Conserved chamber-specific polyploidy maintains heart function in Drosophila

Abstract

Developmentally programmed polyploidy (whole-genome duplication) of cardiomyocytes is common across evolution. Functions of such polyploidy are essentially unknown. Here, in both Drosophila larvae and human organ donors, we reveal distinct polyploidy levels in cardiac organ chambers. In Drosophila, differential growth and cell cycle signal sensitivity leads the heart chamber to reach a higher ploidy/cell size relative to the aorta chamber. Cardiac ploidy-reduced animals exhibit reduced heart chamber size, stroke volume and cardiac output, and acceleration of circulating hemocytes. These Drosophila phenotypes mimic human cardiomyopathies. Our results identify productive and likely conserved roles for polyploidy in cardiac chambers and suggest that precise ploidy levels sculpt many developing tissues. These findings of productive cardiomyocyte polyploidy impact efforts to block developmental polyploidy to improve heart injury recovery.

Keywords: Drosophila; Endocycle; Heart; Human; Polyploidy.

Fig. 1.

Embryonic and larval cell cycle activity of Drosophila cardiomyocytes. (A,A′) Drosophila cardiac organ (dorsal vessel, red) in embryos (A) and larvae (A′). Aorta: segments A1-A4; heart: A5-A6; apex: A7. Each segment contains two ostial pairs (blue) and eight cardiomyocytes (red), except A7 (four cardiomyocytes). (B-B‴) Example embryonic cardiomyocytes expressing Gal4-UAS-Fly-FUCCI. GFP-E2F1_1-230+: G1-phase (B), RFP-CycB_1-266+: S-phase (B′), RFP-CycB_1-266+/GFP-E2F1_1-230+: G2-phase (B″), PH3+/RFP-CycB_1-266+/GFP-E2F1_1-230+: M-phase (B‴). Scale bars: 3 µm. (C) Representative twi-GAL4+ stage 8 S and G2 cardiomyocytes. Scale bar: 20 µm. (D) Representative HandC-Gal4+ stage 16-17 G1 and G2 cardiomyocytes. Scale bar: 20 µm. (E) Quantification of FUCCI data in stage 8-17 embryos. Drivers: stage 8, twi-GAL4; stage 9-11, twi-GAL4 and pnr-GAL4; stage 14-15, Mef2-GAL4; stage 16-17, Mef2-GAL4 and HandC-GAL4. Colors for each cell cycle phase match B-B‴. n=10 embryos/group. (F) Embryonic cardiomyocytes undergo two types of mitotic cycles during stages 8-15 and arrest at G1 or G2 at stage 16-17. (G,H) Cardiomyocyte nuclei (DAPI) in embryonic (G) and wandering larval third instar (WL3; H) stages. Scale bars: 10 µm. (I) Cardiomyocyte number from segments A1-A7 in first instar larval (L1) and WL3 stages. n=10/group. Cardiomyocyte nuclei: NP5169-Gal4>UAS-mCherry-NLS+. (J-J‴) Representative BrdU⁺ cardiomyocyte (heart chamber) at WL3 for NP5169-Gal4>UAS-mCherry-NLS. Cardiomyocyte nuclei: mCherry (red), anti-BrdU (green) and Hoechst (blue). Dotted outline delineates the nucleolus. Scale bars: 5 µm. (K) WL3 cardiac organ. Cardiomyocytes: NP5169-Gal4>UAS-mCherry-NLS (NP>; red); pericardial cells: HandC-GFP (green; outlined); nuclei: DAPI (blue). Scale bar: 100 μm. Each dataset includes at least two biological repeats.

Fig. 2.

