To distinguish cell division from cell cycle variations in cardiomyocytes, we present protocols using two transgenic mouse lines: Myh6-H2B-mCh transgenic mice, for the unequivocal identification of cardiomyocyte nuclei, and CAG-eGFP-anillin mice, for distinguishing cell division from cell cycle variations.
Cardiomyocytes are prone to variations of the cell cycle, such as endoreduplication (continuing rounds of DNA synthesis without karyokinesis and cytokinesis) and acytokinetic mitosis (karyokinesis but no cytokinesis). Such atypical cell cycle variations result in polyploid and multinucleated cells rather than in cell division. Therefore, to determine cardiac turnover and regeneration, it is of crucial importance to correctly identify cardiomyocyte nuclei, the number of nuclei per cell, and their cell cycle status. This is especially true for the use of nuclear markers for identifying cell cycle activity, such as thymidine analogues Ki-67, PCNA, or pHH3. Here, we present methods for recognizing cardiomyocytes and their nuclearity and for determining their cell cycle activity. We use two published transgenic systems: the Myh6-H2B-mCh transgenic mouse line, for the unequivocal identification of cardiomyocyte nuclei, and the CAG-eGFP-anillin mouse line, for distinguishing cell division from cell cycle variations. Combined together, these two systems ease the study of cardiac regeneration and plasticity.
The correct identification of cardiomyocyte nuclei and the cell cycle status is of crucial importance for the determination of cardiac muscle turnover and regeneration. This is especially true for the use of nuclear markers, such as pHH3, Ki-67, or Thymidine analogs, for identifying cell cycle activity. As the proliferative capacity of adult mammalian cardiomyocytes is very small1, a false identification of a nucleus positive for a proliferation marker of a cardiomyocyte nucleus could make a crucial difference in the outcome of a proliferation assay. Moreover, cardiomyocytes are prone to variations in the cell cycle, such as endoreduplication and acytokinetic mitosis, which result in polyploid and multinucleated cells rather than in cell division. To this end, the interpretation of antibody staining against common cell cycle markers is not conclusive in all cases.
Here, we present methods for the straight-forward recognition of mouse cardiomyocytes and their nuclearity in native isolated cells and thick tissue sections at postnatal and adult stages by the unequivocal identification of their nuclei. For that purpose, a transgenic mouse line with cardiomyocyte-specific expression of a fusion protein consisting of human histone H2B and mCherry under the control of the Myh6 promoter (Myh6-H2B-mCh) was used2. Cross-breeding this mouse line with a transgenic proliferation indicator mouse line, in which the expression of an eGFP-anillin fusion protein is under the control of the ubiquitous chicken actin promoter with a CMV enhancer (CAG-eGFP-anillin), allows for the determination of the cell cycle status. The scaffolding protein anillin is specifically expressed in cell-cycle active cells3, and its differential subcellular localization during the cell cycle allows for live-tracking cell-cycle progression with a high resolution of M-Phase4. Therefore, the double transgenic mice can be used to discriminate between proliferating cardiomyocytes and those that undergo cell-cycle variations. This proves especially useful in screening for proliferation-inducing substances in vitro.
All procedures of this protocol involving animals were in accordance with the ethical standards of the University of Bonn and complied with the guidelines from Directive 2010/63/EU of the European Parliament on the protection of animals used for scientific purposes.
1. In Vitro Visualization of Cell Cycle Activity in Postnatal Cardiomyocytes
2. Determination of Nucleation in Adult Cardiomyocytes by Langendorff Dissociation and Thick Tissue Sections
In order to analyze the effects of siRNAs/miRNAs on the cell cycle activity of postnatal cardiomyocytes in vitro, cardiomyocytes of double-transgenic Myh6-H2B-mCh/CAG-eGFP-anillin mice were isolated on postnatal day 3 (P3) and transfected with cell cycle activity-inducing miR1995, siRNA p27, and siRNA Fzr1. Compared to the negative control (Figure 1A), the pictures of miR199- (Figure 1B) and siRNA p27- (Figure 1C) transfected cardiomyocytes show an induction of cell cycle activity. In the eGFP-anillin mouse model, siRNAs against Fzr1 can be used as transfection controls, as the inhibition of Fzr1 leads to the accumulation of eGFP-anillin fusion protein in the nuclei of transfected cells and a loss of Fzr1 leads to the inhibition of APCFzr1. Fzr1 is a cofactor of the anaphase-promoting complex E3 ligase, which targets anillin for proteasomal degradation. Figure 1D shows a confocal overview picture of siRNA Fzr1-transfected cardiomyocytes 3 days after transfection, indicating a transfection efficiency of ~ 45%. Cardiomyocytes that perform endoreduplication (e.g., after knockdown of p276) show exclusively nuclear eGFP-anillin expression (Figure 1E) or are eGFP-anillin negative (see discussion). They do not express eGFP-anillin in M-phase-typical localizations (Figure 1F and G), as endoreduplication only consists of an endo-S phase (eGFP-anillin-positive) and an endo-G phase (eGFP-anillin-negative). In the endo-G phase, the APC is active, resulting in the ubiquitination and degradation of eGFP-anillin in the proteasome.
