O-ring-induced transverse aortic constriction (OTAC) is a new simple method to develop cardiac hypertrophy and heart failure in mice

Suture-based transverse aortic constriction (TAC) in mice is one of the most frequently used experimental models for cardiac pressure overload-induced heart failure. However, the incidence of heart failure in the conventional TAC depends on the operator’s skill. To optimize and simplify this method, we proposed O-ring-induced transverse aortic constriction (OTAC) in mice. C57BL/6J mice were subjected to OTAC, in which an o-ring was applied to the transverse aorta (between the brachiocephalic artery and the left common carotid artery) and tied with a triple knot. We used different inner diameters of o-rings were 0.50 and 0.45 mm. Pressure overload by OTAC promoted left ventricular (LV) hypertrophy. OTAC also increased lung weight, indicating severe pulmonary congestion. Echocardiographic findings revealed that both OTAC groups developed LV hypertrophy within one week after the procedure and gradually reduced LV fractional shortening. In addition, significant elevations in gene expression related to heart failure, LV hypertrophy, and LV fibrosis were observed in the LV of OTAC mice. We demonstrated the OTAC method, which is a simple and effective cardiac pressure overload method in mice. This method will efficiently help us understand heart failure (HF) mechanisms with reduced LV ejection fraction (HFrEF) and cardiac hypertrophy.

OTAC promotes myocardial hypertrophy and heart failure. OTAC results in reproducible and graded myocardial hypertrophy. Representative images showing cardiac morphology (Fig. 3A). Pressure overload by OTAC promoted increased whole heart weight ( Fig. 3B) and LV hypertrophy represented by increased LV weight/TL (Fig. 3C). At 4 weeks post-procedure, LV weight/TL was, on average 1.4-and 1.5-fold significantly higher in OTAC0.50 and OTAC0.45, respectively, than in Sham mice ( Fig. 3C and Table 1). Moreover, at 8 weeks post-procedure, the whole heart and LV weight in OTAC-operated groups had significantly increased compared to the Sham group (Fig. 3B,C, and Table 1). LV weight/TL was, on average, 1.8-and 2.1-fold significantly higher in OTAC0.50 and OTAC0.45, respectively, than in Sham mice ( Fig. 3C and Table 1). The highest increase in LV weight/TL was observed for OTAC0.45. Thus, OTAC resulted in a reproducible and graded hypertrophic response. In addition, the right ventricle (RV)/TL weight of OTAC groups was significantly higher than in Sham   www.nature.com/scientificreports/ at 8 weeks post-procedure ( Fig. 3D and Table 1). The atria (AT)/TL weight of OTAC0.50 was significantly higher in Sham at 8 weeks post-procedure. Pressure overload by OTAC also promoted increased lung weight, indicating severe pulmonary congestion, as previously reported 13,20 . At 4 weeks post-procedure, the lung weight/TL ratio was on average 1.2-fold higher in both OTAC0.50 and OTAC0.45, respectively, than in Sham mice ( Fig. 3F and Table 1). At 8 weeks postprocedure, the lung weight/TL ratio significantly increased by 1.4-and 1.7-fold in OTAC0.50 and OTAC0.45, www.nature.com/scientificreports/ respectively, compared to Sham ( Fig. 3F and Table 1). The highest increase in lung weight/TL was observed for OTAC0.45. Further analysis revealed that at 4 weeks post-procedure, 14% (2/14) of OTAC0.50 and 21% (3/14) of OTAC0.45 mice developed a lung weight/TL ratio above the maximum value for Sham mice, and at 8 weeks post-procedure, 57% (8/14) of OTAC0.50 and 43% (6/14) of OTAC0.45 mice developed a lung weight/TL ratio above the maximum value for Sham mice ( Fig. 3G and Table 1).

Echocardiographic findings of OTAC-induced cardiac hypertrophy and reduced cardiac function.
