CPF impedes cell cycle re‐entry of quiescent lung cancer cells through transcriptional suppression of FACT and c‐MYC

Abstract Blockade of cell cycle re‐entry in quiescent cancer cells is a strategy to prevent cancer progression and recurrence. We investigated the action and mode of action of CPF mixture (Coptis chinensis, Pinellia ternata and Fructus trichosanthis) in impeding a proliferative switch in quiescent lung cancer cells. The results indicated that CPF impeded cell cycle re‐entry in quiescent lung cancer cells by reduction of FACT and c‐MYC mRNA and protein levels, with concomitant decrease in H3K4 tri‐methylation and RNA polymerase II occupancy at FACT and c‐MYC promoter regions. Animals implanted with quiescent cancer cells that had been exposed to CPF had reduced tumour volume/weight. Thus, CPF suppresses proliferative switching through transcriptional suppression of FACT and the c‐MYC, providing a new insight into therapeutic target and intervention method in impeding cancer recurrence.


| INTRODUC TI ON
With improvements in treatment, a growing list of cancers has become chronic diseases. In 2016, there were 14 million cancer survivors in the United States alone but many of them could later die from cancer recurrence. According to GLOBOCAN 2018, global cancer deaths could reach 9.6 million in 2018. 1 Hence, prevention of cancer recurrence is a major concern facing the growing population of cancer survivors, given the reality that no secondary recurrence prevention method is available after completion of primary treatment.
G 0 phase is a cellular state in which cells are in reversible cell cycle arrest. To multicellular organisms, the G 0 state is necessary for development as cells at different stages of development lineage, including stem cells, slow-cycling cells and differentiated mature cells, all adopt the G 0 state temporally or spatially. 2,3 Cancer cells can also share the G 0 state and are also known as dormant or quiescent cancer cells. 4 Together with insufficient angiogenesis and immune surveillance, quiescent cancer cells are one mechanism explaining clinically observed cancer dormancy. 4 The presence of dormant cancer cells is a two-edge sword. While it is beneficial to a multicellular organism for cancer cells with uncontrolled proliferation to be halted, the same cancer cells in quiescence can survive most chemoand radiotherapy and may, with time, acquire additional mutations and gain metastatic potential as they re-enter the cell cycle. 5 The therapeutic strategy of eliminating G 0 cancer cells or promoting cell cycle re-entry by quiescent cancer cells is regarded as risky, as treatment efficacy in the G 0 phase is uncertain, and surviving cells may be selected out for more aggressive characteristics. Also, there is no assurance that antiproliferative drugs will be sufficiently effective to eliminate the reactivated G 0 cancer cells. 5 Considering that proliferation of G 0 cancer cells upon cell cycle re-entry can simply replace the eradicated cancer cells during primary treatment, 6,7 the development of a therapeutic strategy aimed at impeding cell cycle re-entry by quiescent cancer cells is of importance. A major prerequisite to be able to implement this strategy is to identify the therapeutic target responsible for the proliferative switch and to establish an effective and safe method capable of exerting action on the therapeutic target.
Our knowledge of the mechanisms underlying cell cycle re-entry from a G 0 state is growing. Apart from cyclin-dependent kinase inhibition, 8 histone modification, RNA interference and autophagy are all implicated in the proliferative switch from quiescent state. 3,[9][10][11][12][13] Facilitates Chromatin Transcription (FACT) is a member of histone chaperone family and consists of subunits of structure-specific recognition protein 1 (SSRP1) and suppressor of Ty homolog-16 (SPT16). In the nucleosome, FACT facilitates the passage of DNA and RNA polymerase by temporal eviction of histones. 14 FACT then promotes the deposition of histones to re-establish the nucleosome. 14 However, against intuition, genetic silencing of the histone chaperones affects only about 2% of gene transcription and elongation in lung cancer cells. 15 Also, FACT is aberrantly overexpressed in cancers of breast, lung, pancreas and brain. 16,17 Recently, we have shown that FACT mRNA and protein levels oscillate between the quiescent and proliferative state, and FACT is necessary and sufficient in the proliferative switch of quiescent lung cancer cells through transcriptional regulation of c-MYC, which in turn influence cyclin-dependent kinase inhibitor p27 and its regulatory proteins such as SKP2. 18 Ideally, in cancer patients who have completed intensive treatment, sometimes with considerable toxicity, any subsequent treatment should be not only effective but possess few side effects, especially if quiescent cancer cells are to be maintained in the G 0 phase for a considerable period. Based on the principles of effective therapy, minimal toxicity and economical use, as well as the notion of 'an old drug for anew use', we have investigated Chinese herbal medicines for preventing cell cycle re-entry of quiescent lung cancer cells. The ancient book 'Treatise on Miscellaneous Diseases' written by Zhang Zhongjing (AD 150-219) is the first known complete collection of Chinese medicine prescriptions. 19 Known as the 'ancestral book' in China, the prescriptions are respected as 'Kampo' (Han dynasty prescriptions) in Northeast Asia and are still used in universities of Chinese medicine. CPF (Chinese name: Xiao-Xian-Xiong Tang) was first described in the 'Treatise on Miscellaneous Diseases' and consists of three herbs: Coptis chinensis, Pinellia ternata and Fructus trichosanthis ( Figure S1A). Based on the HPLC results, of the 63 compounds identified, 43 were from Coptis chinensis (Figure S1B,C and Table S1). According to the theory of traditional Chinese medicine, CPF is used for treating diseases of the respiratory system, including pneumonia, asthma and pulmonary fibrosis in China. Here, we present evidence that CPF is able to impede the proliferative switch of quiescent lung cancer cells by transcriptional suppression of FACT and c-MYC genes.  c-MYC, G 0 cell cycle re-entry, lung cancer, structure-specific recognition protein 1, suppressor   of Ty homolog-16 ATCC and grown in RPMI 1640 supplemented with 10% v/v foetal bovine serum (AusGeneX), penicillin (100 U/mL) and streptomycin (100 μg/mL). The cells were cultured at 37°C with 5% CO 2 /95% air.

