Wnt3a signaling with serum supply induces replication stress in cultured cells

Abstract

Wnt signaling plays a central role in tissue development and homeostasis, and its deregulation is implicated in many human diseases, including cancer. As an essential posttranslational modification, protein phosphorylation is critical in Wnt signaling and has been a focus of investigation using systematic approaches, including proteomics. Typically, studies were conducted by applying purified Wnt ligands to cells in a “starvation” condition to minimize the background noise. Despite leading to many important discoveries, such an approach may omit pivotal integrative effects of Wnt signaling in a complex physiological environment. In this study, we investigated the temporal dynamics of the phosphoproteome following treatments of Wnt3a conditioned medium (CM) with serum supply. This revealed three clusters of phosphoproteome changes with distinct temporal profiles with implications in gene expressions and chromatin organizations. Among these, we observed enhanced phosphorylation at the Thr543 residue of 53BP1, which is a key event in the cellular response to DNA damage. Functionally, it triggered the replication stress response pathway mediated by γH2AX accumulation and Chk1 activation, leading to a significant reduction of cells in the S phase of the cell cycle. Intriguingly, Wnt3a treatment in the serum-free condition did not activate 53BP1-Chk1 and replication stress response. Our study indicates the importance of noting the presence or absence of serum supply when studying the signaling pathways.

Highlights

• •Wnt3a stimulation induces different patterns of dynamic phosphorylation on proteins.
• •Wnt3a causes different cell responses under serum and serum-free conditions.
• •Wnt3a induces replication stress response under serum supply.
• •Wnt3a leads to a reduction of cells in the S phase of the cell cycle in serum conditions.

1. Introduction

Wnt proteins are a family of highly conserved signaling molecules that regulates a broad range of metazoan development and tissue homeostasis [1]. At the molecular level, the Wnt signaling activities are organized into several major pathways: the canonical/β-catenin dependent pathway and the so-called noncanonical pathways, such as the planar cell polarity pathway and calcium pathway [2]. Different Wnt pathways are orchestrated by intricate molecular events that integrate into the cellular signaling networks. The canonical/β-catenin signaling is the most extensively studied Wnt pathway. At the intermediate level, the Wnt/β-catenin cascade is propagated by a wide range of signaling events that involve changes in protein stability, interactions and posttranslational modifications (PTMs) [3]. Such proteomic changes have been a focus of investigations to understand Wnt signaling mechanisms over the past years [[4], [5], [6]].

Protein phosphorylation is an indispensable PTM event that regulates signaling propagation. Several studies using mass spectrometry (MS)-based proteomics have investigated the phosphoproteome during activation of the canonical Wnt pathway and provided pivotal insights into how phosphorylation is implicated in regulating this important pathway [7,8]. Typically, investigations of Wnt signaling involved using purified recombinant Wnt ligands to stimulate cells cultured under serum starvation, which is a commonly performed procedure in probing signaling activities. Although serum starvation is frequently assumed to minimize the basal activity of cells, it may induce complex and cell-type dependent effects [9]. In addition, an important activity of canonical Wnt signaling is to stimulate cell proliferation, which can be compromised by serum starvation. Therefore, certain physiological effects of Wnt regulation may be omitted in conventional serum-starvation approaches.

In this study, we examined the temporal dynamics of phosphoproteome stimulated by canonical Wnt signaling in cell cultures with serum supply. Label-free quantitative MS and multivariate analyses revealed the complex phosphoproteome dynamics. Interestingly, we identified that Wnt3a treatment with serum supply activated the 53BP1-Chk1 mediated replication stress, which was absent in the serum-starvation condition. Our study provided novel insights into the importance of serum supply in studying cell signaling.

