Angelica Sinensis Polysaccharides Ameliorate 5-Flourouracil-Induced Bone Marrow Stromal Cell Proliferation Inhibition Via Regulating Wnt/ β-Catenin Signaling

Background: Chemotherapy-induced bone marrow hematopoietic microenvironment oxidative damage is closely related to myelosuppression. Angelica Sinensis Polysaccharides (ASP) are major effective ingredients of traditional Chinese medicine Angelica with multi-target anti-oxidative stress features. In this study, we investigated the potential benets and mechanism of action of ASP on chemotherapy-induced bone marrow stromal cell (BMSC) damage. Methods: The human bone marrow stromal cell line HS-5 cells were divided into control group, 5-FU group, 5-FU + ASP group, and 5-FU+ LiCl group to investigate the mechanism of ASP to alleviate 5-FU-induced BMSC proliferation inhibition. Results: The results showed that 5-FU inhibits the growth of HS-5 cells in a time and dose-dependent manner; however, ASP partially counteracted 5-FU-induced decrease in cell viability, whereas Wnt signaling inhibitor Dkk1 antagonized the effect of ASP on HS-5 cells. ASP reversed the decrease in cytoplasmic total β-catenin, p-GSK-3β, and CyclinD1 following 5-FU treatment, and modulated nuclear expression of β-catenin, Lef-1, and C-myc proteins. Furthermore, ASP also enhanced the antioxidant capacity of cells and reduced 5-FU caused oxidative stress, attenuated FoxO1 expression, thus weakened its downstream apoptosis-related proteins and G 0 /G 1 checkpoint-associated p27 Kip1 expression to alleviate 5-FU-induced apoptosis promote cell cycle progression. Conclusion: The protective role of ASP in BMSCs proliferation for the chemotherapy may be related to its activating Wnt/β-catenin signaling and keeping homeostasis between β-catenin and FoxO1 under oxidative stress. The study provides a potential therapeutic strategy for alleviating chemotherapeutic damage on BMSCs.

The HS-5 cells were seeded in 96-well plates at a density of 5 × 10 3 cells per well and treated as described in groups of experiment. After 48 h treatment, cells were exposed to 10 μmol/L EdU solution for 24 h. Cells were washed and xed in 4% paraformaldehyde at room temperature for 30 min. After washing, cells were permeabilized in PBS containing 0.5% Triton X-100 for 20 min. Then, cells were washed and incubated with 1X Apollo ® reaction cocktail for 30 min. Subsequently, cells were stained with Hoechst33342 for 30 min and observed under a uorescence microscope (Olympus, Japan). Counting 200 cells at random, the proliferation rate of HS-5 cells was de ned as the ratio of EdU-positive cells (green cells) to Hoechst33342-positive cells (blue cells).

Flow Cytometry Analysis
For cell apoptosis assay, the HS-5 cells were cultured then treated as described in groups of experiment. After 48h treatment, cells were harvested and centrifuged at 1000r/min for 5min.Subsequently, cells were resuspended with 500µ L PBS solution for each tube. Cell apoptosis was detected by the ow cytometry. For cell cycle assay, the HS-5 cells were cultured then treated as described in groups of experiment. After 48h treatment, cells were harvested and xed with pre-cooled 75% ethanol at 4°C for at least 5h. After centrifugation, cells were incubated with propidium iodide (PI) and RNase A at 37°C for 30 min in dark. Cell cycle was detected by the ow cytometry. The apoptosis and cell cycle were analyzed on a FAC-Scan laser ow cytometry (BD Biosciences, New Jersey, the USA). The data were processed by Cell Quest software (BD Biosciences, New Jersey, the USA).

Immuno uorescence Staining
Sterile glass slides were put into 24-well plates, the HS-5 cells were cultured at a density of 5× 10 4 cells per well in 24-well plates then treated as described in groups of experiment. After 48h treatment, cells were xed with 4% paraformaldehyde for 30 min at room temperature. After washing, cells were permeabilized in PBS containing 0.5% Triton X-100 for 20 min then blocked with 10% goat serum for 1h. Subsequently, cells were incubated with monoclonal antibody β-catenin (1:150) overnight at 4°C. After being washed thrice with PBS solution, cells were incubated with Cy3-labeled goat-anti-rabbit immuno uorescent secondary antibody (1:300) at 37°C for 2h in the dark. The nuclei were stained with 4′,6-diamidino-2-phenylindole (DIPI)for the last 5 min. The images were observed and acquired under the uorescence microscope.

