A minimally invasive approach to induce myocardial infarction in mice without thoracotomy.

Acute myocardial infarction (MI) is a leading cause of morbidity and mortality in the world. Traditional method to induce MI by left coronary artery (LCA) ligation is typically performed by an invasive approach that requires ventilation and thoracotomy, causing serious injuries in animals undergoing this surgery. We attempted to develop a minimally invasive method (MIM) to induce MI in mice. Under the guide of ultrasound, LCA ligation was performed in mice without ventilation and chest-opening. Compared to sham mice, MIM induced MI in mice as determined by triphenyltetrazolium chloride staining and Masson staining. Mice with MIM surgery revealed the reductions of LVEF, LVFS, E/A and ascending aorta (AAO) blood flow, and the elevations of S-T segment and serum cTn-I levels at 24 post-operative hours. The effects of MI induced by MIM were comparable to the effects of MI produced by traditional method in mice. Importantly, MIM increased the survival rates and caused less inflammation after the surgery of LCA ligation, compared to the surgery of traditional method. Further, MIM induced angiogenesis and apoptosis in ischaemic hearts from mice at postoperative 28 days as similarly as traditional method did. Finally, the MIM model was able to develop into the myocardial ischaemia/reperfusion model by using a balloon catheter with minor modifications. The MI model is able to be efficiently induced by a minimally invasive approach in mice without ventilation and chest-opening. This new model is potentially to be used in studying ischaemia-related heart diseases.

Traditionally, a ventilation-based thoracotomy to establish MI model by ligating left main descending coronary artery (LCA) in mice has been presented firstly by Johns and Olson in 1954, in which many surgical manipulations have been made to induce the cardiac ischaemic event. 12 Although LCA ligation remains the most commonly practiced ischaemic injury, this remains to utilize methodology requiring ventilation and full opening of the chest, resulting in extensive tissue damages and high surgery-related death. 13 To solve these problems, Gao et al 14,15 improved a traditional method to induce MI injury in mice without ventilation. However, this method also needs to open the thoracic cavity rapidly because LCA is not visible if the chest is closed. Thoracotomy may cause some traumas, which not only increases animal pains and mortalities, but also exerts some detrimental effects on the whole body immune homoeostasis, a situation that limits clinical relevance to human patients. 16 In this study, we established a new method to induce MI in mice

| Reagents and animals
Primary antibodies against CD31 were obtained from Cell Signaling Company. All drug concentrations were expressed as the final molar concentration in the buffer. Male wild-type (C57BL6) mice, 8-12 weeks of age, 20-25 g of body weight, were obtained from the Jackson Laboratory (Bar Harbor, ME). Mice were housed in temperature-controlled cages with a 12-hour light-dark cycle and given free access to water and normal chows. All animal studies were conducted at the Animal Institute of Central South University according to the protocols approved by the Medical Experimental Animal Care Commission of Central South University.

| Protocols of permanent coronary artery
occlusion without opening thoracic cavity 1. Sterilize surgical instruments with a dry bead sterilizer (Germinator 500).

2.
Mouse (generally 2-3 months of age or at least 18 g body weight) is anaesthetized with 2%-3% isoflurane inhalation in an inducing chamber.

3.
Once anaesthetized, the mouse is removed from the inducing chamber to the surgical board, immobilized with tape and continuously anaesthetized with 2% isoflurane via coaxial breathing apparatus but not ventilated.

4.
Remove the fur with a standard depilatory (eg, Nair) and clean the skin with water and then betadine and alcohol pads. To perform this procedure more efficiently, the step of fur-removing could be done earlier.

5.
Two small incisions (0.5 cm long) are made on the left and right chest skin with the scissors to expose the 3rd intercostal space as shown in Figure 1A.

8.
Then, the needle is inserted back from the right to of the left ( Figure 1D). When the needle passes through the heart, it goes through above LCA ( Figure 1E) under ultrasound and came out the skin from the same site in the left chest. For this step, the needle only passes through the space between the anterior chest wall and the anterior wall of the heart by gently pressing down both sides of suture to expand the space.

9.
Once a loose knot is made ( Figure 1F), the needle is inserted back from the left to the right in the chest. The LCA is now located inside of the knot ( Figure 1G).

11.
The mouse is then allowed to breathe room air and monitored on a heating blanket during the recovery period, which is generally complete within 3-5 minutes.