2. Materials and methods

2.1. Cell cultures, reagents, and antibodies

HEK293T, A549, mouse control L cells, and Wnt3a-expressing L cells were obtained from American Type Culture Collection (ATCC). All the cells were cultured under the standard condition in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (Hyclone, USA) and 1% penicillin-streptomycin (Gibco, USA). Anti-Dvl2 (sc13974), anti-Chk1 (sc-8408), anti-β-catenin (sc-7963) and anti-β-actin (sc-47778) antibodies were purchased from Santa Cruz Biotechnology. Anti-phospho-53BP1(Thr543) (#3428S), anti-phospho-Chk1 (Ser354) (#2341) and anti-phospho-Histone H2A.X Ser139 (#2577) antibodies were purchased from Cell Signaling Technology and anti-53BP1 (ab21083) antibody was purchased from Abcam. Recombinant Wnt3a protein (#5036-WN) was purchased from R&D Systems.

2.2. Conditional medium preparation, and cell stimulation

Control L cells and Wnt3a-expressing L cells were cultured in DMEM supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin at 37 °C. At confluency, DMEM with 5% FBS and 1% penicillin-streptomycin was applied to cells for conditional medium (CM) collection every 24 h for three days. The three batches of CM were pooled and centrifuged at 2500 rpm and 4 °C for 10 min to remove cell debris, followed by storage at −80 °C. For stimulation, HEK293T cells were washed with DMEM and cultured in control- or Wnt3a-CM and cell lysates were collected at different time intervals (0 min, 5 min, 30 min, and 120 min) for MS sample preparation. The experiments at each time point were conducted in biological triplicates. For validation experiments, HEK293T or A549 cells were cultured with control-CM or Wnt3a-CM for 2 h before collecting cell lysate for Western blot analysis.

2.3. Protein extraction and quantification

HEK293T cells were collected by centrifugation and washed three times with cold PBS. The cell pellets were gently homogenized in an ice-cold SDS lysis buffer (50 mM Tris pH 7.4, 4% SDS, 65 mM DTT, 2% protease inhibitor cocktail (v/v), 1% Triton X-100 (v/v), 1 mM EDTA, 1 mM AEBSF, 1 mM NaF and 1 mM Na₃VO₄. The homogenates were chilled on ice for 2 min before sonication (400 W, 10 s) and centrifugation at 20, 000 g at 4 °C for 0.5 h. The supernatant was collected, and the extracted proteins were precipitated with 5 vol of ice-cold acetone/ethanol/acetic acid (v/v/v = 50/50/0.1) at −20 °C, followed by centrifugation at 14,000 g for 30 min. The pellet was dissolved in buffer containing 50 mM Tris/HCl, pH 8.2 and 8 M urea. The protein concentration was determined by Bradford assay.

2.4. Western blot

Equal amounts of total proteins (20–40 μg) were resolved on 4–20% gradient Tris-Gly PAGE (Beyotime, P0468 M) gels and transferred to nitrocellulose membranes (NC) (BioRad, 170–4159) using constant current mode at cold room. Following blocking in 5% skim milk (Biofroxx, 927–40000) dissolved in TBST (20 mM Trizma base, 150 mM NaCl, 0.5% Tween-20) for 1 h at room temperature. Membranes were incubated with following primary antibodies: Anti-phospho-53BP1(Thr543) (#3428S), anti-53BP1 (ab21083), anti-phospho-Chk1 (Ser354) (#2341), anti-Chk1 (sc-8408), anti-phospho-Histone H2A.X Ser139 (#2577), anti-Dvl2 (sc13974), anti-β-catenin (sc-7963) and anti-β-actin (sc-47778) overnight at 4 °C. The primary antibodies were diluted at 1:1000 in TBST. Membranes were incubated with secondary antibodies for 1 h at room temperature after washing with TBST four times and visualized using BioRad ChemiDoc (BioRad) to capture the results after washing with TBST four times.

