Strategy 1 Tnnt2-Dre; R26-iCre; IR1 non-muscle cells do not contribute to myocytes in adult hearts

Tnnt2-Dre; R26-iCre; IR1 mice were used to simultaneously track cardiomyocytes and non-cardiomyocytes in adult hearts. Immunostaining of tdTomato, ZsGreen and TNNI3 showed that tdTomato+ cells were TNNI3+ cardiomyocytes, whereas ZsGreen+ cells were not TNNI3+ cardiomyocytes (Fig. 1G). Isolation of Tnnt2-Dre; R26-iCre; IR1 cardiomyocytes did not detect ZsGreen+ cardiomyocytes in vitro (Fig. 1H). Thus, the Tnnt2-Dre; R26-iCre; IR1 system labeled myocytes and non-muscle cells with two different fluorescent colors (Fig. 1I).

Figure 1. GI

Does non-muscle cells differentiate into myocytes in adult hearts? If so, the new cardiomyocytes derived from non-muscle cells will be ZsGreen+. The results of the experiment were: after Dox induction, Tnnt2-Dre; R26-iCre; IR1 mice were ligated to the left anterior descending coronary artery to induce MI modeling (Fig. 2A). After myocardial infarction, a large number of myocardial cells died and cardiac function was impaired. ZsGreen, tdTomato and TNNI3 immunostaining were performed on heart sections 2-4 weeks after injury, and no ZsGreen+TNNI3+ cardiomyocytes were found (Fig. 2B). ZsGreen+tdTomato-cardiomyocytes were not detected in more than 600 sections of 5 myocardial infarction tissues. Flow cytometric analysis (Fig. 2D) and fluorescence microscopy (Fig. 2E) also confirmed that ZsGreen+ cardiomyocytes were not observed in the hearts of Tnnt2-Dre; R26-iCre; IR1 mice after MI modeling. The RT-PCR results also showed only tdTomato expression and no ZsGreen mRNA expression (n=5, Figure 2F). These data indicate that non-cardiomyocytes do not contribute to muscle cells after cardiac damage in adult hearts.

Figure 2. AF

In parallel, the cell fate of non-muscle cells to muscle cells in skeletal muscle after skeletal muscle injury was converted as a positive control. During the steady state maintenance of skeletal muscle, myofibroblasts are fused with non-muscle cells, and myofibroblasts are in a multinuclear state. Therefore, in normal skeletal muscle (sham operation group), tdTomato+ZsGreen+ muscle cells can be detected (Fig. 2G). When BaCl2 injury was applied to the tibialis anterior muscle (TA muscle) of Tnnt2-Dre;R26-iCre;IR1 mice, muscle fiber cells in skeletal muscle died in large numbers, and weak tdTomato fluorescence signal and strong ZsGreen expression were detected after TA muscle injury. It was shown that ZsGreen+ cells gradually replaced tdTomato+ cells after BaCl2 modeling (Fig. 2G). ZsGreen+tdTomato-muscle cells (arrow, Figure 2G) were detected in damaged TA muscle compared to the sham-operated group, indicating that non-muscle cells re-formed muscle cells.

Figure 2. G

Combining the above results, the dual homologous recombination lineage tracing technique of Tnnt2-Dre;R26-iCre;IR1 (strategy 1) revealed the transformation of non-muscle cells into myocytes in embryonic heart and adult skeletal muscle, but in adult heart Neutral myocytes do not contribute to myocytes.

Strategy 2 Tnnt2-Cre; R26-DreER; IR3 Reverse Cre/Dre System Labels Two Cells in Non-Myocyte to Myocyte Transformation Studies

A second double homologous recombination lineage tracing system was designed to validate the transformation of non-myocytes into myocytes. The system consists of Tnnt2-Cre (Cre expressed in myocytes), tamoxifen-inducible R26-DreER and IR3 reporter mice. The non-muscle cells were labeled with tdTomato red fluorescence exclusively by IR3 reporter mice (see figure below), and the muscle cells were specifically labeled with ZsGreen green fluorescence.

Tnnt2-Cre; R26-DreER; IR3 mice can simultaneously label muscle cells (ZsGreen+) and non-muscle cells (tdTomato+) (Fig. 3A).

Figure 3. A

ZsGreen was primarily labeled with the original cardiac tube in the embryo's E8.0 day heart (Fig. 3B). Tamoxifen was induced on day E8.0, and tdTomato+ non-muscle cells were found to have a significant contribution to muscle cells in the right ventricle at E13.5 days (Fig. 3C). This result was also confirmed by immunostaining of heart sections (Fig. 3D-F).

Similarly, in adult hearts and skeletal muscle, the transformation of fate of non-cardiomyocytes was observed by MI and BaCl2 modeling (Fig. 3G). Flow cytometric analysis (Fig. 3H) and fluorescence microscopy (Fig. 3I) of isolated cardiomyocytes also showed no tdTomato+ cardiomyocytes, respectively. Morphologically, these tdTomato+ cells were also significantly different from ZsGreen+ cardiomyocytes (Fig. 3J).

Immunostaining of tdTomato and TNNI3 showed that these tdTomato+ cells were not transformed into TNNI3+ cardiomyocytes in the infarct, border and distal regions of the MI heart (Fig. 3K). TdTomato+TNNI3+ cardiomyocytes were not detected in more than 600 tissue sections from 5 Tnnt2-Cre; R26-DreER; IR3 mouse hearts. RT-PCR results also indicated that isolated cardiomyocytes expressed only ZsGreen and did not express tdTomato (Fig. 3L).