Polyploidization of Drosophila larval cardiomyocytes is chamber specific. (A-C) Representative BrdU⁺ (red) cardiomyocytes (green) in the WL3 heart chamber. NP5169-Gal4>UAS-GFP-NLS (NP>) animals were pulse-fed BrdU (Materials and Methods) for 24 h at the indicated times after hatching. Scale bars: 50 µm. (D) Percentage of BrdU⁺ cardiomyocytes in the heart (red) and aorta (blue) during larval development. n=5 animals/group. (E) Ploidy distribution in the heart (A5-A6; red) and aorta (A1-A4; blue) for NP5169-Gal4>UAS-mCherry-NLS (NP>). ****P<0.0001, unpaired two-tailed Student's t-test. n=10 animals/group. (F-G′) GFP⁺ single-cell FLP-out clones (green; Materials and Methods) in the heart (F,F′) and aorta (G,G′). Dashed line marks one cardiomyocyte. Cardiac organ is stained with phalloidin (Phal; red). Scale bars: 20 µm. (H) Single-cell area of FLP-out GFP⁺ cardiomyocytes in each chamber (taken from 2D renderings of 3D image). The GFP clone in the heart wraps around the cardiac tube. ****P<0.0001, unpaired two-tailed Student's t-test, n=25 single-cell clones/group. (I) Larval cardiomyocyte endocycles lead to chamber-specific ploidy levels. (J,K) Representative 3D-rendered sections of heart (J) and aorta (K). Chamber wall, green (from phalloidin staining); cardiomyocyte nuclei, yellow (from DAPI and NP5169-Gal4>UAS-mCherry); non-cardiomyocyte nuclei, blue (from DAPI and location). (L,M) Representative transverse two-dimensional OCT images of the WL3 heart (L) and aorta (M) with End Diastolic Dimension area (EDD or a_D) pseudo-colored in gray for NP5169-Gal4> UAS-mCherry (NP>). Scale bars: 100 µm. Total wall thickness (t) is calculated as the sum of posterior (t_P) and anterior (t_A) wall thickness (yellow bracket). (N) OCT measurements of total wall thickness in the heart and aorta. ****P<0.0001, unpaired two-tailed Student's t-test, n=10/group. (O) OCT measurements of EDD (a_D) in the heart and aorta. ****P<0.0001, unpaired two-tailed Student's t-test, n=10/group. Each dataset includes at least two biological repeats. Data are mean±s.d. (represented as error bars in D and dashed and dotted lines in E,H,N,O).

Fig. 3.

Chamber-specific asymmetry in nuclear volume and insulin signaling in human hearts. (A) Method for analysis of human cardiomyocytes (see also Materials and Methods). Nuclear volume and ploidy (relative Hoechst intensity) were measured in 10 µm sections of the left ventricle (LV) and left atrium (LA) from the Duke Human Heart Repository (DHHR). (B-C″) Representative LV (B-B″) and LA (C-C″) cardiomyocytes, showing sarcomeres stained with phalloidin (Phal; green), membranes stained with Wheat Germ Agglutinin (WGA; red) and nuclei stained with Hoechst (blue). B″ and C″ show 3D-rendered LV and LA cardiomyocytes in yellow (Materials and Methods). Scale bars: 10 µm. (D) Nuclear ploidy (relative Hoechst intensity) of LV and LA cardiomyocytes. Dashed and dotted lines indicate mean±s.d.; ****P<0.0001, unpaired two-tailed Student's t-test, n=5 subjects. (E) Volcano plot showing differentially expressed genes in ventricular and atrial cardiomyocytes. KEGG analysis (Materials and Methods) show ventricular upregulation of insulin signaling (red). Data taken from Litviňuková et al. (2020). (F-G′) Representative LV (F,F′) and LA (G,G′) cardiomyocytes stained with human anti-INSR antibody (white), phalloidin (Phal; green) and Hoechst (blue). Scale bars: 50 µm. Inset in F shows a higher magnification of enrichment of insulin receptor at intercalated discs. Scale bar: 25 µm. (H,H′) Single z-section showing LV cardiomyocytes stained with INSR antibody (white), WGA (red) and Hoechst (blue). Arrows indicate intercalated regions. Scale bars: 10 µm. (H″) Relative INSR intensity at intercalated regions for LV and LA cardiomyocytes (Materials and Methods). Line shows mean and shaded area represents s.d. **P<0.01, Sidak's multiple comparisons test. Dashed line indicates the point at which the intensity is significantly different. Each dataset includes two biological repeats. (I) Schematic of chamber-specific ploidy asymmetry and insulin signaling in humans. A higher expression of insulin receptor in LV cardiomyocytes may lead to increased nuclear polyploidization compared with LA cardiomyocytes through upregulation of insulin signaling.

Fig. 4.