In response to pressure overload, echocardiography showed that PWTd and IVSd had significantly increased in both OTAC groups than the Sham group at 1-8 weeks post-procedure ( Fig. 4A-E and Table 3). The LV mass also increased during hypertrophy (Fig. 4F). To determine whether the increased mortality and incidence of HF by OTAC, which was related to a more severe cardiac dysfunction, we also assessed LV systolic function by echocardiography post-procedure. Echocardiographic analysis revealed that LVFS was reduced early in both OTAC groups and continued to decline ( Fig. 4G and Table 3). LVFS in both OTAC groups was significantly reduced compared to the Sham group at 1-8 weeks. LVIDd was significantly larger in OTAC0.45 at 4-8 weeks, and LVIDs were significantly larger in OTAC0.50 at 4-8 weeks and in OTAC0.45 at 1-8 weeks than in the Sham group (Fig. 4H, I and Table 3). Representative cardiac hypertrophy and HFrEF cases are shown in Fig. 4A-C.
Signature molecules of cardiac remodeling, failure, and fibrosis. In mice that survived 4 and 8 weeks after the procedure, we performed real-time quantitative PCR (qPCR) to determine the gene expression levels of the markers of HF, hypertrophy, and fibrosis in the LV of OTAC-induced HF model mice. The mRNA expression levels of natriuretic atrial peptide (Nppa), natriuretic brain peptide (Nppb), and actin, alpha 1, skeletal muscle (Acta1) were significantly higher in both OTAC groups than in the Sham group ( Fig. 7A-C).
We also assessed the gene expression of fibrosis-related molecules in the myocardial tissue. The mRNA levels of collagen type Iα1 (Col1a1) had significantly increased in OTAC0.50 at 4 weeks and OTAC0.45 at 8 week, and collagen type IIIα1 (Col3a1) were significantly increased in both OTAC groups compared to the Sham group at 4 and 8 weeks (Fig. 7D,E). These results are consistent with the histological findings.

Discussion
This study demonstrated an innovative method to induce cardiac pressure overload using the OTAC technique in mice. This method can be easily performed without intubation and is highly reproducible for simulating cardiac hypertrophy, fibrosis, and HFrEF (Tables 1 and 3). In addition, the lower inner diameters of the o-ring can induce an increasing degree of LV hypertrophy, LV dysfunction, and lung congestion. Our method can also be applied to cardiac hypertrophy models with or without HF by simply selecting an appropriate-sized o-ring. This feature is similar to the selection of the needle size in a conventional TAC procedure 13 .
Various surgical approaches have been developed to mimic patients with hypertensive heart disease to study heart failure with reduced ejection fraction induced by chronic pressure overload of the left ventricle. Rockman et al. first reported the TAC method 9 , the most widely used to study the LV pressure overload that induces HFrEF 21,22 . TAC causes an increase in afterload due to stenosis of the transverse aorta, resulting in concentric hypertrophy, interstitial and perivascular fibrosis, and HFrEF 9,12,13,23,24 . However, the TAC procedure has certain drawbacks. As it is highly operator-dependent, it has inter-operator variability and is technically demanding. This drawback might lead to variable degrees of aortic constriction, resulting in heterogeneous phenotypes. The mortality rate of mice undergoing TAC is difficult to estimate because the exact cause of the death is often not declared, and only successfully performed TAC procedures have been mentioned. Surgical variables such as mechanical or spontaneous ventilation 15 , type of approach (thoracotomy 9 vs. mini-sternotomy 15 ), site of the constriction, severity of the stenosis produced (cannula size 13 ), and operator experience could affect the mortality in the conventional mice TAC model. Therefore, the mortality range in the TAC mouse model varies remarkably in several studies 15,16 .