2.5. Protein digestion and phosphopeptides enrichment

Protein samples were first reduced by DTT at 37 °C for 1 h and then alkylated by 25mM iodoacetamide in the dark at room temperature for 30 min. Then the urea concentration in the solution was diluted to 1 M with 50 mM Tris/HCl (pH 8.2). Trypsin digestion was performed at 37 °C overnight with an enzyme-to-protein ratio of 1/50 (w/w). The resulting peptide mixture was desalted using pre-equilibrated C18 SepPak cartridges before phosphopeptide enrichment, or storage at −80 °C before analysis of protein abundance.

For phosphopeptide enrichment, the digested proteins were subjected to enrichment with immobilized titanium (IV) ion affinity chromatography (Ti⁴⁺-IMAC) for phosphopeptides as previously reported [10]. Briefly, a sieve plate with 2 mm diameter was packed into a tip (200 μL). The peptides from 0.5 mg of initial cell lysate were then dissolved in loading buffer containing 80% (v/v) acetonitrile and 6% (v/v) TFA. The sample was incubated with Ti4+-IMAC beads at room temperature for 30 min after centrifugation. The Ti4+-IMAC beads were washed twice with washing buffer I (50% (v/v) acetonitrile, 6% (v/v) TFA, 200 mM NaCl) after centrifugation, then twice with washing buffer II (30% (v/v) acetonitrile, 0.1% (v/v) TFA) after centrifugation. The phosphorylated peptides were eluted with 100 μL of 10% (v/v) ammonia directly into 100 μL of 10% formic acid. Samples were vacuum-dried down and stored at −80 °C until subjected to LC−MS/MS.

2.6. Mass-spectrometry and data processing

The obtained peptide samples of total proteome or phosphoproteome were reconstituted in 20 μL 0.1% formic acid in water and performed on an Easy-nLC 1200 system (Thermo Scientific) fitted with a reverse phase C18 column (75 μm i.d. × 15 cm, 3 μm particle size), analyzed by an Q-exactive HF Orbitrap mass spectrometer (Thermo Scientific). Each sample was loaded in 0.1% formic acid in ultrapure water (buffer A) and eluted in 0.1% formic acid in 80% acetonitrile (buffer B) run with a linear 75-min gradient of 7–95% buffer B at flow rate of 250 nL/min. In the mass spectrometer, samples were prepared in triplicate and operated in positive ion mode acquiring a survey mass spectrum with a mass resolution (R) of 120,000, m/z = 350–1800 using an automatic gain control (AGC) target of 3 × 10⁶. The 12 most intense ions were selected for higher-energy collisional dissociation (HCD) fragmentation (normalized collision energy 27) and MS/MS spectra were generated with an AGC target of 1 × 10⁵ at a resolution of 30,000. The dynamic exclusion time was set to 30 s.

The MS/MS spectra raw data were searched against the human proteome using Sequest HT node integrated with the Proteome Discoverer (PD) software (Version 2.2, Thermo Scientific). The precursor and fragment mass tolerances were set to 10 ppm and 0.02 Da, respectively. Two missed cleavages were allowed for trypsin digestion. Carbamidomethyl (C) was chosen for fixed modifications; oxidation (M), phosphorylation (S, T, Y) and N-terminal acetylation as variable modifications. False discovery rate (FDR) of peptide spectrum matches (PSMs) was determined by the Percolator algorithm at 1% based on Q-value. For quantification, the precursor ion areas in a node-based processing and consensus workflow in PD 2.2 were used. Minora Feature alignment and feature mapping calculated the areas of the ions in the MS1 scan. The normalized abundance values of proteins were obtained via a label-free quantification method.