Quantification of mono- and binucleated cardiomyocyte portions at the adult stage can be performed either at the single-cell level after the Langendorff dissociation of Myh6-H2B-mCh transgenic hearts or in thick cryoslices of transgenic hearts. After the enzymatic digestion of the heart tissue at the Langendorff apparatus, atria and ventricles can be mechanically separated and analyzed independently of each other. Figure 2A shows a representative picture of non-fixed H2B-mCh transgenic ventricular cardiomyocytes after Langendorff isolation with a high degree of binucleated cardiomyocytes, as indicated by the nuclear H2B-mCh expression. By contrast, the majority of atrial cardiomyocytes are mononucleated (Figure 2B). As the enzymatic digestion does not result in 100% single cells, the pattern of cross-striation revealed by α-actinin staining facilitates discrimination between binucleated cardiomyocytes (continuous pattern of cross-striation, Figure 2C) and cell doublets. Figure 2D shows an example of the identification of a binucleated cardiomyocyte in a thick slice.
3D reconstructions of thick slices of adult Myh6-H2B-mCh transgenic hearts can be used to determine the proportion of cardiomyocyte nuclei under physiological conditions within the tissue. Using the 3D module of imaging software, Hoechst-stained nuclei and H2B-mCh-positive nuclei can be detected and counted automatically, as illustrated in Figure 2E. The final result should be corrected manually for doublets, meaning nuclei that directly touch each other, in this case. To analyze the nucleation index of cardiomyocytes in thick slices, it is necessary to manually scroll through the stack, as the nuclei do not necessary lie within one z-plane. WGA staining allows for the detection of cell borders.
Figure 1: Examples of In Vitro Visualization of Cell Cycle Activity in Postnatal Cardiomyocytes after Transfection with siRNAs. (A-D) P3 cardiomyocytes from eGFP-anillin/Myh6-H2B-mCh hearts stained for α-actinin (white). Cardiomyocyte nuclei are identified by the H2B-mCh signal (red), cell cycle activity by eGFP-anillin signals (green), and nuclei by Hoechst nuclear dye (blue). (A) P3 cardiomyocytes transfected with scramble siRNA serve as the negative control. The bar is 100 µm. (B) P3 cardiomyocytes transfected with miRNA-199 display significantly more eGFP-anillin signals than the control (A). The bar is 100 µm. (C) Example of exclusively nuclear eGFP-anillin signals after transfection with p27 siRNAs, indicating endoreduplication. The bar is 80 µm. (D) Transfection with siRNA against Fzr1 for the determination of transfection efficiency. As eGFP-anillin accumulates, the number of eGFP-anillin+ cardiomyocytes indicates the transfection efficiency. The bar is 80 µm. (E-G) Different localizations of eGFP-anillin (green) in cardiomyocytes during the cell cycle: nuclear localization (arrow in E), contractile ring (arrows in F), and midbody localization (arrow in G). The bar is 20 µm in (E) and 10 µm in (F and G). Please click here to view a larger version of this figure.
Figure 2: Examples for the Assessment of Multinucleation by the Langendorff Isolation of Cardiomyocytes from H2B-mCh Mice and by the 3D Analysis of Thick Slices. (A,B) Ventricular (A) and atrial (B) cardiomyocytes from hearts from adult H2B-mCh transgenic mice after isolation by Langendorff dissociation. The bars are 50 µm. (C) Cardiomyocytes from Myh6-H2B-mCh hearts stained for α-actinin (green). Cardiomyocyte nuclei are identified by the H2B-mCh signal (red). The bar is 10 µm. (D) Binucleated cardiomyocyte in a thick slice (arrows). Cardiomyocyte nuclei are identified by the H2B-mCh signal (red), cell borders by WGA staining (green), and nuclei by ToPro3 (white). The bar is 50 µm. (E) Workflow for the 3D analysis of binucleation in thick slices. Please click here to view a larger version of this figure.
