Methanolic extract of Potentilla fulgens root and its ethyl-acetate fraction delays the process of carcinogenesis in mice

People of north-eastern states of India consume raw areca-nut (RAN) and lime which could lead to oral, esophageal and gastric cancers. However, the incidence of these cancers are significantly lesser in those who consume pieces of Potentilla fulgens root along with RAN. Since evaluation of anticancer role, if any, of P. fulgens on RAN-mediated genetic alterations in human is difficult because of other compounding factors, this study was undertaken in mice to focus on gastric carcinogenesis since ad libitum administration of RAN extract with lime in drinking water induced stomach cancer due to greater exposure of its lining. A total of 160 mice were used at different time points and either methanol extract of P. fulgens roots (PRE) or mixture of four compounds of ethyl-acetate fraction (EA-mixture) was mixed with mice feed. Histological studies revealed that RAN + lime induced cancer in all the mice and interestingly only 20% developed cancer when PRE/EA-mixture was provided along with RAN + lime. Higher frequency of precocious anaphase and over expression of p53 and Securin genes were significantly reduced by PRE/EA-mixture. Thus PRE/EA-mixture mitigates the RAN-induced tumor-initiating process in stomach by maintaining expression of tumor suppressor and check-point genes under control.


Results
General observation. In total, 160 mice were used at different time points for different experimental analysis (Table 1). Mice treated with RAN + lime with and without PRE for 260 days (n = 10 at each point) were used for both immunohistochemistry and histopathological evaluation. The histological section clearly differentiated among untreated normal (Fig. 1A), hyperplasia (Fig. 1B) and adeno-carcinoma ( Fig. 1C) in stomach. The histological preparation of stomach tissue was made from all the mice in both the treated and untreated group (n = 5). All 10 mice treated with RAN + lime developed cancer in stomach. However, two out of 10 mice treated with PRE + RAN + lime showed carcinoma whereas 5 showed hyperplasia, a precancerous lesion in stomach. Interestingly, 3 mice in the PRE + RAN + lime group showed normal stomach. After 260 days of treatment, cancer was observed to be induced in all the ten mice treated with RAN + lime whereas only 20% mice developed cancer after treatment with PRE + RAN + lime.
Studies on metaphase spreads. Continuous exposure to RAN extract with lime induced chromosomal instability which was evaluated in mouse BMC. We scored metaphase spreads 3 h after colchicine treatment and the data revealed a gradual but significant increase in the frequency of prematurely separated sister chromatids (Fig. 1D) in RAN + lime treated mouse compared to untreated one ( Table 2). The frequency of premature anaphase separation (PAS) was 23.8% and 27.8% after 120 and 180 days of continuous administeration of RAN + lime, respectively. Interestingly, the frequency of PAS was reduced significantly if either PRE or mixture of epicatechin (E), catechin (C), gallic acid (G) and ursolic acid (U) (ECGU) was present along with RAN + lime (Fig. 1D, Tables 2 and S2 [Supplemetary section]). A normal metaphase cell showed 40 chromosomes (2 N = 40) in mouse BMC (Fig. 1E) whereas Fig. 1F,G showed metaphase spreads with prematurely separated sister chromatids. We performed chromosome count on metaphase spreads to test the importance of precocious anaphase on chromosome stability. The untreated mice showed a stable (2n = 40) karyotype without showing any precocious anaphase and aneuploid cells ( Table 2). The frequency of aneuploidy cells was increased gradually from 0.7 to 2.6% after 60 to 180 days of treatment with RAN + lime, respectively. Interestingly, the frequency of both precocious anaphase and aneuploidy cells were reduced significantly when mice fed with either PRE (  Total  Mice  60d  120d  180d  260d  60d  120d  180d  260d  60d  120d  180d   Metaphase  preparation  5  5  5  5  5  5  5  4  4  4  47   Western Blot  5  5  5  5  5  4  29   qPCR  10  6  6  6  6  34   8-Oxo-dG  ELISA  5  5  5  5  5  www.nature.com/scientificreports www.nature.com/scientificreports/ for 180 and 260 days (5 mice per point) ( Fig. 2A,B). It was clear that the level of 8-OHdG was increased after RAN + lime exposure and the degree of induction was more in stomach than esophagous. Interestingly, when mice were consumed PRE and RAN + lime together the level of 8-OHdG was reduced. pRe maintained the expression status of mitotic checkpoint genes in RAn + lime fed mice. We examined the expression of Mad2 and AuKA genes in cells from the inner layer of stomach of mice untreated (n = 10) or administered RAN + lime with or without PRE for 180 and 260 days (n = 6 per point). The quantitative RT-PCR results (Fig. 2C,D) showed that the expression of Mad2 and AuKA genes was reduced considerably in RAN + lime treated mice with respect to untreated control. However, such reduction in the expression was not seen when PRE was present. In fact, presence of PRE maintained the expression of both these genes almost similar to untreated control level. PRE or ECGU maintained the level of p53 and Securin protein: Immunoblotting studies. Levels of p53 and Securin in cells from inner layer of esophagous and stomach of those mice which were untreated (n = 5) or administered RAN + lime with or without PRE for 180 and 260 days (n = 5 per point) were evaluated by immunoblotting. Similar analysis was also performed in mice administered RAN + lime with or without ECGU for 180 days. Results indicate that the level of these proteins were elevated significantly in both the tissues after administration of RAN + lime. However, such elevated level was reduced significantly in presence of either PRE or ECGU along with RAN + lime ( Fig. 3A-F).