12.
The sham group undergoes the same surgical procedures except that the LCA is not occluded.

13.
One dose of buprenorphine (0.1 mg/kg) is administered subcutaneously and immediately after the incision is closed.
14. Clean surgical tools with PBS and alcohol.
For detailed procedures, please watch Video S1.

| Traditional method to ligate LCA with ventilation and thoracotomy
The traditional, ventilation-based method of MI in mice has been fully described by previous investigators with minor modifications. 17

| ECG analysis
Quantification of ECG was recorded before and after CICAL procedure using electrophysiological recording system (Taimeng Inc., Chengdu, China) as described previously. 18 Otherwise, changes in ECG in echocardiography imaging system were also recorded during the whole CICAL and CICAL I/R procedures. Briefly, the electrodes of the plate were connected with three mouse limbs skin to monitor the changes in ECG in echocardiography.

| Measurement of plasma cTn-I levels
The mice were killed 24 hours after the surgery. ELISA kits were used to assess the levels of serum cTn-I (Life Span Bio Sciences, Seattle, USA), following the manufacturer's instructions.

| Microfil angiography
Microfil (Flow Tech Incorporation, MA, USA) was injected into the left ventricle until it accumulated in the coronary vessels. After the casting material was completely hardened, the perfused organ was removed and placed in paraformaldehyde for several hours. Images were acquired on a Leica M205 FC stereo microscope.

| Echocardiography measurements in vivo
After anaesthetization, individual mouse was placed on a heating pad (37°C) and underwent echocardiography using a VEVO 2100 micro-ultrasound system with the echocardiography probe (MS-400).
Percentage changes in both LVEF and LVFS, the peak velocity (Vmax) of the AAO blood flow and E and A peaks were calculated and recorded using Visual Sonic analysis software (Visual Sonics Inc. Toronto, Canada).

| Determination of angiogenesis by immunohistochemistry
Histological analysis was assessed in perfusion/fixed hearts collected from mice at 28 post-operative days. The right atrium was then cut, and the myocardial vasculature was perfused, followed by 10 minutes perfusion with 10% formalin. The hearts were harvested and fixed in 4% formalin for 24 hours. The formalin-fixed tissues were embedded in paraffin wax and cut into 4 lm sections. For the measurement of capillary density (counts/mm 2 ), we performed immunohistochemical analysis of CD31 as described previously. 20 with Log Rank statistical method. A two-sided P-value <.05 was considered significant.

| Novel MIM is successful to establish MI model in mice
As described in Section of Methods, we used the new method to induce MI injury without ventilation and thoracotomy. To determine whether this new method is efficiently to induce MI, we visualized the heart and coronary arteries using microfil angiography and examined the MI size as described previously. 22 In all cohort (63 mice) used for MIM surgery, 2 mice died of LCA angiorrhexis within 6 hours after the surgery, and 3 mice were excluded because S-T enhancement was not observed during the MIM surgery. We did not observe the major complications such as pneumothorax in all mice.
As shown in Figure 2A-a,b, microfil angiography clearly outlined the LCA, which was ligated in mice hearts with the MIM surgery.
The area of MI was also evidently identified by the microfil angiography imaging (

| Serum cTn-I level is increased in mice after the MIM surgery
Acute MI, defined as an elevation of cardiac troponin I (cTn-I) resulting from ischaemia, is associated with substantial mortality in surgical patients. 23 Thus, we assayed serum cTn-I levels to further confirm the effects of MI induced by MIM. As indicated in Figure 2E, serum cTn-I levels were totally increased at 24 postoperative hours, compared to sham mice, suggesting the new method is effective to induce myocardial injury in mice.

| Cardiac dysfunction is induced by new method in mice
Cardiac dysfunction is a key factor contributing to the mortality in patients with MI. 24 We then examined heart function to assess the effects of MI in mice with MIM surgery by electrocardiogram (ECG) at 24 post-operative hours. As shown in Figure 3A, the S-T segments were markedly elevated in mice with MIM surgery, compared to mice with sham surgery. Moreover, both LVEF and LVFS were decreased in mice with MIM ( Figure 3B,C), indicating the impaired anterior wall motion of the heart caused by the new method.
The systolic and diastolic functions of hearts were also determined by calculating the Vmax of ascending aorta (AAO) blood flow and the E/A peak ratio. As depicted in Figure 3B,D,E, both Vmax of AAO and the E/A peak ratio were noticeably decreased in mice with MIM surgery, compared with the mice following sham surgery, demonstrating that this new method is a useful approach to induce cardiac dysfunction in mice.

