Protective effect of limonin against doxorubicin-induced cardiotoxicity via activating nuclear factor - like 2 and Sirtuin 2 signaling pathways

ABSTRACT The anti-tumor and anti-inflammatory effects of limonin have been established, here, we aim to explore whether limonin can induce protective effects against doxorubicin (DOX)-mediated cardiotoxicity which limits its clinical application. We found that limonin attenuated DOX-mediated cytoxicology of myocardial cell line H9C2 by measuring cell viability and reactive oxygen species (ROS) level. Additionally, limonin ameliorates DOX-induced cardiac injury in rat by examining the activity of lactate dehydrogenase (LDH), superoxide dismutase (SOD), glutathione (GSH), glutathione peroxidase (GSH-Px) and malondialdehyde (MDA) concentration, and histopathological changes. Mechanistically, it was shown that limonin partially abrogated the inhibition of Nuclear factor – like 2 and Sirtuin 2 signaling induced by DOX. Furthermore, limonin-mediated protective effects on DOX-mediated cytoxicology of H9C2 were rescued by a Sirt2-specific inhibitor or siRNA against Sirt2. Thus, this work reveals that limonin can suppress DOX-mediated cardiotoxicity by activating Nrf2 and Sirt2 signaling.


Introduction
Doxorubicin (DOX) is a highly effective anthracycline chemotherapeutic drug, but its clinical application is limited due to its dose-dependent cardiotoxicity. However, the sensitive predictive indicators and concrete mechanisms underlying DOX-mediated cardiotoxicity are still confusing and effective therapeutic drugs are urgently needed.
Increasing evidences have shown that the monomers or extracts from Chinese herbal medicine not only have a low toxicity, but also have significant anti-tumor effects, which have become a hot spot in the research of anti-cancer drugs. In 2017, European scientists found that citrus fruit can induce preventive effects on digestive system and upper respiratory tract cancer. With the development of comprehensive utilization of citrus, various effective components have been gradually extracted and separated, especially limonin in orange peel and pomelo peel. Limonin has been shown to exert inhibitory effects on cancer progression, such as limonin can inhibit lung cancer cell proliferation [1] and intestinal carcinogenesis [2]. Specifically, limonin was recently been confirmed to attenuate the stemness of various cancer cells [3][4][5][6]. Additionally, thanks to its low toxicity and mild adverse reactions, it has attracted a growing number of attention and favor of scholars on other diseases, such as Mahmoud et al. indicate that limonin can attenuate the hepatocellular injury induced by live ischemia and reperfusion [7]; Zhang et al. reveal that limonin ameliorates acetaminophen-induced hepatotoxicity [8]; and limonin also can attenuate dextran sulfate sodium-induced chronic colitis [9]. However, its effects on DOXmediated cardiotoxicity and related signaling pathways are never been reported.
In the present work, we first show that limonin could induce protective effects on DOX-mediated cardiotoxicity through in vitro and in vivo model. Mechanistic studies revealed that limonin rescued the inhibition of Nuclear factor -like 2 (Nrf2) and Sirtuin 2 (Sirt2) signaling induced by DOX, which is also essential for limonin-mediated protective effects.

Animal model
Sprague-Dawley rats (200-220 g) were purchased from GemPharmatech (Nanjing, China). Rats were housed in a standard experimental room (12 h light/dark cycle) and left for 1 week to acclimatize before starting the experiment. This study was followed by the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85-23, Revised 2016). All experimental procedures were following with the guidelines of the research ethics committee of Nanjing Medical University (Nanjing, China). Here, rats were randomly separated into four groups: control (0.9% saline), DOX treatment (10 mg/kg), and DOX plus Limonin treatment (5 and 10 mg/kg, Cat # 1180-71-8, Solarbio, Beijing, China). For control groups, it was administered the vehicles (0.9% saline) of the used drugs (every other day); For the DOX groups, they were received the vehicle (0.9% saline) of limonin for 1 week (every other day) and i.p. injection of DOX (10 mg/kg) on the fifth day of the experiment (every other day for four injections) without limonin injection; For limonin group, it was administered limonin (10 mg/kg) (every other day); For DOX and limonin treatment groups, they were received a daily dose of limonin (5 mg/kg i.p or 10 mg/kg i.p) for 1 week and i.p. injection of DOX (10 mg/kg) on the fifth day of the experiment (every other day for four injections) without limonin injection. Seven days after the DOX last administration, the rats were euthanized and subjected to other experiments.