Immunohistochemical staining for p53 and Securin in RAN induced cancer samples. We studied p53 and Securin expression by immunostaining in a panel of mouse stomach samples collected from untreated (n = 5) and treated with RAN + lime (n = 10) or PRE + RAN + lime (n = 10) for 260 days (Fig. 4A-F). The sections of stomach from RAN + lime treated (Fig. 4B,E) showed significantly higher expression of both p53 and Securin than samples collected from untreated mice (Fig. 4A,D). The expression of both p53 and Securin was observed in the nucleus but in cancer samples Securin was also seen in the cytoplasm as well (Fig. 4E, mice 2 and 4 in the enlarged inset with an arrow). This elevated expression of both p53 and Securin genes were reduced www.nature.com/scientificreports www.nature.com/scientificreports/ significantly when mice were treated with PRE + RAN + lime (Fig. 4C,F). H-score of p53 varied from 100 to 194 and for Securin it varied from 94 to 190 in RAN + lime treated samples (Fig. 4G,H). The mean H-score for p53 came down from 152 to 94 and for Securin it came down from 166 to 84 in PRE + RAN + lime treated mice.

Discussion
Natives of north-eastern regions of India use P. fulgens root-stock and herb as folk-medicine against various ailments 16 . Evaluation of the total phenolic and flavonoid contents of P. fulgens root extract revealed the gallic acid and quercetin to be 138.8 ± 1.6 mg and 401.6 ± 4.6 mg per gm, respectively 22 . The root extract increased the survival of mice bearing Ehrlich ascites tumour cells and reduced cell viability in a dose-dependent manner in MCF-7 cells 22 . The ethyl-acetate (EA) soluble fraction of methanolic extract of P. fulgens root exhibited more efficient growth-inhibitory effect in MCF-7 and U87 cancer cell lines than hexane and n-butanol soluble fraction 23 . This observation prompted further purification of EA-fraction and a total of nine compounds including two new ursane type triterpenoids, Fulgic acid A and Fulgic acid B with a good antioxidant activity were identified 24 . Since the natural compounds act through diverse mechanisms and show minimal undesirable side effects 25 , it is attractive to evaluate anticancer activities of natural plant-based compounds in mammalian system.
In this study, mice were given the whole RAN-extract and lime ad libitum in drinking water with dose increased every two months to mimic the human habit of consumption of RAN. In earlier studies with similar mode of treatment, parameters like precocious anaphase and expression of Securin and p53 genes were monitored  www.nature.com/scientificreports www.nature.com/scientificreports/ at different time-points starting from 30 days to 300 days for understanding the process of RAN-induced carcinogenesis 8 . It was found that after 240 days of ad libitum administration of RAN extract with lime in drinking water all the mice developed gastric tumour. Therefore, in the present study we monitored the influence of PRE on carcinogenesis after 180 and 260 days of treatment.
We examined the anticancer potency of PRE or of a mixture of epicatechin (E), catechin (C), gallic acid (G) and ursolic acid (U). in mice in vivo. These four compounds were selected primarily because of their higher yields: E (2.7%), C (1.6%), G (3.6%) and U (3.3%) in crude methanol extract 24 . Following the practice in Potentilla based traditional Chinese medicine 26 , we provided PRE or ECGU to experimental mice one month before RAN + lime treatment in order to strengthen their physiological system. In fact, it was suggested that such pretreatment of Potentilla based traditional Chinese medicine is fortifying the stomach, promoting the production of body fluid, invigorating the vital energy and nourishing the blood 26 .