| New method is comparable to traditional method to induce MI and cardiac dysfunction
We

| MIM surgery shortens recovery time and causes less inflammation
To evaluate the side injuries induced by MIM surgery, we compared the post-operative recovery time in mice between undergoing new method and traditional method. The recovery time was defined as the interval from surgical completion to the recovery of animal consciousness or motion. 15 As indicated in Figure 5A,

| New method of MIM surgery increases survivals of mice
The short/long survival rate was also calculated in mice with new method and traditional method. As shown in Figure 5D, from the beginning of the surgery to 6 hours after surgery, no mice died when each sham surgery was performed. The survival rate in mice with MIM surgery was approximately 95%, while it was about 75% in mice with MI by traditional method. The overall survival rates (including surgery-related death) 28 days after surgery was 100% in mice with sham surgery of new method and 95% in mice with sham surgery of traditional method ( Figure 5E). The overall survival rate at the 28th post-operative day was 90% in mice with MIM surgery, which was significantly higher than 45% in mice with MI by traditional method ( Figure 5E). These data confirm an improved outcome using this new method of MI.

| New method of MIM surgery induces angiogenesis
Angiogenesis is critical for re-establishing the blood supply to the surviving myocardium after MI and, consequently, to the recovery of cardiac function. 25 To test whether this model is suitable for studying angiogenesis, we detected the angiogenesis by analysis of CD31, which is a biomarker for endothelial cells. 26 As shown in Figure 6A,

| New method of MIM surgery induces cardiac myocyte apoptosis
Studies have also demonstrated an important role of apoptosis in ischaemic heart disease, contributing to myocyte cell death in MI and left ventricular remodelling. 27 To examine whether this model is applicable to study cardiac myocyte apoptosis induced by MI, we detected the apoptosis by TUNEL. As shown in Figure 6B, the apoptosis rate of cardiac myocyte in ischaemic area was completely increased in mice with the surgeries of both traditional method and new method, compared to sham mice. It reveals that MIM surgery is the same effective as the traditional method to be used in studying ischaemia-induced angiogenesis.

| Application of MI model established by new method to I/R model
The I/R model is generally used to examine the short-term consequences of ischaemic injury and has been used to investigate the development of innovative cardioprotective therapies. 14 Figure 8B). These results show that the modified MIM surgery is applicable in studying I/R injury in mice.

| DISCUSSION
In this study, we described a minimally invasive approach to ligate coronary without thoracotomy under the guide of ultrasound. This new method is more efficient to induce MI and causes less injuries, | 5217 thoracotomies. 28  A difficulty of this technique is that the real-time image of the heart as viewed on the monitor will be distorted which is also reflected in the video, due to movement in the chest wall musculature as well as the LV wall. The same would happen every time when the needle is inserted and there is movement in musculature.
Therefore, the ultrasound probe will have to be constantly readjusted to get a real-time accurate image of the heart, hence adding to the time taken to perform the procedure.
Another difficulty of this technique is how to control the depth of the needle so that it can precisely cross the LCA rather than punctures into the heart chamber. For this question, we should identify the LCA, needle and heart wall under ultrasound correctly (Figure 1b). When the needle reaches the inferior of LCA as monitored by ultrasound, the needle tip is going up a little by pressing the side of needle connected to the suture ( Figure S2 as a flash and Video S2). This manipulation is critical to avoid the incorrect puncture of heart chamber. For this step, it is better to use a straight needle but not a curved needle to do puncture.
Frankly, the traditional method can be performed with a very small incision in the chest cavity and minimal trauma. The careful intubation and a weight based artificial respiratory equipment are mandatory to obtain safe ventilation during the procedure. Due to evacuation of the thorax, intensive care after the surgery and good pain management, the mortality rate of traditional operated LCA ligation can be reduced to a minimum. Furthermore, the correct ligation can be evaluated directly while performing the surgery of traditional method of LCA ligation. Also, the traditional method can be trained with dead and intubated animals while the MIM method has to be performed under the ultrasound of a living animal. For these reasons, the traditional method still has its advantages, compared to the MIM method.
In conclusion, we have efficiently established a minimally invasive MI model without thoracotomy and ventilation. This new model is not only applied to study ischaemia-induced angiogenesis and apoptosis, but also developed to I/R model with minor modifications.

ACKNOWLEDG EMENTS
This study was supported by grants from the Natural Science Foun-