| CPF preparation
CPF consists of Coptis chinensis, Pinellia ternata and Fructus trichosanthis. All the herbs obtained from Huayu Pharmacy Company were certificated based on the authentication by Chinese Pharmacopoeia, heavy metal and pesticide residue standards. In a weight ratio of 3:6:10, CPF was boiled in six volumes of pure water for 60 minutes twice. The combined supernatants after filtration were vacuum-dried at 60°C and reconstituted in DMSO (500 mg/mL) as stock and kept in 4°C.

| Retroviral transduction and plasmid transfection
The mVenus-p27K −20 was provided by Dr Toshihiko Oki (The University of Tokyo, Japan). It was mixed with packaging plasmids as described 18 and transfected into 50% confluent HEK293T cells with a calcium phosphate precipitation method. The resulting lentiviral particles were used to infect A549 cells with 8 µg/mL polybrene. The infected cells with mVenus-p27Kconstruct were selected by 0.2 mg/mL puromycin. The SPT16 (OHu22815D) and SSRP1 (OHu17195D) plasmid were obtained from GenScript Company and transfected into cells with Lipofectamine™ RNAiMAX (Invitrogen, Life Technologies).

| SYBR green assay
Quiescent A549 (10 000 cells/well) and H1975 (7000 cells/well) cells were seeded in 96-well plates. The same number of cells/well was kept as a baseline and stored at −80°C. After treatment, the medium was gently aspirated and replaced with 100 µL of lysis buffer as described. 18 The cells were then lysed in the dark for 2 hours with shak-

| Flow cytometry assay
To distinguish G 0 and G 1 cells, the cells incubated with Hoechst 33258 and Pyronin Y 21 were injected into Gallios flow cytometer (Beckman Coulter). To determine mVenus-p27Kfluorescence, the experimental cells were fixed with paraformaldehyde for 10 minutes at room temperature and washed with PBS prior to data acquisition using a flow cytometer (FACSCalibur II) equipped with CellQuest Pro software (BD Biosciences). Flow cytometry data were analysed using FlowJo software (version 10.0.8).

| Quantitative reverse transcription-PCR
Real-time PCR was performed as described previously. 21 The cDNA generated from total RNA (iScript™ cDNA Synthesis Kit, Bio-Rad) was mixed with SYBR Green containing SensiMix™ Master Mix (Bioline). The reaction was subjected to a thermocycle at 95°C for 10 minutes followed by 45 cycles at 95°C for 15 seconds, at 60°C for 15 seconds and at 72°C for 15 seconds. The primer sequences are described in the supplementary information (Table S2).

| Immunoblotting
A549 and H1975 cells were treated in 6-well plates, and cell lysates were prepared as described ref. 19 . Primary antibodies against SPT16 GAPDH antibody (Cat. #: ab8245) was obtained from Abcam.

| Chromatin immunoprecipitation (ChIP)
ChIP analysis was conducted with immunoprecipitation assay kit (Cat #: 17-295, Merck) together with ChIP-grade rabbit anti-H3K4me3 (Cat #: ab8580, Abcam) and anti-RNA polymerase II antibodies (Cat #: 664911, BioLegend) and isotype rabbit IgG (Cat #: sc-2027, Santa Cruz Biotechnology). Immunoprecipitated DNA was quantified by quantitative real-time PCR. See primer sequences in Table S3. ChIP assays were repeated thrice and calculated as fold enrichment relative to the control IgG and normalized by input DNA.