There is a controversy over whether cardiomyocytes are able to reenter the cell cycle and divide after injury and during tissue homeostasis. Values for the basic turnover of cardiomyocytes have been given in the range between 1%1 and 80%7. Also after a cardiac lesion, the induction of cell cycle activity and the generation of new cardiomyocytes has been reported in the border zone, with values between 0.0083%8 and 25 – 40%7. These discrepancies can partly be explained by different experimental approaches to identify cardiac nuclei, a process that is very challenging on histological sections9 and during cell division. As cardiomyocytes are prone to variations of the cell cycle, it is of crucial importance to distinguish authentic cell division from endoreduplication and acytokinetic mitosis, which lead to polyploid and multinucleated cells. For the identification of cell division, it is necessary to visualize hallmarks of cell division, such as contractile rings and midbodies, as well as to determine the percentage of binucleated cardiomyocytes. We have developed techniques to analyze cell cycle activity in cardiomyocytes in detail and to determine their degree of nuclearity in isolated cells or in thick sections.
It is important to note that the eGFP-anillin system provides direct proof of cell division, through the visualization of a symmetrical contractile ring (Figure 1F) and the midbody (Figure 1G), and indirect proof for endoreduplication and acytokinetic mitosis. As previously described, the occurrence of a unilateral ingression of the contractile ring is an indicator of binucleation10, and this can be observed with the eGFP-anillin system.
Endoreduplication is suggested as long as no contractile ring or midbody are observed, which makes calculations of the probabilities of detecting these localizations mandatory. Assuming a cell cycle duration of 25 h, an average duration of contractile ring visibility of 20 min, and a midbody persistence of 1 h, there should be 1 contractile ring in 100 dividing eGFP-anillin-positive cells and 4 midbodies. Statistically speaking, a minimum of 25 eGFP-anillin-positive cells would need to be analyzed to distinguish cell division from variations of the cell cycle. This number changes accordingly as the cell cycle duration (which is often unknown) increases or decreases. This also implies that the majority of eGFP-anillin signals will be nuclear (Figure 1E), as M-phase, the only phase with non-nuclear localizations, lasts only approximately 1 h.
During postnatal development, binucleation takes place in mouse hearts and increases to 90% in ventricular cardiomyocytes (~ 25% in humans)11. For the analysis of regeneration in the heart, it is important to determine the degree of binucleation. We describe two methods for addressing this important morphological feature: namely, Langendorff dissociation and the creation of thick sections of cardiac tissue. While Langendorff isolation is easier and faster, the morphology of thick sections is more time consuming but also more accurate. Interestingly, we have found that Langendorff isolation overestimates the percentage of binuclear cardiomyocytes. This could be due to different survival rates of mononuclear and binuclear cells during this rather rigid procedure.
As the induction of proliferation in postnatal cardiomyocytes is an emerging approach in cardiac regeneration, this protocol describes a screening system by combining two transgenic mouse lines, Myh6-H2BmCh and CAG-eGFP-anillin. Cardiomyocytes isolated from hearts on postnatal days P1 – P6 can be easily cultured and transfected with miRNAs or siRNAs or treated with libraries of small molecules. Cardiomyocyte nuclei can be manually or automatically detected by software algorithms, and potential "hits" can be determined by an increase in the mCh+ nuclei number. Discrimination of cell division from endoreduplication and acytokinetic mitosis can be done by quantification of the different localizations of the eGFP-anillin signals. This system should shed some new light onto the regulatory mechanisms underlying postnatal cardiomyocyte proliferation and may lead to the discovery of new therapeutic agents for the treatment of cardiac diseases.
The authors have nothing to disclose.
We thank S. Grünberg (Bonn, Germany) and P. Freitag (Bonn, Germany) for their technical assistance.