Autoxidation of areca-nut polyphenols at alkaline pH and consequent generation of hydrogen peroxide and superoxide radicals accelerates RAN-induced carcinogenesis 2,27 . Further, transition metal ions such as Cu −2+ , Mn ++ , Fe 2+ and Fe 3+ present in RAN and betel-leaf, facilitate the production of more reactive oxygen species which in turn contributes to initiation and promotion of cancer 28 . DNA is biologically significant target of oxidative attack, and it is widely believed that continuous oxidative damage to DNA is a significant contributor to the development of the major disorders including cancers. Among the diverse oxidative DNA damages, 8-hydroxy-deoxyguanosine (8-OHdG) is one of the most important DNA lesions and has been used as a marker for DNA oxidation 29 . Present results indicate that the high level of 8-OHdG following RAN + lime treatment was considerably reduced when PRE was co-administered. Thus, the presence of PRE or ECGU reduced DNA damages induced by RAN + lime mediated reactive oxygen species because both EA and butanol fractions of PRE showed a significant antioxidant potential, as evaluated by DPPH• and ABTS + • (hydrogen atom transfer based) and MTT assay system 24 . Moreover, Corosolic acid which is present in EA-fraction of PRE, is known to deplete the level of 8-OHdG and reduce the oxidative stress 24,30 .
Earlier studies highlighted the importance of precocious anaphase, which leads to aneuploidy, as a potential screening marker for identification of mitotic checkpoint defects during early days of RAN exposure in both mice and human 8,9,31 . As observed earlier 8 , present results also showed a gradual increase in the frequency of precocious anaphase and aneuploidy in the BMC of mice following RAN + lime administration. Very significantly www.nature.com/scientificreports www.nature.com/scientificreports/ frequencies of both were reduced following co-administeration of both PRE or ECGU. It is known that the percentage of aneuploid cells correlates with the severity of premature sister-chromatid separation, suggesting a direct relationship between the severity of loss of checkpoint control and chromosome mis-segregation 8,9,31 . Partial loss of spindle assembly checkpoint gene like MAD2 in Hct116 cells and also in mouse primary embryonic fibroblast cells showed increased premature sister chromatid separation in the presence of spindle inhibitors and higher rate of chromosome mis-segregation events in the absence of spindle inhibitors 31 . Another study, showed that arecoline, an alkaloid component of areca-nut, upregulated the spindle assembly checkpoint genes like Aurora A, BubR1 and Mps1 which led to mitotic spindles distortion and misalignment of chromosomes 32 . In this study, downregulation of MAD2 and AuKA in stomach cells of the mouse was noted after administeration with RAN + lime but in the presence of PRE, their expression was similar to untreated control. Thus, the inhibitory effect of RAN + lime on some of the spindle assembly checkpoint genes can be abrogated significantly by either PRE or ECGU.
Elevated expression of p53 protein is well documented in head and neck squamous cell carcinoma 33,34 and also in oral dysplastic lesions and therefore such alteration in p53 is considered to be an early event in oral carcinogenesis 35 . Securin or pituitary tumor transforming gene, is known to be involved in the regulation of chromatid separation at metaphase-anaphase interface of the cell cycle 36 . Overexpression of Securin has been associated with the aneuploidy formation due to chromatid mis-segregation and is demonstrated in multiple cancer types 37,38 .
Overexpression of p53 as well as Securin genes in non-target cells like mouse BMC and in human peripheral blood lymphocytes as well as in stomach and esophageal cells after exposure to RAN + lime has also been reported 8,30 . Interestingly, similar elevated expression of both these genes in stomach as well as in esophageal cells seen in this study after RAN + lime treatment was found to be considerably reduced by PRE or ECGU exposure. Since cancer induction mainly happened in stomach due to the greater exposure, further immunohistochemical analysis was performed only in stomach, which fully agreed with the immunoblotting data.