Sample collection
By the seventh day of the experiment, rats were anesthetized with urethane (20% in a dose of 1 g/ kg) i.p. Blood was withdrawn from the abdominal aorta, and then centrifuged at 5000 rpm for 15 min (JanetzkiT30 centrifuge, Germany) for separation of clear sera to be used in the detection of cardiac enzymes. Hearts were dissected, washed with normal saline. Part of the heart tissue was homogenized in 5% w/v ice-cold phosphate-buffered saline (PBS) (0.01 M, pH 7.4); then, homogenate was centrifuged at 3000 rpm for 20 min and the supernatant was stored at −80°C till used for biochemical analysis. Another part of heart tissue was stored in −80°C for western blotting detection of proteins' level. Lastly, samples of cardiac tissue were stored in 10% formalin for histopathological and immunohistochemical study.

Histopathological examinations
The heart was fixed with 4% paraformaldehyde, embedded in paraffin and dehydrated in a series of ethanol. Then, tissue samples were cut into 5 μm slicers which were used for staining. Briefly, paraffin sections were dewaxed and hydrated followed by dying with hematoxylin and eosin (HE) solution for 20 mins and washed with tap water. Then, sections were differentiated with differentiation medium. Subsequently, sections were soaked in warm water (about 50°C) for 5 mins and placed into eosin dye solution for 2 mins, and rinsed with tap water. Sections were soaked in tap water for another 5 mins, dehydrated, transparent and sealed. Finally, the stained slicers were photographed under a microscope (VHX -7000, Keyence, Osaka, Japan). The cardiac injury was classified on the degree of vascular congestion, hemosiderosis, cloudy fatty degeneration, loss of muscle striation, apoptosis, and necrosis. The severity of the pathological lesions was graded semi-quantitatively depending on the degree and extent of the alteration as follows: score 0 was considered normal, score 1 (+) for mild changes, score 2 (++) for moderate changes, and score 3 (+++) for severe changes.

TUNEL staining assay
Cardiomyocyte apoptosis was analyzed by TUNEL staining using In Situ Cell Death Detection Kit (Roche Diagnostics GmbH, Mannheim, Germany). Briefly, after being fixed with 4% paraformaldehyde, CMs were incubated with permeabilizing solution for 30 min and were then treated in TUNEL reaction mixture for 1 h at 37°C. Morphological assessment was performed by fluorescence microscopy (20 × objective). Nine microscopic fields were randomly selected to observe at least 100 cells to assess apoptosis.

Small interfering RNA (siRNA) synthesis and transfection
The siRNA against Sirt2 and corresponding negative control (NC) were purchased from GenePharma (Shanghai, China). The transfection procedure was performed using Magic™ siRNA (Cat # CBMS250) according to the manufacturer's recommendation.

Statistical analyses
Data were denoted as the mean ± standard deviation (SD), single factor analysis of variance (ANOVA) was used for analysis, and non-paired t test was used for inter-group comparison. P < 0.05 was considered to be statistically significant.