Immunoblot Assay
The HS-5 cells were treated as described in groups of experiment. After 48h treatment, cells were incubated with PIPA lysis buffer containing 1% protease inhibitor and phosphatase inhibitor for 30 min on ice and proteins were isolated after centrifugation. The concentrations of proteins were detected by the BCA Protein Assay Kit (Beyotime, China). The protein samples were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to polyvinylidene uoride (PVDF) membranes (Millipore, the USA). The membranes were blocked with 5% skim milk for 1h at room temperature and subsequently incubated overnight at 4°C with β-catenin, Cyclin D1, p-GSK-3β, GSK-3β, Lef-1, C-myc, FoxO1, p-FoxO1, p27 Kip1 , Bim, Bcl-2, Bax and caspase-3 primary antibodies (1:1000). After washing three times with Tris-Buffered Saline and Tween-20(TBST), the membranes were incubated with secondary antibodies for 1h at room temperature. The enhanced chemiluminescence (ECL)kit (Millipore, the USA) was used for color development and Image Lab 5.2.1 software was used for semi-quantitative analysis. The relative expression levels of the target proteins were determined by the ratio of the target protein gray value to internal reference protein gray value.

Oxidation-Associated Biological Indicators Assay
For the detection of intracellular ROS, the HS-5 cells were seeded in 6-well plates at a density of 2 × 10 5 cells per well then treated as described in groups of experiment. After 48h treatment, the cells were washed thrice by serum-free medium and then incubated with 2',7'-dichlorodihydro uorescein diacetate (DCFH-DA) at 37°C for 20-30 min in the dark. The content of intracellular ROS was observed and acquired under the uorescence microscope. The average optical density per unit area was analyzed using ImageJ software. For the detection of MDA content and SOD, CAT activity, the HS-5 cells were cultured and treated as described in groups of experiment. After 48h treatment, cells were harvested, lysed and centrifuged to collected the supernatant. MDA, SOD, and CAT were measured by the corresponding assay kits according to the manufacturer's instruction.

Statistical analysis
For all assays, the experiments were performed at least three times. All the results were analyzed by Oneway analysis of variance (ANOVA) with SPSS 20. 0 statistical software. All the data were expressed as mean ± standard deviation (SD). P < 0.05 was considered statistical signi cance.

5-FU Inhibits the Growth of HS-5 Cells by down-regulating Wnt/β-catenin Signaling Pathway
To assess the effect of 5-FU on proliferation, HS-5 cells were treated with 5-FU at different concentrations for 72 hours. As shown in Fig.1A, with the increase of 5-FU concentration and the extension of treatment time, the cell growth was signi cantly inhibited, suggesting that 5-FU has a dose-dependent and timedependent inhibiting effect on HS-5 cells proliferation. HS-5 cells were incubated with 5-FU at concentration of 25μg/mL for 48h, half of the cells were suppressed. Thus, 25μg/mL 5-FU treated cells for 48h was selected for the subsequent experiments. Interestingly, in the current study we found the relationship between 5-FU inhibition of HS-5 cell proliferation and Wnt/β-catenin signaling pathway. Western blot assay demonstrated that 5-FU downregulated the cytoplasmic levels of p-GSK-3β, total βcatenin and Cyclin D1 in HS-5 cells, followed by nuclear protein expression of β-catenin, Lef-1, and C-myc downregulation (Fig.1B). These results hinted that 5-FU promoted the ubiquitination degradation of βcatenin mediated by GSK-3β-complex, inhibits the nuclear translocation of β-catenin and downregulate the downstream target genes. We hypothesize that the growth inhibition of HS-5 cells caused by 5-FU may be related to the down-regulation of Wnt/β-catenin signaling.
Hence, the activator and antagonist of Wnt/β-catenin signaling pathway were used to illustrate the question above. In vitro culture, treated with different concentration of LiCl, which could activate the Wnt/ β-catenin signaling pathway, under light microscope, it was shown that the cells in 5, 10, 20mmol/L LiCl groups increased the number, however, the cellularity in 40 mmol/L group signi cantly dropped concomitant with smaller and loosely dispersed shape (Fig.1C). Also, cell viability was tested by CCK-8 assay. As shown in Fig.1D, cells treated with 5 mmol/L ~ 20 mmol/L LiCl for 48h showed different degrees of proliferation, among which the proliferation rate of the cells peaked in 10 mmol/L LiCl group and was increased to 150% compared with the control group. However, LiCl at a high concentration of 40 mmol/L was presented cytotoxic to HS-5 cells even within 24h. Therefore, the pretreatment with 10mmol/L LiCl was utilized for the subsequent experiment as a positive control. Furthermore, Dkk1, an antagonist for Wnt/β-catenin signaling, was used to get more evidence for Wnt/β-catenin signaling pathway on the cell viability. 50 ng/mL Dkk1 treated for 48h, the cells were dramatically inhibited compared with the control group revealed by the results of CCK-8 (Fig.1E). All the results above indicate that the effect of 5-FU on inhibition of HS-5 cell growth correlates with the suppression of Wnt signaling pathway.