| Immunocytochemistry
Cells detached from culture flasks were fixed in 10% buffered formalin solution, dehydrated and embedded in paraffin blocks. Selected samples were sectioned (5 mm thick) and stained with haematoxylin and eosin, Ki-67 (Abcam, ab92353), as described previously. 22 The primary antibodies were used at 1:500 for Ki-67. The sections were finally mounted with DPX Mountant for histology analysis.

| Imaging study
Nikon Ti-E spinning disc confocal live cell microscope with a 20 × objective lens was used for time-lapse imaging analysis of cells cultured in a 0.17-mm glass-bottom 6-well plate (MatTek). An image was acquired every 15 minutes for 36 hours. Image acquisition and video conversion were performed with NIS-Elements (version 4.5).

| Animals study
Male BALB/c nude mice (6 weeks old) and male ICR mice (4 weeks old) were maintained under specific pathogen-free conditions with constant temperature (23 ± 2°C) and controlled light (12-hour light: 12-hour dark). To examine the effect of CPF on tumour growth, quiescent A549 cells were induced to re-enter the cell cycle by plating at a low density and treated either with CPF at GI90 or with DMSO for 6 hours. The cell viability in both groups was evaluated by trypan blue exclusion, and they were >95% in both groups. Together with proliferative A549 cells, the pre-treated cells, 1 × 10 7 cells in 0.2 mL PBS, were subcutaneously injected into the left flank of BALB/c nude mice. The mice were anaesthetized for determining the tumour volume and weight 16 days after cancer cell injection.

| Statistical analysis
The statistical software SPSS (version 18.0) was used for analysis.
One-way ANOVA was used to determine the difference between individual groups of data. Multiple comparison test was used to determine whether the difference between individual groups (P < .05) was significant.

| CPF suppresses proliferative switch from G 0 state in lung cancer cells
As described previously, 18 Figure 1C). Since cells at G 0 and G 1 are both diploid, we used a double staining method to quantify the G 0 fraction. 21 The reduced G 0 fraction seen in controls as early as 12 hours following release from quiescence was prevented to a large extent by CPF ( Figure 1C). In control cells, a complete return of cell cycle distribution to proliferative state was achieved at 36 hours following release from quiescence (all P > .05 compared with No CI and FCS).
However, treatment with CPF caused delay in the re-entry in both cell lines, as the distribution of all cell cycle phases remained different significantly from control at 36 hours ( Figure 1C). No statistically significant change in cell viability was observed on treatment with CPF at GI90 as cell cycle analysis showed <2% cell population in the sub-G 1 fraction.
To further validate the CPF action, we transduced A549 cells with mVenus-p27K − plasmid. This mutant p27 cannot bind cyclin-dependent kinase but maintains an intact domain for ubiquitination.
CPF was introduced when mVenus-p27K − cells were released from quiescence following 3-day contact inhibition. There was a significant increase in fluorescent signal over the course of contact inhibition compared to the proliferating cells (Figure 2A,B). The signal was reduced upon cell cycle re-entry. However, treatment with CPF at GI50 and GI90 led to a clear retention of mVenus-p27K − signal compared with vehicle-treated control ( Figure 2C and Video). Taken together, these data underscore that CPF can impede cell cycle re-entry of quiescent lung cancer cells.

| CPF treatment decreases FACT mRNA and protein levels and transcriptional activity
We have shown recently that FACT is required in cell cycle re-entry by quiescent lung cancer cells. 18 To establish the mechanism by which CPF impedes cell cycle re-entry, we examined the impact of CPF on mRNA and protein levels of FACT in cell cycle re-entry. Treatment with CPF efficiently reduced the surged mRNA ( Figure 3A) and protein ( Figure 3B, Figure S2A) levels of FACT subunit SSRP1 and SPT16 over the time period of 6-24 hours compared with vehicle control cells. Moreover, consistent with our previous finding that FACT promotes cell cycle re-entry via p27 and its regulatory proteins including c-MYC and SKP2, 18 the CPF treatment reduced c-MYC and changed SKP2 and p27 over the same period ( Figure 3C, Figure S2B).
To verify that the CPF-led inhibition of cell cycle re-entry is mediated by reduction of FACT mRNA and protein levels, we transfected quiescent A549 cell line with an expression vector containing SSRP1 or SPT16 for 4 hours after release from quiescence. At 36 hours, the ectopic expression of either FACT subunit significantly reduced the G 0 fraction compared to control. While CPF alone increased the cells at G 0 fraction compared to control, the effect of CPF on G 0 fraction was diminished when FACT subunits were overexpressed simultaneously ( Figure 4A, Figure S2C). Hence, CPF action on cell cycle re-entry could be mediated by its impact on FACT. We then used ChIP assays to determine the effect of CPF on H3K4 tri-methylation and RNA polymerase II occupancy at FACT gene promoters in A549 cell line. There was a significant reduction of H3K4 tri-methylation and RNA polymerase II recruitment at SSRP1 and SPT16 gene promoter regions in the presence of CPF compared with vehicle control ( Figure 4B).
Considering that the CPF treatment reduced c-MYC almost at same time as FACT, we used ChIP assays to determine the occupancy of H3K4 tri-methylation and RNA polymerase II at c-MYC gene promoter in the presence of CPF. Indeed, compared with vehicle control, CPF treatment led to a significant reduction of transcriptional activity at c-MYC promoter ( Figure 4B) and mRNA levels ( Figure S2D). While it is clear that FACT can regulate c-MYC in lung cancer cells, 18