In contrast, in positive control experiments, tdTomato+ZsGreen-muscle cells (arrows, Figure 3M) were easily detected after tamoxifen induction in the TA muscle injury model, whereas the sham group did not.

image 3.

Combining the above results, the second strategy uses Tnnt2-Cre; R26-DreER; IR3 also shows results consistent with strategy 1: non-cardiomyocytes are converted to myocytes in embryonic heart and adult skeletal muscle, but in adult hearts Myocytes did not contribute.

Strategy 3 Tnni3-Dre; R26-iCre; IR1

To further verify the above findings, another muscle cell marker, TNNI3, was used to drive the Dre recombinase. Tnni3-Dre effectively and specifically labels muscle cells in adult hearts, but does not label non-muscle cells. Muscle cells and non-muscle cells were labeled with Tnni3-Dre; R26-iCre; IR1 mice, respectively (Fig. 4A).

Figure 4. A

Immunostaining by ZsGreen, tdTomato, and different cardiac cell line markers showed that myocytes express tdTomato, non-muscle cells (eg, endothelial cells, fibroblasts, etc.) express ZsGreen (supplemental material). No ZsGreen+ cell tissue was observed in Tnni3-Dre; R26-iCre; IR1 mice without Dox induction (Fig. 4B).

Two weeks after Dox administration, MI or BaCl2 injection (Fig. 4C) was performed, and it was found that in the sham-operated heart, tdTomato+ cardiac cells were also labeled with ZsGreen green fluorescence after Dox induction (Fig. 4D). Immunostaining of tdTomato, ZsGreen and TNNI3 showed that all cardiomyocytes were tdTomato+ZsGreen- (Fig. 4E). After MI modeling, the number of tdTomato+ myocytes in the infarcted heart tissue was significantly reduced and replaced by ZsGreen+ non-muscle cells (Fig. 4F).

The myocardial cells after separation of MI were observed in vitro as tdTomato+ZsGreen- cells (Fig. 4G). Immunostaining of tdTomato, ZsGreen and TNNI3 showed that ZsGreen+ cells did not turn to cardiomyocyte fate after injury (Fig. 4H, I). No ZsGreen+tdTomato-muscle cells were detected in the MI heart by split flow cytometry and RT-PCR (Fig. 4J, K).

In contrast, in positive control experiments, ZsGreen+tdTomato-muscle cells (arrow, Figure 4L) were detected in injured TA muscle, indicating that non-myocyte neoplasms are myocytes.

Figure 4.

Therefore, the third model based on Tnni3-Dre; R26-iCre; IR1 also revealed that non-muscle cells do not transform into myocytes in adult hearts.

Strategy 4 Tnnt2-Dre; Actb-Cre; NR1 studies the transformation of non-muscle cells into myocytes through the NR1 system

Although inducible Cre or Dre driven by a broad promoter can effectively label most non-muscle cells, in practice the labeling efficiency does not reach 100%. It is still possible that a small number of unlabeled non-muscle cells produce new muscle cells in the adult heart after injury, although the likelihood is not significant because the labeling process during lineage tracing is completely random. To achieve 100% non-muscle cell labeling efficiency, a fourth strategy was designed using NR reporter mice (see figure below).

In this system, ZsGreen labels all cells in mice except for newly formed muscle cells labeled with tdTomato. All cells in NR1 mice after Actb-Cre-loxP recombination were first labeled as ZsGreen green fluorescence, and only new myocytes were labeled with tdTomato red fluorescence upon formation due to recombination of the second Tnnt2-Dre-rox (Fig. 5A). Green fluorescent protein has a half-life of 12-24 hours in mammalian cells, so although Dre-rox recombination in myocytes leads to tdTomato expression, ZsGreen protein still has a stable labeling time before degradation. Therefore, double positive cells can be detected during the transformation of non-muscle cells into muscle cells.

Figure 5. A

The system was first tested by analyzing the embryonic heart of the previous experiment (Fig. 5B). It is easy to detect tdTomato+ZsGreen+ myocytes in the heart outflow tract developed in E8.5 days (arrow, Fig. 5C), which is consistent with previous studies, indicating that the second heart region progenitor cells differentiate into cardiomyocytes at this stage.

Mouse heart samples were then collected every 12 hours during the 28 days of MI modeling for analysis (Fig. 5B). For ZsGreen, tdTomato and TNNI3 immunostaining, the results showed no ZsGreen+tdTomato+ muscle cells in the infarcted myocardium (Fig. 5D). At all 56 time points, Tnnt2-Dre; Actb-Cre; NR1 mouse heart samples, no ZsGreen+tdTomato+ muscle cells were detected in more than 6000 heart sections. Similarly, isolation of cardiomyocytes for flow cytometry did not reveal muscle cells in ZsGreen+ cells (Fig. 5E).

In contrast, ZsGreen+tdTomato+ myocytes were easily detected in the injured TA muscle in the positive control experiment (Fig. 5F). Non-muscle cells were 100% labeled with ZsGreen (supplemental material) in Tnnt2-Dre; Actb-Cre; NR1 mouse heart sections.

Figure 5.

In summary, the results of the fourth lineage tracing strategy based on Tnnt2-Dre; Actb-Cre; NR1 showed that the transformation of non-muscle cells into myocytes appeared only in the embryonic phase and not in the adult heart.