In agreement with the above, only 20% mice developed RAN + lime induced stomach cancer in presence of PRE whereas in its absence all the RAN + Lime exposed mice developed cancer. Thus the presence of PRE or ECGU delays the RAN + lime induced stomach carcinogenesis by maintaining the normal expression status of www.nature.com/scientificreports www.nature.com/scientificreports/ tumor suppressor and check-point genes. This seems to explain the greatly reduced incidence of oral and esophageal cancers in human when P. fulgens root is consumed with RAN + lime. It is interesting that the mixture of four compounds of EA-fraction showed similar potentiality in anticancer property as of the PRE. Both catechin and epicatechin are well known natural antioxidant and showed anticancer property 39,40 . Gallic acid can induce cell death of various cancer cells 41 by depleting cellular glutathione and blocking EGFR signal pathway 42 . A pentacyclic triterpene ursolic acid has also been shown to have anti-inflammatory, antioxidant, and antitumor effects 43 . Consequently, all these compounds show anticancer potentialities with different targets.
This in vivo study thus demonstrated that the presence of either PRE or ECGU mitigates the RAN + lime induced tumor-initiating processes. Most strategies on drug development use either combination of several monotargeted drugs or a compound with multitargeting properties 20 . A recent, multiscale model for association of frequently mutated genes and inflammation mediated stomach, colon and liver cancer, based on information retrieved from the Gene Expression Omnibus, the Cancer Genome Atlas, and Gene Ontology database has suggested two stages of inflammation-induced tumorigenesis 44 . The present study, in agreement with earlier results 22,23 indicated that PRE or ECGU targets precancerous state as well as the transition from the precancerous to tumorigenic states. Further in-depth studies are needed to identify the driver pathways targeted by PRE or ECGU in inflammation-induced tumorigenesis.

Materials and Methods
plant material and extraction procedure. Roots of Potentilla fulgens were collected from ten different plants from Shillong peak regions of Meghalaya state of India (altitude 1700 m above sea level; 25°34 North latitude and 91°54 East longitude) after obtaining a proper approval of the forest officer. Taxonomic identification of the plant material was kindly confirmed by senior taxonomist of Department of Botany, North-Eastern Hill www.nature.com/scientificreports www.nature.com/scientificreports/ University, Shillong and a voucher specimen was deposited in the herbarium of the department (accession number 11906).
Extraction and isolation were performed earlier where ethyl acetate fraction was purified and all the compounds were identified and characterised 24,45 . Chemical structures were elucidated by spectroscopic methods, especially ESIHRMS and 2D NMR techniques 24,45 . In brief, P. fulgens roots (1 Kg) were shade dried and extracted with methanol using Soxhlet apparatus and obtained 250 gm of crude Potentilla root extract (PRE) which was suspended in water-methanol mixture (80:20) and partitioned with hexane, chloroform, ethyl acetate (EA), butanol and water soluble fractions. The EA-fraction was further subjected to vacuum liquid chromatography in silica-gel using hexane-ethyl acetate (0 to 90% EtOAc) and then chloroform-methanol gradients (1% to 50% MeOH) as elutent to yield five major fractions. It was already reported that nine compounds including new triterpenes and phenolics were identified and characterized from this EA-fraction of the methanolic root extract of P. fulgens 13 (pl see the supplementary section for detailed isolation scheme Fig. S1). In this study, we used PRE or mixture of 50 µg each epicatechin (E), catechin (C), gallic acid (G) and ursolic acid (U) obtained from EA-fraction.
Animals maintenance and treatment. Swiss albino mice weighing 25-30 gm and aged 2-3 months were maintained in community cages in a room with controlled temperature (20 ± 2 °C) and controlled lighting (12 h light; 12 h dark). Standard mouse diet (NMC Oil Mills Ltd., Pune, India) and water ad libitum were used in all experiments. The experiments were conducted in compliance with institutional guidelines and approved by North-Eastern Hill University "Institutional Ethics Committee (Animal Models)" Board.
In total, 160 mice were distributed into 4 groups: Group 1 was treated with simple drinking water considered to be untreated whereas group 2, 3 and 4 were administered RAN extract ad libitum in the drinking water with lime (pH 9.8) ( Table 1). It was estimated that each mouse consumed 1 mg of extract per day till 60 days after which the dose was increased to 2 mg per day till 120 days. Likewise, every 60 days later the dose was increased by 1 mg per day consumption till it reached to 4 mg per day. In this study, mice were fed till 260 days.