Angelica Sinensis Polysaccharides Antagonize Growth Inhibition of 5-FU-treated HS-5 Cells through Wnt/ β-catenin Signaling Pathway
EdU is a thymidine analog that replaces thymidine in the DNA being synthesized during DNA replication. The conjugation reaction of EdU with uorescent dyes can detect the cell proliferation. As shown in Fig.2A, B, the proportion of proliferating cells in 5-FU group was signi cantly lower than that of the control group; after pretreatment with ASP and LiCl, the proportion of EdU proliferating cells increased markedly compared with 5-FU group. The results of CCK-8 showed an obvious reduction after a 48h incubation with 5-FU compared with untreated control cells; however, ASP pretreatment partially reversed with ASP, the reduction of cellar viability induced by 5-FU. Moreover, ASP-induced increase in the viability was weakened by Dkk1 (Fig. 2C). The results above suggested that ASP may mitigate the proliferation inhibition of HS-5 cells after 5-FU treatment via regulating Wnt signaling. β-catenin protein is a key effector protein of Wnt/β-catenin signaling pathway, as shown in Fig.3A by immuno uorescence assay, the cytoplasmic and nuclear expression of β-catenin was decreased obviously after 5-FU treatment, however, ASP and LiCl pretreatment respectively rescued the expression of β-catenin and its nuclear translocation. Also, the results of western blot revealed that ASP and LiCl pretreatment signi cantly reversed the 5-FU-induced decrease in cytoplasmic expression of total β-catenin, p-GSK-3β, and CyclinD1, meanwhile modulated nuclear expression of β-catenin, Lef-1, and C-myc proteins ( Fig.3B-I). These data suggested that ASP may activate Wnt signaling, which may be one mechanism that ASP counteract the inhibiting effect of 5-FU on HS-5 cell growth.