Limonin improves cell viability of H9C2 cells suppressed by DOX treatment
We initially confirmed that DOX significantly reduced H9C2 cell viability in a concentrationdependent manner (Figure 1(a)). As shown in the dose-response curves, the median lethal concentration (LC 50 ) was about 1 μM for the 48 h time point, which was chosen for the treatment in this work. Then, H9C2 cells were pre-treated with different concentrations of limonin before being treated with DOX, it was found that limonin pretreatment attenuated DOX-mediated inhibition on the cell viability of H9C2 cells in a dose-dependent manner (Figure 1(b)). Notably, a higher concentration of limonin more than 25 μM showed no additional protective effects on cell viability; thus, the 5 μM of limonin was selected for the subsequent experiment in vitro. Additionally, LDH level was markedly upregulated by DOX treatment, which was rescued by limonin treatment (Figure 1(c)). Consistently, DOX-induced upregulation of CK level was partially abrogated by limonin treatment (Figure 1(d)).

Limonin attenuates DOX-mediated apoptosis on H9C2 cells
We continued to investigate the effects of limonin on DOX-mediated apoptosis of H9C2 cells. As shown in Figure 2(a,b), DOX remarkably induced H9C2 cell apoptosis as evident by the increase TUNEL-positive cells, this effect was attenuated by limonin treatment. Additionally, the expression of apoptotic executors (cleaved caspase3, Bax) was increased by DOX treatment, while the expression of apoptotic suppressor (Bcl-2) was decreased by DOX treatment, this effect was also rescued by limonin treatment (Figure 2(c,d)). Furthermore, DOX-induced upregulation of ROS was reduced by limonin treatment (Figure 2(e)). These results suggest that limonin can attenuate DOX-induced cytoxicology of H9C2 cells.

Limonin ameliorates DOX-induced cardiac injury in rat
As oxidative stress is regarded as the primary cause of DOX-induced cardiomyopathy, we detected the levels of oxidative stress markers (CAT, SOD, MDA, GSH, GSH-Px) in rat heart tissues. As shown in Figure 3(a,c), the cardiac activity of CAT and SOD, and ROS level were remarkably decreased by DOX treatment, which was rescued by limonin treatment. Additionally, MDA activity was increased by DOX treatment, this effect was attenuated by limonin treatment (Figure 3(d)). Furthermore, DOX-induced downregulation of GSH-Px level was partially reversed by limonin treatment (Figure 3(e)). Consistently, the heart tissues from rat exposed to DOX showed myocardial cell injury, as characterized by the interstitial edema, capillary congestion and inflammation infusion, this effect was ameliorated by limonin treatment (Figure 3(f)).

Limonin attenuates the Nrf2 and Sirt2 signaling inactivated by DOX
We then explored the mechanisms underlying limonin-mediated protective effects on DOXinduced cardiotoxicity. As the inhibition of Nrf2 and Sirt2 signaling has been shown to be responsible for DOX-mediated cardiotoxicity, the expression levels of Nrf2, Sirt2 and some of their downstream effectors (HO-1, Gst, NQO1, FOXO3a, GCLM, Keap1) were examined. As shown in Figure 4(a), the protein expression of Nrf2, Sirt2, HO-1, Gst, NQO1, FOXO3a, GCLM was inhibited by DOX treatment, while Keap1 expression was increased by DOX treatment, this effect was rescued by limonin treatment. Consistently, western blot analysis on rat heart tissues obtained similar results (Figure 4(b)).

Limonin ameliorates DOX-induced cardiotoxicity dependent on Sirt2
Finally, we investigated the essential role of Sirt2 on limonin-mediated protective effects on DOXinduced cardiotoxicity. Sirt2-specific inhibitor (AGK2) was added or Sirt2 siRNA was transfected in H9C2 cells with DOX treatment plus limonin incubation, it was found that limonin-mediated decrease of LDH and CK level in DOX-treated H9C2 cells was rescued by Sirt2 inhibition or knockdown ( Figure 5(a,b)). The inhibition efficiency of AGK2 and knockdown efficiency of Sirt2 siRNA were confirmed in H9C2 cells, respectively ( Figure 5(c,d)). Additionally, the protective effects of limonin on DOX-induced oxidative injury on H9C2 cells were partially abrogated by Sirt2 inhibition or knockdown, as evident by the change of TUNEL-positive cells ( Figure 5(e,f)), apoptotic regulator expression ( Figure 5(g)) and ROS level ( Figure 5(h)). Therefore, these results indicate that SIRT2 is necessary for limoninmediated protective effects on DOX-induced cardiotoxicology.