2.7. Immunofluorescence

The immunofluorescence procedure was briefly described as follows. Firstly, HEK293T cells were cultured on the coverslips overnight. After changing the medium to an FBS-free medium and starving at 37 °C under 5% CO₂ for 3 h, the cells were treated with Ctrl-CM and Wnt3a-CM for 2 h. Subsequently, the cells were washed with warm PBS twice and fixed the cells with 4% PFA at room temperature for 10 min. After washing the fixed cells with PBS three times for 5 min per wash, the cells were treated with 0.5% triton in PBS at 4 °C for 10 min, then washed with PBS-Tween buffer (0.05% Tween 20 in PBS) three times, for 10 min per wash. The reaction was blocked with IF buffer (1% BSA in PBS-Tween buffer) for 1 h, and then incubated with anti-phospho-53BP1 (Thr543) and anti-phospho-Histone H2A.X (Ser139) antibody diluted in IF buffer at 4 °C for overnight, respectively. After finishing the primary antibody incubation, the coverslips were washed with PBS-Tween buffer three times, for 5 min per wash, then reacted with DAPI and secondary antibody in IF buffer for 1 h at room temperature. Finally, the coverslips were washed with PBS-Tween buffer three times for 5 min to prepare the sample for imaging using the Nikon A1HD25 High speed and Large Field of View Confocal Microscope (Nikon).

2.8. Edu cell cycle flow cytometry

After plating HEK293T cells in 6-well plates overnight, the medium was changed to an FBS-free medium for starvation at 37 °C under 5% CO₂ for 3 h. Then, the cells were treated with Wnt3a-CM and Ctrl-CM for 2 h. After washing three times with PBS containing 1% BSA, the cells were collected and stained by the Click-iT™ Plus EdU Flow Cytometry Assay Kit (Invitrogen, C10632). Cells were collected by centrifugation and mixed well with 100 μL of Click-iT™ fixative. The cells were then incubated for 15 min at room temperature in the dark. After incubation, we washed the cells with 3 mL of 1% BSA in PBS and then collected the pellets. Subsequently, the cells were resuspended with 100 μL of 1X Click-iT™ permeabilization reagent and incubated for 15 min. Afterward, we incubated the cells in the Click-iT™ Plus reaction cocktail for 30 min at room temperature in the dark. After washing with 3 mL of 1X Click-iT™ permeabilization and wash reagent, stained the DNA by DAPI for 10 min. Finally, we used the Beckman Coulter CytoFLEX S Flow cytometer analyzer (Beckman, B49008AC) to analyze the cell cycles.

2.9. Data analysis

Label-free quantitative data from PD 2.2 were filtered with a cut-off of ≥2 observations in the 3 repeats of at least one time point. Log10 transformation was performed before the missing data points were imputed using random values generated from a normal distribution centered at the 1% quantile and the median SD. The abundance of phosphopeptides was normalized to the abundance of corresponding parental proteins. The Limma package was used in the R environment [11] to identify differential proteins or phosphorylation with an FDR-corrected ANOVA p-value of 0.05. The k-means clustering approach [12] was exploited to analyze the dynamic patterns of the differential proteins and phosphorylation events. For functional annotation and network analysis, the differentially regulated proteins were searched against the String database [13], and the PPI networks were visualized by using Cytoscape [14].

3. Results

3.1. Study the phosphor-proteome dynamics of Wnt3a signaling in the presence of serum

We performed label-free quantitative MS/MS profiling of phosphoproteomic dynamics of canonical Wnt signaling in serum-supplied context. The overall scheme of the study is illustrated in (Fig. 1). As a tool commonly used in studying canonical Wnt signaling, we generated conditional media from L cells stably expressing Wnt3a (Wnt3a-CM) or from control L cells (Ctrl-CM) constituted with 5% fetal bovine serum. Wnt3a-CM led to Dvl2 phosphorylation (Supplementary Fig. 1A) and β-catenin dephosphorylation and stabilization (Supplementary Fig. 1B), indicating the successful activation of canonical Wnt signaling in the serum-competent condition.

Fig. 1.

Schematic outline for Wnt3a-CM treatment schedule and dynamic phosphoproteome analysis. HEK293T cells were stimulated with Wnt3a-CM or Ctrl-CM in the presence of serum supply for 0, 5, 30, and 120 min. The cell lysates were then subject to trypsin digestion and phosphopeptides enrichment before label-free quantitative mass spectrometry analysis.