Cardiomyocytes in the heart chamber endocycle faster than in the aorta and are more sensitive to growth and cell cycle signaling. (A-B‴′) Control NP5169-Gal4, UAS-mCherry-NLS X w¹¹¹⁸ (NP>) and NP>InR-RNAi cardiac organs at WL3. Cardiomyocytes are visualized with mCherry-NLS (red), phalloidin (Phal; green) and DAPI (blue). Insets show enlarged single nuclei (dotted boxes in lower magnification image) of the DAPI channel for NP> (A′-A‴′) and NP>InR-RNAi animals (B′-B‴′). Scale bars: 20 µm (A,B); 5 µm (A′-A‴′,B′-B‴′). (C,D) Chamber-specific WL3 cardiomyocyte ploidy levels of animals of the indicated genotypes (Materials and Methods) for heart (C) and aorta (D). Dashed and dotted lines indicate mean and s.d. ****P<0.0001, unpaired two-tailed Student's t-test, n=at least 6/group. Each dataset includes two or more biological repeats. Numbers above brackets indicate fold change of the mean. (E) Schematic of larval cardiomyocyte endocycles regulated by insulin signaling. During early larval stages, Drosophila cardiomyocytes endocycle at different rates to induce chamber-specific asymmetry in nuclear polyploidization. Higher insulin signaling in the heart chamber versus aorta chamber is responsible for increased nuclear polyploidization of cardiomyocytes.

Fig. 5.

Size of the heart chamber is reduced in cardiac ploidy-reduced animals. (A-D) Representative sagittal two-dimensional OCT images of diastolic WL3 heart chamber (A7-A5, pseudo-colored) of control NP5169-Gal4, UAS-mCherry-NLS X w¹¹¹⁸ (NP>) (A), NP>InR-RNAi (B), NP>InR^CA (C) and NP>fzr-RNAi (D). Scale bars: 100 µm. (E) Heart chamber length in animals of the indicated genotypes. (F-I) Transverse two-dimensional OCT images of WL3 heart chamber with End Diastolic Dimension area (EDD or a_D) pseudo-colored in gray for control NP5169-Gal4, UAS-mCherry-NLS X w¹¹¹⁸ (NP>) (F), NP>InR-RNAi (G), NP>InR^CA (H) and NP>fzr-RNAi (I). Scale bars: 100 µm. (J) a_D in the WL3 heart chamber in animals of the indicated genotypes. (K-N) Transverse two-dimensional OCT images of a_D, as in F-I, for the aorta of control (NP>) (K), NP>InR-RNAi (L), NP>InR^CA (M) and NP>fzr-RNAi (N). Scale bars: 100 μm. (O) a_D of aorta for animals of the indicated genotypes. (P) Stroke volume in the WL3 heart chamber of animals of the indicated genotypes. Control genotype: NP5169-Gal4, UAS-mCherry-NLS X w¹¹¹⁸ (NP>). Each dataset includes at least two biological repeats and ten animals/group. All statistical tests were unpaired two-tailed Student's t-test; data are presented as mean±s.d. (dashed and dotted lines). *P<0.05, **P<0.01, ***P<0.001. ns, not significant (P>0.05).

Fig. 6.

Cardiac output and aorta hemocyte velocity are impacted in cardiac ploidy-reduced animals. (A-D) OCT orthogonal images (Materials and Methods) of heart chamber for control NP5169-Gal4, UAS-mCherry-NLS X w¹¹¹⁸ (NP>) (A), NP>InR-RNAi (B) and NP>fzr-RNAi (C). (D) Heart rate (beats/min) in the heart of WL3 animals of the indicated genotypes. n=10/group. (E,F) Cardiac output in the WL3 heart (E) and aorta (F) of animals of the indicated genotypes. n=10/group. (G) Control (NP5169-Gal4, UAS-GFP-NLS X w¹¹¹⁸) WL3 heart (Phal; green) hemocytes (Hml-DsRed) and nuclei (DAPI; blue). Arrows point towards the aorta. Scale bar: 100 µm. (H-J) Montages from movies of WL3 aorta (Materials and Methods) of the indicated genotypes. Hemocytes visualized by Hml-DsRed. Time from first frame shown in ms. Scale bar: 100 µm. (K) Hemocyte velocity in WL3 heart and aorta. n=at least 5/group. (L,L′) Schematics showing chamber dimensions and velocity in control (NP>), NP>InR and >fzr-RNAi animals. In control animals, chamber-asymmetry in nuclear polyploidization leads to increased hemocyte velocity while moving from heart to aorta chamber (L). However, this acceleration is dampened in ploidy reduced animals (NP>InR, >fzr-RNAi) (L'). Each dataset includes at least two biological repeats. Data are presented as mean±s.d. (dashed and dotted lines). ns, not significant (P>0.05). *P<0.05, **P<0.01, ***P<0.001, unpaired two-tailed Student's t-test. Anterior to the right in all images.