In Group 3, PRE was given along with RAN + lime. In fact, PRE-treatment was started 30 days before RAN + lime treatment. PRE 2 mg powder was mixed with 6 gm of mice feed and finally made a tablet for each mouse. Thus, each mouse consumed 2 mg PRE every day. In group 4, epicatechin, catechin, gallic acid and ursolic acid mixture (ECGU) were used instead of PRE. A stock solution (5 mg/ml) of each compound was made and from this 10 µl of each was added to 6 gm of mice feed and finally made the tablet for each mouse. This way each mouse was consumed 50 µg of each compound every day. chromosome preparation and scoring. For preparation of metaphase chromosomes, bone marrow cells (BMC) were collected 3 h after colchicine (15 mg/kg) treatment from 5 untreated mice as control. BMC were collected from untreated and 60, 120 and 180 days of different treated groups. The number of mice were used in each treated group are shown in Table 1. Animals were killed by cervical dislocation and femurs were dissected out for BMC which were treated with prewarmed KCl (0.075 M) and kept for 15 min at 37 °C. Then cells were fixed in acetic acid and methanol (1:3). Slides were prepared by flame drying method, stained with 5% Giemsa and mounted in a synthetic medium.
Around 100 well spread metaphase plates were studied for each mouse. We performed chromosome counts on metaphase spread. Values are expressed as mean ± SEMs. Both esophageal and stomach tissues were collected from the mouse of untreated and treated with RAN + lime with and without PRE for 180 and 260 days. For 8-OHdG measurement, DNA were extracted and digested with nuclease P1 (Sigma, USA) and further treated with calf intestinal phosphatase (Sigma, USA) and denatured. 8-OHdG was quantified by quantitative ELISA assay in 96-well plate format. The quantity of 8-OHdG in the specimens were determined by comparing its absorbance with known 8-OHdG standard curve.
RnA isolation and qRt-pcR analysis. Cells were collected from the inner layer of stomach from untreated mice (n = 10), RAN + lime and PRE + RAN + lime treated (n = 6 per point) for 180 and 260 days. Total RNA was extracted using an RNeasy Mini Kit (QIAGEN Co., Limburg, Netherlands). From 1 μg of total RNA, cDNA synthesis was performed using QuantiTect Reverse Transcription kit (Qiagen GmbH, Hilden, Germany) according to the manufacturer's protocol. For qPCR, cDNA was amplified using SYBER Green PCR mastermix according to the manufacture's cycling condition for 40 cycle on a StepOnePlus amplification and detection system (Applied Biosystems). The primers of target genes used for this analysis were Mitotic arrest deficient 2 (Mad2) and Aurora A Kinase (AukA), and the primer sequences are listed in Table S1 (Supplementary section). Data were analysed using the delta-delta Ct method and plotted as fold change versus control (pl see Supplementary section).
immunoblotting. Cells were collected from the inner layer of esophagous and stomach from untreated (n = 5), RAN + lime and PRE + RAN + lime treated mice for 180 and 260 (n = 5 per point) days. In case of 180 days treatment with ECGU + RAN + lime 4 mice were used as a treated group. The cells were washed with ice-cold 0.1 M phosphate-buffered saline (PBS; pH 7.4) and total protein was extracted with a lysis buffer containing 0.1% SDS, 2 mM EDTA, 1% NP-40, 1% sodium deoxycholate, 50 mM sodium fluoride, 100 U/ml aprotinin and 1 mM phenylmethylsulfonyl fluoride. After centrifugation, the cell lysate was collected and the protein concentration was determined using the bicinchoninic acid protein assay. Equal amount of protein (40 µg/well) were subjected to Novex Tris-Glycine 4-20% gradient gels and electrophoresis was performed in NuPAGE immunohistochemistry analysis. For immunohistochemistry (IHC) analysis, a small part of stomach tissue were collected from untreated and treated (both RAN + Lime with and without PRE) mice which were used for histopathological evaluation and kept in formalin. The tissues were dehydrated, paraffin embedded and sectioned with a microtome (Leica). Briefly, after blocking for endogenous peroxidase activity, the sections were incubated with anti-p53 (PAb 240; ab26; Abcam, USA) and anti-Securin (DCS-280; ab3305; Abcam, USA) primary antibody. IHC analysis was performed with a Strept-Avidin Biotin Kit (Dako). The scoring of immunohistochemical stains in each specimen was determined using a histological score (H) 48 (please see Supplementary section).
Statistics. The results are expressed as mean ± SEM for control and treated samples. Statistical analysis was performed by Student's t-test with GraphPad Prism software 5.1. The values were considered statistically significant, if the p value was 0.05 or less.

Data availability
The data that generated and supports the findings of this study will be available by the corresponding author upon request.