Phosphorylation is a fast and dynamic signaling activity. Upon the binding of ligands, the proximal events are initiated within seconds to a few minutes before the phosphorylation cascades propagate downstream for a longer period. Previous Wnt3a phosphoproteomic studies utilized periods of 0∼30 min [7] and up to 240 min [8] and provided important insights. To investigate the effects of Wnt3a signaling with serum supply, we chose to treat HEK293T cells with Wnt3a-CM or Ctrl-CM (with 5% fetal bovine serum) for 5 min, 30 min, and 120 min. Cells were also harvested before the treatment (0 min) as the starting point for comparison. Proteins were extracted and digested using trypsin. The phosphopeptides were enriched and analyzed using label-free quantification before normalizing to the abundance of the corresponding protein as recommended in other studies [15]. We conducted three biological replicates and identified 2162 phosphorylation events from 639 proteins in the Ctrl-CM treated samples and 2215 phosphorylation events from 644 proteins in the Wnt3a-CM treated samples (Supplementary Table S1). Notably, 632 of the phosphorylation proteins were shared between the Ctrl-CM and Wnt3a-CM groups (Supplementary Fig. 1C), suggesting that Wnt3a did not induce large changes in the identity of phosphorylated proteins in the presence of serum. To examine the effects of Wnt3a on the dynamic profiles of protein phosphorylation, we performed label-free quantitation of the data and the quantity of the phosphopeptides were normalized to the changes of the levels of parental proteins for downstream analyses. Differential analysis revealed that 424 and 436 phosphopeptides displayed significant changes over the 120-min time course in the Ctrl-CM and Wnt3a-CM group, respectively (Fig. 2A and Supplementary Table S1), with only 147 phosphorylation events commonly observed in both groups. These results indicate that

Fig. 2.

The changes of phosphoproteome following Wnt3a-CM treatment. (A) The Venn diagram illustrates the number of phosphorylation events with significant changes after Wnt3a-CM and Ctrl-CM treatment. (B–D) Three different changing patterns of phosphoproteins following Wnt3a-CM treatment. The phosphopeptide levels were normalized to the changes of their parental proteins, and the temporally changing profiles were algorithmically subdivided into three clusters using the k-means clustering method. (E) Biological process enrichment analysis of three different phosphoprotein clusters. (F) Network analysis of the changing phosphorylation events of which the parental proteins are associated with chromatin binding and/or remodeling.

Wnt3a led to quantitative changes in the dynamics of protein phosphorylation in the presence of serum.

3.2. Profile patterns of the phosphorylation dynamics upon Wnt3a-CM stimulation

We then focused on the 289 Wnt3a-CM-unique phosphorylation events to analyze the patterns of dynamic changes. Using k-means clustering analysis, we revealed three clusters of temporal profiles of the changing phosphorylation events induced by the Wnt3a-CM. (Fig. 2B–D and Supplementary Table S2). These include overall increase (Cluster 1, Fig. 2B), increase-then-decrease (Cluster 2, Fig. 2C), and a pattern of mild decrease (Cluster 3, Fig. 2D). Go term enrichment analysis of the dynamic phosphorylation events revealed enrichments of gene expression, RNA processing and nucleic acid metabolic processes, indicating their dynamic regulation in response to Wnt3a activation with serum supply (Fig. 2E). Interestingly, all three clusters had proteins of the changing phosphorylation events that possess chromatin binding and remodeling activities. Network analysis of protein-protein interactions using the information from the STRING database revealed TP53BP1 (from Cluster 1) as the center hub with 11 partner proteins. These includes TOP2A, RIF1, and CHD4 in Cluster 1, SMARCA1, CBX3, MRE11, CCAR2, MCM2, and SIN3A in Cluster 2, and MCM3 and TP53 in Cluster 3 (Fig. 2F).