Angelica Sinensis Polysaccharides Relieve 5-FU-induced Intracellular Oxidative Stress
To elucidate the mechanism of 5-FU induced damage and ASP mediated protective effect on HS-5 cell growth, we assessed the indexes of oxidative damage. Increases of intracellular ROS and MDA were found in 5-FU group compared with the control group whereas single pretreatment of ASP or LiCl reversed the increase dramatically (Fig.4A, C). On the contrary, ASP or LiCl administration protected the antioxidant enzymes including SOD and CAT in HS-5 cells (Fig.4D, E). These results demonstrate that 5-FU cause oxidative stress to HS-5 cells, whereas ASP exert a signi cant anti-oxidative role to alleviate 5-FU-induced oxidative stress, which may be related to the activation of Wnt/β-catenin signaling.
Angelica Sinensis Polysaccharides Ameliorate the Activation of FoxO1 Induced by 5-FU FoxOs are transcriptional factors closely related to cellular survival and oxidative stress, Notably, activated FoxO1 may impair Wnt signaling via competitive combination with β-catenin in nucleus. To address the question that whether 5-FU-induced intracellular oxidative stress and mitigative effects of ASP is related to FoxO1 transcription, we analyzed the levels of FoxO1 protein and its phosphorylation. Western blot results showed that compared with the control group FoxO1 expression in 5-FU group rose concurrent with decreased p-FoxO1 expression. However, ASP or LiCl pretreatment signi cantly reduced of FoxO1 expression via degradation of FoxO1 by an increased p-FoxO1 expression (Fig.5A). It was inferred that FoxO1 may be involved in response to 5-FU induced intracellular oxidative stress. Antioxidative properties of ASP may play a role in FoxO1 downregulation, which may be another possible mechanism for ASP upregulation of Wnt/β-catenin signaling. Like p53, FoxOs orchestrate apoptosis through multiple mechanisms, also FoxOs may promote cell cycle arrest by upregulating multiple cell cycle suppressor genes such as CDKI p27 Kip1 . Here, in this context, the proteins correlating to apoptosis and cell cycle modulators including Bcl-2, Bim, Bax, caspase-3 and p27 Kip1 were detected via western blot assay. The results showed that ASP or LiCl counteracted apoptosis and cell cycle arrest by abrogating 5-FU-induced increase in Bim, Bax caspase-3 and p27 Kip1 expression, also by enhancing anti-apoptotic protein Bcl-2 expression (Fig.5B-F, H). The data of owcytometric analysis revealed that compared with the control group there was a 2.2-fold up-regulation of the apoptosis rate in 5-FU group, ASP or LiCl signi cantly decreased the percentage of apoptotic cells compared with the 5-FU group (Fig.5G). Meanwhile, ASP or LiCl pretreatment weakened 5-FU induced G0/G1 phase retard concurrent with S and G2/M phase recovery (Fig.5I). All these data above hinted that the effects of 5-FU on HS-5 growth inhibition may relate to the activation of FoxO1 leading to apoptosis or cycle arrest. Anti-oxidative property of ASP exerts a protective effect against cycle arrest and apoptosis.