10 cm petri dish | Sarstedt | 821472 | |
100 µm cell strainer | Becton Dickinson GmbH/Falcon | 352360 | |
2,3-Butanedione monoxime (BDM) | Sigma-Aldrich | B0753 | |
G20x1 ½ injection cannula, Sterican | Braun, Melsungen | 4657519 | |
20 gauge needle | Becton Dickinson GmbH | 301300 | |
24-well plates | Becton Dickinson GmbH/Falcon | 353047 | |
2-Methyl-butane | Carl Roth GmbH + Co. KG | 3927.1 | |
37% formaldehyde solution | AppliChem GmbH | A0936,1000 | |
3-way stopcock | B. Braun Medical Inc. | 16494C | |
50 ml syringe | B. Braun Medical Inc. | 8728810F | |
70% ethanol | Otto Fischar GmbH | 27669 | |
Alexa-Fluor-conjugated secondary antibody | Jackson ImmunoResearch | 115-605-205 | |
Alpha-Aktinin EA-53, Mouse IgG | Sigma-Aldrich, Steinheim | A7811 | |
CaCl | Sigma-Aldrich | C4901 | |
Cell Culture Microplate, 96 Well, Half Area | Greiner bio-one | 675986 | |
Collagenase B | Roche | 11088815001 | |
confocal microscope Eclipse Ti-E | Nikon | ||
cryostat CM 3050S | Leica | ||
donkey serum | Jackson Immuno Research, Suffolk, GB | 017-000-121 | |
Dulbecco's Phosphate Buffered Saline | Sigma-Aldrich | D8537 | |
EDTA | Sigma-Aldrich | E4884 | |
fetal calf serum | PromoCell, Heidelberg | ||
Formaldehyde solution (4%) | PanReac AppliChem | A3697 | |
Gelatine from porcine skin, Type A | Sigma-Aldrich, Steinheim | G2500 | |
glass coverslips | VWR | 631-0146 | |
Glucose | Sigma-Aldrich | G7021 | |
Heidelberger extension tube | IMPROMEDIFORM GmbH | MF 1833 | |
Heparin-Natrium | Ratiopharm | 5394.02.00 | |
HEPES | Sigma-Aldrich | H3375 | |
HistoBond microscope slides | Marienfeld | 0810000 | |
Hoechst 33342 (1mg/ml) | Sigma Aldrich, Taufkirchen | B2261 | |
Insulin syringe | Becton Dickinson GmbH | 300334 | |
Iscove’s ModifiedDulbecco’s Medium (IMDM) | Gibco/Life Technologies, Darmstadt | 21980-032 | |
KCl | Sigma-Aldrich | P9333 | |
Laminin | Corning | 354221 | |
Laser Scanning Mikroskop Eclipse Ti | Nikoninstruments, Düsseldorf | ||
Lipofectamine RNAiMAX | Invitrogen/Life Technologies, Darmstadt | 13778075 | |
Mouse IgG Cy5 (donkey) | Jackson ImmunoResearch | 715-175-151 | |
MGCl | Sigma-Aldrich | M8266 | |
microcentrifuge tube | Sarstedt | 72690 | |
Mini shaker | VWR | 12620-940 | |
mirVana miRNA mimic, hsa-miR199a-3p | Ambion/Thermo Fischer Scientific | 4464066 | |
Biopsy Mold | Sakura Finetek/ VWR | 4565 | |
M-slide 8-well ibiTreat | ibidi | 80826 | |
NaCl | Sigma-Aldrich | S9888 | |
NaOH | Merck Millipore | 567530 | |
negative control(scrambled RNA) | Ambion/Thermo Fischer Scientific | AM4611 | |
Neonatal Heart Dissociation Kit | Miltenyi Biotech, Bergisch Gladbach | 130-098-373 | |
NIS Elements AR 4.12.01-4.30.02-64bit | Nikoninstruments, Düsseldorf | ||
Non essential amino acids, NEAA | Gibco/Life Technologies, Darmstad | 11140-035 | |
Opti-MEM, Reduced Serum Medium | Gibco | 51985-026 | |
P21-siRNA | Ambion/Thermo Fischer Scientific | 4390771 | |
P27-siRNA | Ambion/Thermo Fischer Scientific | 4390771 | |
Penicillin/Streptomycin | Gibco/Life Technologies, Darmstadt | 15140-122 | |
Phosphate buffered saline (PBS) | Sigma-Aldrich, Steinheim | 14190-094 | |
Polyvinyl alcohol mounting medium with DABCO®, antifading | Sigma-Aldrich | 10981 | |
RNase A | Qiagen | 1007885 | |
RNaseZap | Invitrogen/Life Technologies, Darmstadt | AM9780 | |
sample containers | Vitlab | 80731 | |
Serological pipette | Greiner | 607180 | |
software NIS Elements | Nikon | ||
Sucrose | Sigma-Aldrich | S0389 | |
Tissue-Tek O.C.T. Compound | Sakura Finetek/ VWR | 25608-930 | |
ToPro3 iodide (642/661) | Molecular probes/ThermoFisher Scientific | T3605 | |
Tris | Sigma-Aldrich | T1503 | |
Triton X | Fluka | 93418 | |
Triton X-100 | Fluka | 93418 | |
Trypsin | Sigma-Aldrich | T1426 | |
Wheat germ agglutinine (WGA) Fluorescein labeled | Vector Laboratories | VEC-FL-1021-5 | |
α-actinin antibody | Sigma-Aldrich | A7811 | |
β-Mercaptoethanol | Sigma-Aldrich, Steinheim | M3148 |