| D ISCUSS I ON
The histone chaperone FACT contributes to the maintenance of a flexible chromatin landscape. By removing and presenting histones to naked DNA, FACT affects selectively nucleosome disassembly and assembly thereby regulating transcription and elongation of a small fraction of genes. 15 Hence, a therapy targeting FACT can influence neither globally nor on FACT only. This multitarget approach, in theory, should be beneficial considering the ability of cancer cells to acquire resistance to single-target therapies. FACT is highly expressed in poorly differentiated tumours with unfavourable outcome. 15 Reducing expression of either FACT subunit leads to inhibition of proliferation and tumour cell death, probably due to the mechanism that FACT selectively promotes transcription of genes that stimulate proliferation and prohibit cell death and differentiation. 16,17 We have shown recently that FACT is necessary and sufficient in the switch from G 0 to the proliferative state in lung cancer cells. 18 FACT protein levels are low at G 0 compared to the proliferating state but quickly surge upon cell cycle re-entry.
Knockdown of FACT hindered cell cycle re-entry of quiescent cells, Pinellia ternata or Fructus trichosanthis. We will screen the identified 43 compounds from Coptis chinensis, and, if they are not the major active ingredients, we will use HPLC to obtain the fraction of Coptis chinensis and test each fraction in our platform of cell cycle re-entry. The effective fraction will be used for isolation of the active compound, which will then be validated by comparing its action and mode of action with CPF and Coptis chinensis.
The presented work also reflects our effort to use modern research tools to develop a system to scientifically determine the efficacy of ancient Chinese medicine recipes. In 2015, the Chinese scientist Youyou Tu was awarded the Nobel Prize for the development of an antimalarial drug extracted from Artemisia annua L. 29 Realgar-Indigo naturalis receipt and its ingredients have been proven to be effective in treating human acute promyelocytic leukaemia. 30 Although these are evidences of the presence of effective compounds in traditional Chinese medicines, for most Chinese medicine receipts the exact action and mode of action are not well defined. Since a great population is using traditional medicine, 31 it is necessary to evaluate and validate the biomedical potential of Chinese medicine so that evidence can be provided for each recipe for its disease indication, molecular target and active ingredients.

CO N FLI C T O F I NTE R E S T
The authors declare no competing interests.

AUTH O R CO NTR I B UTI O N S
LB, CX, LJ, SJ, SH, MY, YW, QW, GG, YW, XS and YK conducted experiments, analysed data and wrote the manuscript. XZ, PD, TL and F I G U R E 5 Quiescent lung cancer cells exposed to CPF in vitro reduced growth in vivo. Quiescent A549 cells were induced to re-enter the cell cycle by plating at a low density and treated either with CPF at GI90 or with DMSO for 6 h. Together with proliferative A549 cells, the pre-treated cells were subcutaneously injected into the left flank of 6-week-old male BALB/c nude mice (1 × 10 7 cells per mouse). Tumour volume (A) and weight (B) on day 16 of experimentation were shown. *P < .05 compared with the proliferative group; #P < .05 compared with the DMSO pre-treated control. C, The response of normal human bronchial epithelial cells to CPF for indicated dose and duration was analysed by CCK-8 reagent LX and QD designed the study.

E TH I C S A PPROVA L A N D CO N S E NT TO PA RTI CI PATE
The animal study was approved in Sino-British SIPPR/BK Lab Animal Ltd (animal authorization reference number: SCXK2013-0016) and performed in accordance with the Declaration of Helsinki.

DATA AVA I L A B I L I T Y S TAT E M E N T
The original data of this study are available from corresponding author upon request.