Recently, two modified aortic constriction methods have been reported. One is the double-loop-clip technique 25 . In this method, the transverse aorta diameter is measured using echocardiography before the procedure to calculate the inter-knot span of the suture, which can decrease variability. This new procedure results in a far more accurate, reproducible stenosis that decreases mouse mortality and increases the homogeneity of structural and molecular features after aortic constriction 25 . The other method is the ORAB method 18  www.nature.com/scientificreports/ method involves constriction of the ascending aorta of mice using an o-ring with fixed inner diameters. A fixed inner diameter can easily achieve ascending aorta constriction for hypertrophy and HF. These methods are more reproducible than conventional TAC and may be used in place of TAC in the future, further advancing HF research. In addition to the two modified aortic constriction methods, our proposed OTAC method has various advantages. First, a pressure overload can be applied by tying the thread on the o-ring [ Fig. 1D-F and Supplemental Fig. 4 (10, 11)]. In the conventional TAC method, there is a possibility that the thread may loosen when the www.nature.com/scientificreports/ needle is removed from the thread; this is one of the causes of phenotypic variation and is improved by OTAC, wherein there is no concern regarding the thread loosening. The other possible cause was intraluminal suture migration. This phenomenon, resulting from tying with single-loop banding, appears in up to 30% of the animals after the conventional TAC procedure 26,27 . The o-ring, which is made of non-slip rubber and has a width greater than that of the thread, helps overcome this drawback. OTAC can develop concentric hypertrophy, interstitial and perivascular fibrosis, and has a high incidence of HF compared to conventional methods. (Fig. 3, Tables 1  and 3). In the comparison between OTAC (the surgeon with about 100 case experiences of OTAC and 20 case experiences of TAC) and TAC using the needle of 26 gauge (26G) (the surgeon with about 350 case experiences of TAC) in our laboratory, IVSd and PWTd were higher, and LVFS was lower in OTAC mice than 26G TAC mice (Supplemental Fig. 2). The number of cases experienced by the surgeons was relatively small, and they were considered to be beginners. Mice of 26G TAC were performed for the same weeks, body weight, and C57BL/6 J mouse as those of OTAC. This result shows that OTAC is a useful method for beginners to create the HFrEF model easily. Moreover, the features of HFrEF in our model are more represented than those in previously reported TAC mouse models [14][15][16] . OTAC can be easily diverted from the conventional TAC, and it may be useful even in laboratories with expertise in the conventional TAC mouse model. Second, OTAC can perform with mini-sternotomy and does not require intubation. The fact that OTAC does not require intubation of the mouse is an important advantage over ORAB. In addition, perioperative blood loss is minimal, and there was no risk of complications associated with intubation during OTAC. Anesthesia during OTAC involves intraperitoneal anesthesia. Therefore, the time needed for intubation can be reduced. In addition, the cost of the o-ring is low. Third, periprocedural mortality is likely comparable to that of conventional TAC. Periprocedural mortality is an essential point from an animal welfare perspective. According to past TAC reports, the mortality rate is approximately 0-45% [13][14][15][16][17] , which is considered acceptable from the point of our mortality rate of 11% (Table 4). OTAC is placed on the transverse aorta, which may slowly progress toward hypertrophy and HF compared to the constriction of the ascending aorta with o-rings. Although we could not directly compare OTAC and ORAB in Table 3.  www.nature.com/scientificreports/ this study, hemodynamics and phenotypes depending on the location of aortic constriction have been previously reported 28 . The constriction of the ascending aorta leads to the immediate onset of pressure overload and trauma to the vascular wall. Additionally, aortic occlusion produces LV overdistention and acute myocardial damage that may kill the animal or exhibit myocardial artifacts at termination studies. This occlusion is particularly relevant for ascending aortic constriction techniques 17,29,30 , where catastrophic consequences quickly follow complete aortic occlusion. OTAC for the transverse aorta could prevent these traumas, because in this procedure, the presumed cause of death post-procedure has been suspected to be HF without bleeding in the thoracic cavity by necropsies. Moreover, unlike the TAC procedure, it is possible to reduce the mortality rate during the surgery such that the OTAC could avoid complete occlusion of the aorta 29,31 . This OTAC method may also have better reproducibility of HF occurrence if the following features are improved. OTAC used the same inner diameter of o-rings regardless of the body weight and aortic size of the mice. O-rings form similar aortic lumen areas in the mice despite different aortic diameters. It has been reported www.nature.com/scientificreports/ that the double loop-clip technique causes a similar percent reduction of aortic flow area in dissimilar animals using aortic diameter measured by echocardiographic techniques before the TAC procedure 25 . Similarly, we believe that measuring the aortic diameter of mice by echocardiographic techniques before OTAC and using an appropriate size of the o-ring may lead to high reproducibility of left ventricular hypertrophy and HF in mice.
