eclosion results in increased arrest/fibrillation rate upon pacing stress (29 °C) compared to uninduced controls (18 °C). See Table S1 for details of statistical analysis. (G-K) α-Actinin labeling of the adult heart (one segment). (G and H) Confocal sections through the outer longitudinal heart-associated muscle layer indicate no difference between nmr2-kockdown hearts and controls. (I-K) Sections through the inner cardiomyocytes show abnormal arrangement of the spiral myofibers (arrows) in nmr2 adult specific mutant hearts using either the 24B-Gal4 (J) or tinCΔ4-Gal4 driver (K).

Fig. 2.

Arrhythmias in nmr mutant adult hearts. (A-C) M-mode traces prepared from high-speed movies of semiintact flies (for methods see Ocorr, et al. in ref. 26). (A and B) M-modes from control flies show regular beating pattern. (C) Arrhythmic heart beats are evident in cardiac nmr2 knockdown flies. Representative traces are shown. (A'-C') Histograms of heart periods (method as described in ref. 26). No differences were observed between males and females of each genotype that were pooled (n as indicated above each M-mode). (D-F) Statistical analysis of heart contraction in nmr2 adult specific mutant hearts. nmr2-RNAi expression in the heart results in slower heart beat due to longer Diastolic Intervals (D) and in a dramatic increase in the incidence of arrhythmias (F), without affecting the Systolic Interval (E).

Fig. 3.

Abnormal cardiac function and morphology in nmr,tin transheterozygotes. (A) Heart arrest/fibrillation rates of transheterozygous combinations of neuromancer/Tbx20 (Df(2L)Exel6012) with tinman (tin³⁴⁶) or dorsocross/Tbx5/6 (Df(3L)DocA), and corresponding controls. Sample sizes are 50–100 flies/data point. χ² analysis [with respect to all controls; see Wessells, et al. (2004) in ref. for details]: *, P < 0.01 (see Table S1 for individual values). (B) Increased heart arrest/fibrillation rate in transheterozygotes of nmr and tinman or doc is rescued by expressing nmr2 cDNA in the heart. (χ² analysis: *, P < 0.01; **, P < 0.001, see Table S1 for individual values). (C-E) Representative M-mode traces from each genotypes. Note the increased incidence of arrhythmia in nmr+/−;tin+/− flies (arrows in H indicate shortened heart beats, arrowheads in H show heart beats with longer diastolic interval). (F and G) α-Actinin labeling of adult heart. As compared to control (nmr+/−, F), nmr+/−;tin+/− (G) transheterozygous mutants show disruption in the regular myofibril alignment and Z-lines spacings. (H) Quantification of myofiber irregularities. Severity is determined by the average number of segments (5 in total as shown in Fig. 1A) with significant abnormal myofiber structure (as in G). Student t test: *, P < 0.05.

Fig. 4.

Quantitative real-time PCR on candidate genes in neuromancer, tinman mutants. Relative expression of nmr1, nmr2, Calcium ATPase at 60A (Ca-P60A; a SERCA homolog), dystrophin (dys), and eag-like K⁺ channel (elk) in 1-week-old adult hearts were standardized by rp49 expression. Wild-type control (w¹¹¹⁸) was set as one (indicated by thicker horizontal line). (A-D) The transheterozygotes of nmr1&2 deficiency [Df(2L)Exel6012] and tinman null mutants (tin³⁴⁶/+ or tin^EC40/+), Df(nmr)/+;tin/+, showed lower expression than heterozygous Df(nmr)/+ or tin/+ alone, indicating that nmr1 itself, nmr2, Ca-P60A, and dys were positively regulated by Nmr and Tinman. The expression level of nmr2 is also significantly reduced in the progeny of a cross between the heart-specific driver tinCΔ4-Gal4 and UAS-nmr2 RNAi line. nmr2 expression in the nmr2 knockdown flies is significantly reduced, compared to wild type (w¹¹¹⁸) as well as to Df(nmr)/+ or tin/+. (E) In contrast, the transheterozygotes showed higher level of elk expression indicating that elk may be negatively regulated by Nmr and Tinman. The normalized (to w¹¹¹⁸) average of independently prepared samples plus standard error is shown. We used 20 fly hearts for each sample (RNA extraction). Two samples were prepared for each genotype, thus tin/+ and Df(nmr)/+;tin/+ lanes are the normalized averages of four samples and Df(nmr)/+ and tinCΔ4>nmr2-RNAi the normalized averages of two samples each. For statistical analysis we used non-repeated measures of ANOVA, then the differences were calculated between all combinations of each experimental group by Student-Newman-Keules test (parametric) shown in the following triangles; **, P < 0.01; *, P < 0.05.

Fig. 5.

Tbx20 mutations in humans with cardiomyopathy. (A) Patient cohort details. Four types of Tbx20 mutations are identified from patients diagnosed with dilated cardiomyopathy (see SI Materials and Methods: Clinical data, for details). (B and C) Pedigrees of the families carrying the TBX20 mutation. (B) In family 3, the 7-year old proband with the W349R mutation died of DCM. The father also died of acute idiopathic DCM (no sample available for genetic studies). (C) In family 4, the 16 year proband and his affected sister were detected with R334Q mutation. One of his unaffected brothers, his father, and some other extended family members are carriers of the mutation. In this family the grandfather and one of the uncles died suddenly of DCM.