Discussion
Myelosuppression is one of the common side effects of chemotherapy, characterized by depletion of cells within the bone marrow [40,41]. In general, myelosuppression is primarily attributable to the direct cytotoxicity to bone marrow cells, inhibition of bone marrow precursor or progenitor cell proliferation, the reduction in HSC reserves and impairment in HSC self-renewal. Notably, because of the reduction of HM cellularity in varying degrees, the damaged-hematopoietic microenvironment may result in diminished or delayed hematopoiesis function, immune-related disorders, as well as long-term damage to the bone marrow recovery [42,43]. It has been shown that chemotherapeutic treatment damage the hematopoietic microenvironment in vitro and vivo [5,44,6,7,9,10]. As chemotherapy disrupts the steady-state function of hematopoietic and stromal cell, disruptions over time may cause severe bone marrow toxicity and the failure of cancer treatment. To ensure this does not occur, nding appropriate agents to promote the recovery process following discontinuation of chemotherapy and to lessen the bone marrow damage has a profound signi cance.
Since it was rst synthesized in 1957, 5-FU has remained one of the most widely used chemotherapeutic agents with broad-spectrum activity against many solid tumors [45]. 5-FU exerts its anticancer effects through inhibition of thymidylate synthase (TS) and incorporation of its metabolites into RNA and DNA, leading to cytotoxicity and cell death [12]. Recent studies have indicated that 5-FU suppressed the proliferation of HSCs and induced the myelosuppression of mice by down-regulating PI3K-AKT signaling pathway [14,46]. However, the de nite mechanism for 5-FU caused myelosuppression remains unclear. Focused on bone marrow stromal cells, we provided the evidence that 5-FU inhibited stromal cell growth and induced apoptosis, which was related to downregulation of Wnt/β-catenin signaling, also upregulation of FoxO1 concomitant with an increase of cellular oxidative stress. Furthermore, the current work revealed that anti-oxidative property and role in Wnt signaling regulation might be the key mechanisms of ASP to prevent against 5-FU-induced stromal damage.
Stem cells display the de ning capacity to self-renew, and their fate are primarily dictated by extrinsic, short-range signals, which typically emanated from the stem cell niche [47]. The non-hematopoietic cells in hematopoietic microenvironment have a functional role in regulating hematopoiesis and the signaling pathways that regulate HM may be necessary for the development of functional niches that regulate hematopoietic stem cells and their progenitors [48,49]. The Wnt signaling pathway exerts a variety of effects on target cell developmental processes, including cell proliferation, apoptosis, and differentiation. The canonical Wnt pathway affects cellular functions by accumulating of β-catenin in the cytoplasm and eventually translocating into the nucleus. Within the nucleus, β-catenin binds to T cell factor (TCF) family/lymphoid enhancer factor (LEF) and regulates cell proliferation through Wnt downstream target genes [16,50,51]. It was reported that Wnt/β-catenin signaling regulates HSCs function in dosagedependent manner [52][53][54][55]. Various degrees of activation of the pathway may cause different outcomes, leading to either enhanced repopulation capacity or exhaustion of the HSCs. A mild increase in Wnt signaling enhanced HSC function [18,56]. However, a high Wnt level in HSCs eventually leads to stem cell exhaustion and impairment of reconstitution in irradiated recipients [57][58][59]. Most importantly, Wnt signaling regulates HSC reconstruction in a stromal-dependent manner. It was found that when hematopoietic cells were co-cultured with BMSCs supplemented with Wnt3a conditioned medium, the cellularity of Lin − Sca-1 + c-kit + hematopoietic stem cells were increased, and the hematopoietic transplantation and reconstruction capability were enhanced [60,20]. Hence, in the current study we focused on the Wnt signaling regulation on BMSCs following chemotherapy. It was found that 5-FU induced a decrease in cytoplasmic expression of total β-catenin, p-GSK-3β, and CyclinD1, meanwhile weakened nuclear expression of β-catenin, LEF-1, and C-myc proteins, causing HS-5 cells proliferation inhibition. The results herein are in line with the other data related to the relationship between canonical Wnt signaling and cell proliferation, which has con rmed that Wnt/β-catenin signaling positively stimulates cell growth via cell cycle regulation [61,62].
Reactive oxygen species (ROS) are free radicals and active metabolites of oxygen containing unpaired electrons, which take a signi cant role in cell signal transduction and regulation [63]. Chemical agents, as well as irradiation can cause a persistent ROS production. This accumulation of ROS may lead to excessive oxidative stress and DNA damage such as DSBs (double-strand breaks), which is considered to be the main potential mechanisms causing cellular damage [64,65]. Previous study in our group has demonstrated that 5-FU weakened the antioxidant capacity of HS-5 cells and caused high sensitivity of cells to ROS, thus HS-5 cells underwent DSB which eventually resulted in either apoptosis or senescence [15]. Oxidative stress is also related to cell cycle arrest. DSBs initiate DNA damage response through sequential stimulation of ATM, Chk2, and p53 [66]. Activation of p53 and its downstream p21 may induce the cell cycle arrest. Meanwhile ROS can activate p38 MAPK pathway [67]. Activation of p53 and p38 pathways can converge at p16 and augment of p16 expression may also lead to permanent cell cycle arrest [68,69]. Interestingly, it is reported that β-catenin may be critical for antagonizing oxidative stress. Exposing β-catenin knock-down mice to chemotherapeutic agent or radiation caused a decreased expression of the hydrogen peroxide (H 2 O 2 ) detoxifying enzyme catalase and led to the accumulation of ROS and superoxide (O2¯) free radicals in cells and an inability to repair DNA damage [70]. On the opposite, effector molecules generated from oxidative DNA damage may also down-regulate the Wnt pathway by inhibiting transcriptional activity or participating in post-translational modi cations to enhance ubiquitination degradation [71]. These evidences above hint that Wnt signaling also closely corelated with oxidative stress. Therefore, in the current study, increased oxidative stress may be one of the reasons for down-regulation of wnt signaling induced by 5-FU treatment. Whereas decrease in βcatenin protein accompanying reduction of antioxidase SOD and CAT induced by 5-FU treatment may be another mechanism of cell proliferation inhibition.