The hypertrophic response to pressure overload and progression to HF depends on the genetic background of the mice. Lorena et al. reported that the C57BL/6J substrain appears to be a model of sustained cardiac hypertrophy, and C57BL/6NCrl and C57BL/6NTac substrains are more suitable for HF 20 . Therefore, according to this report, OTAC may be more reproducible in C57BL/6NCrl and C57BL/6NTac substrain mice for HF studies.
The severity of the constriction is usually assessed by measuring pulsed-wave Doppler images of the aortic arch 16 . However, in this OTAC model, the artifacts of the o-ring make it difficult to assess the signals. The measurement of the peak flow velocity difference between the right and left carotid artery post-procedure enables the quantification of the pressure gradient 32 . The present study could not quantify the pressure gradient because it was not measured invasively; therefore, the pressure gradient might vary among the OTAC groups, contributing to the phenotypic differences. It could be considered to measure pressure gradients noninvasively by measuring blood pressure in the right and left arms of mice to solve this limitation. If these measurements can be applied to the OTAC model, a more uniform phenotype can be achieved. Therefore, research projects on cardiac hypertrophy, fibrosis, and HFrEF will make considerable progress.
This study had several limitations. First, the echocardiographic and tissue assessments were not randomized and could be biased. However, since OTAC mice in this study showed overt hypertrophy, the effect of this may be relatively small. Second, the OTAC is performed by a single operator and the inter-operator variability has not been verified. However, in previous reports there was little inter-operator variability when o-rings were used for the pressure overload model in mice 18 . In the future, it will be necessary to verify the accuracy of the operators to ensure that inter-operator variability does not cause any significant differences in the procedure. Finally, there may be a potential bias due to the small number of mice that underwent surgery. OTAC0.45 mice scheduled for sampling at eight weeks had more cases of heart failure death from late-phase post-procedure to one week after OTAC than OTAC0.45 mice scheduled for sampling at four weeks after OTAC (5/14 vs. 0/14). Therefore, due to censoring, the Kaplan-Meier curve produced a higher survival rate than the actual sampling group at eight weeks after OTAC. Furthermore, mice that underwent OTAC0.45 had a higher mortality rate than those who underwent OTAC0.50. Therefore, the exact phenotype of OATC0.45 could not be reflected because the survived mice probably had a lower pressure overload than the dead mice. Therefore, although the degree of hypertrophy with OTAC 0.45 was expected to be higher than that with OTAC 0.50, IVSd and PWTd by echocardiography, cardiomyocyte hypertrophy by histology, and actual left ventricular weight were similar between the two groups. These results were suspected to occur due to the bias, and the difference between the two groups would resolve if the number of cases increases.
In the present study, we demonstrated a murine model of heart failure, which is a simple and highly efficient LV pressure overload model. This model will efficiently help scientists conduct research related to HFrEF and cardiac hypertrophy.