3.3. Wnt3a induces 53BP1 phosphorylation at Thr543 with serum supply

Among these Wnt3a-responsive events, our attention was attracted to an increase of 53BP1 Thr543 phosphorylation (Fig. 3A), which is known to plays a key role in initiating the double strand DNA breaks (DSB) repair and replication stress [16]. Notably, the 53BP1 phosphorylation was long-lasting (more than 120 min in our experiments) and was in conjunction with the dynamic changes of phosphorylation in a number of 53BP1-interacting proteins (Fig. 2F). This result suggested that 53BP1 phosphorylation at Thr543 may play a role in regulating Wnt3a-induced chromatin reorganization and gene expression.

Fig. 3.

Wnt3a induces phosphorylation of 53BP1 at Thr543 site in serum supply.

(A) Thr543 phosphorylation of 53BP1 expression pattern based on time course after Wnt3a stimulate. Data are Mean ± SE from three biological replicates of each sample. (B and C) Immunoblotting and Immunofluorescence analysis of p-53BP1(Thr543) expression of HEK293T treated with Control-CM and Wnt3a-CM after 2 h. (D and E) Immunoblotting and Immunofluorescence analysis of p-53BP1(Thr543) expression of HEK293T treated with Medium and Wnt3a ligand after 2 h.

To validate this induction of 53BP1 phosphorylation at Thr543, HEK293T cells were treated by Wnt3a-CM or Ctrl-CM in the presence of serum for 2 h, before anti-53BP1 phospho-Thr543 assays. Western blot confirmed the increased 53BP1 phosphorylation at the Thr543 site (Fig. 3B) induced by Wnt3a-CM treatment. In line with this, immunolabeling also revealed a significantly higher level of phosphor-53BP1(Thr543) in HEK293T cells treated with Wnt3a-CM than Ctrl-CM (Fig. 3C). In parallel, a similar inducing effect was observed in A549 lung cancer cells treated with Wnt3a-CM (Supplementary Fig. 2A), supporting the general applicability of 53BP1 activation by Wnt3a in the presence of serum.

Next, we examined whether Wnt3a could lead to 53BP1 phosphorylation in the absence of serum. Intriguingly, Western blot (Fig. 3D) and immunolabeling (Fig. 3E) both showed that Wnt3a ligands did not lead to induce Thr543 phosphorylation of 53BP1 in serum-starvation cultures. The absence of induction was also observed in A549 cells (Supplementary Fig. 2B). These results indicated that the presence of serum was necessary for Wnt3a signaling in promoting 53BP1 phosphorylation at Thr543.

3.4. Wnt3a induces Chk1 activation and cell cycle arrest

53BP1 has been revealed as a key regulator of the ATR-Chk1 signaling during response to replication stress [17]. We hypothesized that the Wnt3a treatment in the presence of serum led to replication stress and 53BP1 activation. To address this, we examined Chk1 activation (indicated by phosphorylation at S345) and the phosphorylated form of H2AX S139 (γH2AX), a well-known marker of replication stress [18]. Immunoblotting showed that Wnt3a-CM treatment with serum supply induced both phosphor-S345 of Chk1 and the γH2AX levels (Fig. 4A). In parallel, immunolabeling and confocal imaging confirmed that Wnt3a-CM treatment led to a significantly higher level of γH2AX (Fig. 4B) than the Ctrl-CM treatment with serum supply. In contrast, in the serum-free condition, Wnt3a treatment did not induce Chk1 activation or γH2AX accumulation (Fig. 4A and C). Importantly, Wnt3a successful activated the canonical signaling in both serum-competent and serum-starvation conditions, as indicated by the stabilization of β-catenin (Fig. 4A).

Fig. 4.

Wnt3a induces Chk1 activation and cell cycle arrest.