Forkhead box O (FOXO) family are transcription factors, which promote cell survival by regulating the cell cycle, apoptosis and the response to oxidative stress [23]. The accumulation of ROS may interrupt 14-3-3 combine to FoxO via JNK (c-Jun N terminal kinase), permit FoxO entrance into nucleus, and induce its transcriptional activation [72,73]. FoxO can be phosphorylated by phosphatidylinositol 3-kinase-Akt pathway [74,75]. It is of note that FoxO-mediated transcription requires binding of β-catenin. FoxOs can compete with TCF/LEF by directly binding β-catenin, thereby inhibit Wnt/β-catenin downstream signaling [76,77,24,78]. It was demonstrated herein, compared with the control group FoxO1 expression in 5-FU treated HS-5 cells rose dramatically concurrent with decreased p-FoxO1 expression. The reason for up-regulation of FoxO1 may be related to 5-FU triggered oxidative stress, whereas FoxO1 up-regulation may be another reason for 5-FU induced decrease in Wnt signaling [27,26,79]. FoxO transcription factor family regulate the proteins that are crucial for the apoptosis, as well as the proteins involved in proliferative status of a cell. FoxO factors may regulate antiapoptotic and proapoptotic proteins at multiple levels, nally trigger activation of the effector caspases. Bim promotes apoptosis by inhibition of antiapoptotic Bcl-2 family members or through direct activation of Bax-like molecules. FoxO factors may regulate Bim protein expression to cause cell death due to cytokine deprivation. FoxO factors may also repress transcription of Bcl-XL through the induction of the transcriptional repressor [80][81][82]. Caspase-3 is an important effector protease, when it is cleaved, it acts as the nal executor during apoptosis. In the current study, it was found that in 5-FU treated HS-5 cells, FoxO1 targeted apoptosis-related proteins to cause increase in Bim, Bax, and caspase-3 whereas decrease in Bcl-2. FoxO1 targeted apoptosis to disturb the dynamic balance of the cellularity of HS-5 cells, which may be one of reasons for cell growth inhibition. Moreover, the cyclin kinase inhibitor p27 Kip1 , a downstream target of FoxO1, acting as a potent inhibitor of cyclin/cdk complexes in the S-phase of cell cycle progression was also tested [83][84][85][86]. It was found herein that 5-FU increased the expression of p27 kip1 . In addition, 5-FU simultaneously reduced the expression of Cyclin D1. It is of note that transcriptional repression of D-type cyclins is vital to the FoxOinduced cell-cycle arrest, which is evidenced by transcriptional pro ling and mRNA analysis. D-type cyclins are required for phosphorylation and inactivation of the retinoblastoma tumor suppressor protein (pRb), an essential determinant of cell-cycle progression in G1 [87,88]. To sum up, 5-FU-induced HS-5 cell growth inhibition is probably associated with FoxO1 targeted apoptosis or cell cycle arrest.
The traditional Chinese medicine Angelica sinensis which is commonly used to enrich blood, promote blood circulation [39]. The active constituents of Angelica sinensis include polysaccharides, organic acid sand phthalides, among which Angelica sinensis polysaccharides (ASP) are regarded as the main biological activity ingredient responsible for pharmacological effects with multi-target property [89]. ASP have attracted more and more attention to its bene cial effects, such as hematopoietic effects [90], immunologic enhancement [91], anti-tumor activity [92,93], and anti-radiation damage [94]. The antioxidant properties of ASP suppress the production of ROS and protected the endothelial progenitor cells, hepatocytes, myocardial cell and nerve cells from oxidative damage [95,96,40]. Moreover, evidences demonstrated that ASP promote cell proliferation, including in total spleen cells, macrophages [91], and gastric epithelial cells [97]. Our previous studies suggested that ASP reduced oxidative stress and oxidative DNA damage, boosted direct cell-cell contact between stromal cells and hematopoietic cells through Cx43 junctions, regulated cytokines, growth factors and chemokines such as CXCL12, SCF, GM-CSF, RANTES and thus provided a homeostatic microenvironment for hematopoietic stem/progenitor cells to regenerate following chemotherapeutic myelosuppression. In the present study, it was further demonstrated that ASP protected HS-5 cells from 5-FU-induced proliferation inhibition and ameliorated cellular oxidative stress via the mechanism of up-regulation of Wnt/β-catenin signaling. Most importantly, it was rst evidenced herein that ASP balanced the relationship between FoxO-mediated transcription and Wnt signaling in BMSCs under oxidative stress, which might be promising for clinical therapeutic use of ASP to myelosuppression.

Conclusions
In conclusion, the present study has reported that ASP protect stromal cells against 5-FU-induced proliferation inhibition and apoptosis via activating the Wnt/β-catenin signaling pathway directly or the indirect effects on Wnt/β-catenin signaling by down-regulation of its antagonizing FoxO1, suggesting a broad role for ASP as a potential antioxidant protective agent for chemoradiation therapeutic preventive agents.  Model of Wnt signaling cascade.