Methods
Experimental animals. Male, 9-10 weeks old (26.0 ± 0.5 g) C57BL/6J mice (Clea Japan, Inc., Tokyo, Japan) were randomized for sham operation (Sham) or two OTAC operation groups using 0.50 mm inner diameter o-ring (OTAC0.50) or 0.45 mm inner diameter o-ring (OTAC0.45). The o-rings for this study were confirmed with an optical microscope before the procedure, and inferior ones, such as roughness on the inner surface, were excluded. The mice were observed up to 4 and 8 weeks after the procedure. Regarding our preliminary experiment, the number of mice that underwent the OTAC procedure was set as 14 (total n = 84) in each of the six groups (Sham, OTAC0.50, or OTAC0.45 followed up for 4 or 8-weeks) at the early phase post-procedure (48 h after the procedure). All mice were maintained in ventilated cages with free access to standard rodent feed and sterile water in a temperature-controlled environment under 12/12 h light/dark cycles.  www.nature.com/scientificreports/ OTAC for induction of pressure overload in mice. Mice were initially anesthetized using a mixture of medetomidine, midazolam, and butorphanol (0.3 mg/kg, 4.0 mg/kg, and 5.0 mg/kg, respectively) with intraperitoneal administration. The mice were breathing spontaneously without intubation. A mini-sternotomy was performed by making a 2-3 mm longitudinal cut in the proximal portion of the sternum, and the pectoralis muscles were blunt dissected. After retracting the thymus, the transverse aorta was identified by its location posterior to the thymus gland. For quality control, we carefully dissected the fat tissue that would alter the aortic diameter as much as possible. In OTAC groups, mice were operated with inner diameters of 0.50 mm o-ring purchased from SAKURA SEAL (Tokyo, Japan) or 0.45 mm o-ring purchased from Apple Rubber (Lancaster, NY, USA). The details on the preparation of the o-ring are described in Supplemental Fig. 3. An o-ring was applied around the transverse aorta (between the brachiocephalic and left common carotid artery) and tied with a triple knot www.nature.com/scientificreports/ ( Fig. 1D-F). The details of the OTAC procedure are described in Supplemental Fig. 4. The incision was closed by suturing the layers using 6-0 silk sutures. Sham mice were subjected to the same operation without the insertion of an o-ring. The mice initially recovered over a warming pad and were observed throughout the perioperative period.
Echocardiography. Transthoracic echocardiography (TTE) was performed at baseline and at weeks 1, 2, 3, 4, 5, 6, 7, and 8 after procedures using the VEVO 1100 Imaging System (VisualSonics, Toronto, Canada) with a transducer probe MS400 (VisualSonics, Toronto, Canada, with a frequency of 30 MHz). Fur was removed on the chest using depilatory cream (Veet, USA), ensuring that any residual cream was removed fully with water. For echocardiography in awake mice, we picked up the mouse by the nape and held it firmly in the palm of one hand in the supine position, with the tail held tightly between the last two fingers. The probe was gently placed over the fourth and sixth left ribs. M-mode echocardiography was recorded at the papillary muscle level from www.nature.com/scientificreports/ the parasternal short-axis view without sedation. In each of these captured images were included 10 to 20 cardiac cycles. Left ventricular internal dimension in end-diastolic dimension (LVIDd) and end-systolic dimension (LVIDs), the posterior wall thickness in diastole (PWTd), and interventricular septum thickness in diastole (IVSd) were measured over three cardiac cycles. Left ventricular fractional shortening (FS) was calculated using the formula FS = [(LVIDd − LVIDs)/(LVIDd)] × 100 from the M-mode measurements. These data were averaged from at least three cycles per loop.
Tissue isolation. Mice were euthanized by cervical dislocation under sedation using a mixture of medetomidine, midazolam, and butorphanol (0.3 mg/kg, 4.0 mg/kg, and 5.0 mg/kg, respectively). The heart was excised, AT was removed, and the RV was separated from the left ventricle (including the septum). The individual chambers were weighed to assess cardiac hypertrophy. The LV was divided into the mid-papillary and apical regions. LV was snap-frozen in liquid nitrogen or stored in 10% formalin for histological analysis. The lungs were weighed, and the tibia length (TL) was measured for normalization. TL was also measured during the necropsy.