(A) Immunoblotting analysis of phosphorylation (p) or total (t) abundance of 53BP1, phosphorylation (p) or total (t) abundance of Chk1, and γH2AX expression of HEK293T treated with Control-CM and Wnt3a-CM, Medium and Wnt3a ligand after 2 h. (B and C) Immunofluorescence analysis of γH2AX of HEK293T treated with Control-CM, Wnt3a-CM, Medium and Wnt3a ligand after 2 h. (D and E) Cell cycle detection by Edu staining of HEK293T treated with Control-CM, Wnt3a-CM, Medium and Wnt3a ligand after 2 h.

Replication stress is known to halt cell cycle in the G1-S phase. To further examine the effects of Wnt3a treatment with serum supply, we performed the Edu-incorporation assay and observed a significant reduction of cells in the cell cycle S phase, indicating a consequence of replication stress. In contrast, Wnt3a treatment in the serum-free condition had no significant effect on the cell cycle (Fig. 4E). Taken together, these results indicated that Wnt3a treatment with serum supply could lead to replication stress.

4. Discussion

In this study, we investigated the phosphoproteomic dynamics of the canonical Wnt signaling in cell cultures with serum supply. We identified phosphorylation events with changing patterns by analyzing the Wnt-responsive proteome after 5, 30, and 120 min of Wnt3a-CM treatment. Analysis of the phosphoproteome revealed different patterns of dynamic phosphorylation on proteins with chromatin/histone regulation activities in response to Wnt3a stimulation, indicating an epigenetic reorganization in response to Wnt3a stimulation. Moreover, we identified that Wnt3a treatment with serum supply induced replication stress response, indicated by the phosphorylation of 53BP1 at Thr543 and γH2AX accumulation under serum supply.

The phosphoproteomic dynamics of canonical Wnt signaling have been studied in serum starvation conditions [19]. Although serum starvation could (supposedly) reduce basal signaling activity and make the cell population more homogenous [20], it might obscure the signaling activities of Wnt in the physiological contexts that were highly integrative, especially considering the role of canonical Wnt signaling in promoting the cell cycle [21]. However, serum starvation has prominent effects on cell cycle [22]. To address this, cells in this study were stimulated with the Wnt3a-CM, a common tool of canonical Wnt signaling, in the presence of serum supply. We revealed complex phosphoproteome dynamics of various proteins that regulate gene expression and metabolic processes. It is important to note that Ctrl-CM prepared from the control L cells also led to substantial changes in the phosphoproteome that were not observed in Wnt3a-CM condition (Fig. 2, Fig. 4A). This large difference suggests that the control L cells and Wnt3a-producing L cells may have a different secreting profile in addition to the Wnt3a.

Our integrative analysis revealed the Wnt3a-CM treatment stimulates an increase of 53BP1 phosphorylation at Thr543. 53BP1 is best characterized as a key regulator of the DNA damage response and genomic stability [23]. In addition, 53BP1 also plays intricate roles in the cell cycle and proliferation independently of its DNA repair activity [24]. For example, 53BP1 has been shown to evoke stabilization of p53 and induce cell cycle arrest against aberrant mitosis [25]. The anti-proliferation property of 53BP1 is in line with its activation in the cells with decreased S phase proportion after Wnt3a treatment with serum supply. Interestingly, Tarsounas et al. recently reported that the activation of β-catenin suppresses the proliferation of cells with a deficiency in BRCA1/2, another important player in the DNA damage response [26]. It is plausible that the pro-growth Wnt/β-catenin signaling may elicit replication stress, which activates 53BP1 as a checkpoint to prevent/reduce aberrant replication and DNA damage. Further investigations are necessary to elucidate the mechanisms underlying the Wnt-serum synthetic effects on replication stress, which may provide therapeutic strategies for cancers with abnormal Wnt signaling.

Credit author statement

Z.L. and L.Z. designed the experiments. Y.W. and R.W. performed experiments, H.M. and L.Z. performed data analysis. All authors wrote the manuscript together.

Declaration of competing interest

The authors declare no competing interests.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

The data that has been used is confidential.