Discussion
The cardiotoxicity of DOX has become the major obstacle for its clinical application; however, the underlying mechanisms contributing to DOXmediated cardiotoxicity are still confusing. Thus, it is very important to elucidate the mechanisms, based on which potential drugs can be explored. Here, we identified that limonin can ameliorate DOX-mediated cardiotoxicity through in vitro and in vivo experiments. This is the first study, to our knowledge, revealing the limonin-mediated protective effects on DOX-mediated cardiotoxicity.
An increasing number of evidences have explored potential protective agents based on chemical drugs; however, the toxic dose and effective dose of protective agents based on the research of chemical drugs are similar, and the side effects of are aggravated when they are combined with chemotherapeutic drugs [10,11], which cannot be widely used in clinic. Recently, a large number of studies have found potential drugs that can improve the cardiotoxicity of DOX from the resources of traditional Chinese medicine, such as Zaid H Maayah et al. showed that resveratrol reduced DOX-induced cardiotoxicity in juvenile rat through inactivating the NLRP3-inflammasome [12]; Wenna Zhou et al. indicated that two alkaloids from hippophae rhamnoides linn. inhibited DOX-induced apoptosis of cardiac muscle cells via suppressing JNK signaling pathway [13]; Samar S Elblehi et al. demonstrated that date palm pollen extract averts DOXinduced cardiomyopathy fibrosis by suppressing apoptosis-targeting Bax/Bcl-2 and caspase-3 signaling pathways [14]. Limonin and other analogues are the main reasons for the bitterness of citrus fruits, which are generally enriched in citrus fruits, especially in seeds. Limonin has been shown to hold a wide range of biological activities, such as anti-tumor, anti-virus, anti-inflammatory and so on. However, its roles in protecting DOX-induced cardiotoxicity have never been investigated.
There are several mechanisms contributing to DOX-induced cardiotoxicity, for example, NLRP3inflammasome activation [12,15] and cell death The protein expression of Nrf2 and Sirt2, and their downstream effectors was examined in heart tissues from rat treated differently as indicated. DOX: 10 mg/kg, Limonin: 10 mg/kg. pathways including autophagy, ferroptosis, necroptosis, pyroptosis and apoptosis [16]. Here, we found that Nrf2 and Sirt2 signaling pathways were inhibited by limonin in vitro and in vivo by examining the expression levels of Nrf2, Sirt2 and some of their downstream effectors (HO-1, Gst, NQO1, FOXO3a, GCLM, Keap1). Importantly, Nrf2 and Sirt2 signaling pathways have been confirmed to be critically involved in DOXinduced cardiotoxicity [17,18]. Our results further strengthened these two proteins, Nrf2 and Sirt2, could be potential targets for DOX-induced cardiotoxicity. In addition, the previous study has shown that miR-140-5p could aggravate DOXinduced cardiotoxicity via targeting both Nrf2 and Sirt2 [17]. Plenty of evidences have shown that miRNAs are engaged in DOX-induced cardiotoxicity, such miR-133b [19], miR-98 [20] and miR-152 [21]. However, it is still unclear whether miR-140-5p is involved in limonin-mediated protection on DOX-induced cardiotoxicity, which could be explored in the future.

Conclusions
Taken together, this study first revealed the protective effects of limonin on DOX-induced cardiotoxicity. Although more experiments are needed to reveal its concrete mechanisms underlying limonin-mediated effects, limonin might be a penitential drug for DOX-induced cardiotoxicity.

Limitation/short-coming of the study
The concrete targets of limonin-mediated protective effects on DOX-induced cardiotoxicity are still unclear.

Future direction
Further transcriptome-sequencing and Co-IP combined with LC-MS technology should be constructed to reveal the concrete targets of limonin responsible for DOX-induced cardiotoxicity.