Cancer stem cells have a very typical growth pattern in that they can be grown in suspension in serum-free medium with the addition of growth factors, eventually forming cancer stem cell spheroids with a three-dimensional structure. The ability to form spheroids is an important way to identify cancer stem cells in vitro (Masuda et al., 2016). Therefore, we investigated the effect of TP5 on spheroid formation of HCT116 cancer stem cells. The number of cancer stem cell spheroids was significantly less in HCT116 cells treated with TP5 than in the control group, and the spheroid formation rate was significantly decreased in a concentration-dependent manner (Figures 1A,B), indicating that TP5 reduced the stemness of HCT116 cancer stem cells and inhibited the self-renewal and proliferation ability of cancer stem cells.

To further verify the effect of TP5 on colon cancer stem cells, we used flow cytometry to detect the expression of cancer stem cell markers CD24, CD44 and CD133, which are surface markers that can be specifically expressed by colon cancer stem cells (Jaggupilli and Elkord, 2012; Sahlberg et al., 2014). In the concentration range of 100–800 μM, TP5 could concentration-dependently decrease the expression of CD24, CD44 and CD133 in HCT116 cells (Figures 1C–H). The percentage of CD24-positive cells decreased from 29.7 ± 1.9% to 26.9 ± 2.1%, 24.7 ± 2.9%, 21.5 ± 0.9% and 20.4 ± 1.2%, after 48 h of action of 100, 200, 400, and 800 μM of TP5, respectively (Figures 1C,D). The percentage of CD44-positive cells decreased from 48.0 ± 2.4% to 46.0 ± 2.1%, 44.2 ± 6.6%, 36.6 ± 5.8%, and 34.5 ± 4.3%, respectively (Figures 1E,F). The percentage of CD133 positive cells decreased from 63.4 ± 0.5% to 63.9 ± 0.5%, 63.2 ± 1.5%, 58.5 ± 1.7%, and 53.4 ± 1.0%, respectively (Figures 1G,H).

We also investigated the effect of TP5 on cancer stem cell spheroid formation of another cell line, LoVo. The number of cancer stem cell spheroids was significantly less in LoVo cells treated with TP5 than in the control group, and the rate of spheroid formation was also significantly decreased (Figures 2A,B), indicating that TP5 also reduced the stemness of LoVo cancer stem cells. At a concentration of 800 μM, TP5 reduced the expression of CD24, CD44 and CD133 in HCT116 cells (Figures 1C–H). The percentage of CD24-positive cells decreased from 18.7 ± 1.8% to 14.0 ± 0.4% after 24 h of TP5 at 800 μM (Figures 2C,D). The percentage of CD44-positive cells decreased from 30.7 ± 1.3% to 23.7 ± 1.8%, respectively (Figures 2E,F). After 24 h of TP5 at 400 and 800 μM, the percentage of CD133 positive cells decreased from 54.8 ± 1.1% to 49.3 ± 1.2% and 47.2 ± 1.1%, respectively (Figures 2G,H). These results suggest that TP5 inhibition of cancer stem cells is not limited to HCT116 cells.

Although TP5 inhibits both HCT116 and LoVo cancer stem cells, it has a slightly stronger ability to inhibit HCT116 stem cell formation, we therefore focused on HCT116 as an example and continued to investigate the mechanism of TP5 inhibition of cancer stem cell formation and its application.

ALDH1, OCT4, SOX2 and Nanog are all stemness markers of cancer stem cells and have important roles in maintaining the nature of cancer stem cells and drug resistance (Anorma et al., 2018). We used Real-time PCR to detect the expression of cancer stem cell-related genes ALDH1, SOX-2, OCT-4, and Nanog. The expressions of ALDH1, SOX-2, OCT-4 and Nanog were significantly decreased after 100–800 μM TP5 were applied to HCT116 cells for 48 h (Figures 3A–D). It is suggested that TP5 may inhibit the self-renewal and tumorigenesis of colon cancer stem cells by regulating the important stemness genes ALDH1, SOX-2, OCT-4, and Nanog. Additionally, 100–800 μM TP5 also could significantly decrease the expression of CD24, CD44 and CD133 genes in HCT116 cells (Figures 3E–G). We also examined ALDH enzyme activity as a marker of cancer stem cell stemness (Jiang et al., 2019). TP5 concentration-dependently inhibited ALDH activity in HCT116 cells (after 48 h of TP5 action at 100, 200, 400, and 800 μM, ALDH activity decreased to 0.94 ± 0.10, 0.70 ± 0.05, 0.69 ± 0.02 and 0.33 ± 0.02; Figure 3H).

The Wnt signaling pathway is a key pathway for cancer stem cell development, and the sustained activation of the Wnt signaling pathway leads to the generation of cancer stem cells, which is an important factor for tumor resistance to conventional chemotherapy. Therefore, blocking the Wnt signaling pathway may be the key to removing cancer stem cells for the treatment of colon cancer (Polakis, 2007). β-catenin is a key downstream effector molecule in the Wnt signaling pathway (Liu et al., 2002). After the action of TP5 at 100–800 μM (Figure 4), HCT116 cells exhibited decreased PI3K (110 β) expression (100–400 μM effective), decreased AKT (P-Ser473 + Tyr474) phosphorylation levels (400–800 μM effective), and decreased WNT1 expression (200–800 μM effective); phosphorylated β-catenin (Ser33/37/Thr41) was increased (100–800 μM effective), leading to β-catenin degradation and expression decreased (800 μM effective) (the slightly different sensitivity of these indicators to TP5 may be due to some complex association of these factors), suggesting that TP5 inhibited colon cancer stem cells by affecting the wnt/β-catenin signaling pathway, which in turn inhibited colon cancer stem cells.

As a universal second messenger, calcium ions (Ca²⁺) promote the growth and metastasis of certain malignant tumor cells (Roderick and Cook, 2008; Yang et al., 2021). The self-renewal, drug resistance, and tumorigenic properties of colon cancer stem cells may function by regulating the inward flow of Ca²⁺ (Terrie et al., 2019); Ca²⁺ can promote the growth and spread of cancer stem cells. The percentage of Fluo-4 positive cells decreased from 52.0 ± 2.1% to 47.5 ± 3.0%, 43.5 ± 6.7%, 38.3 ± 2.4% and 32.5 ± 0.6% after 24 h of action of TP5 on HCT116 cells with 100, 200, 400 and 800 μM, respectively (Figures 5A,B), indicating that TP5 can concentration dependently reduce the intracellular Ca²⁺ concentration of HCT116 cells and thus inhibit the properties of cancer stem cells.

Given the possible resistance of cancer stem cells to chemotherapeutic agents (Prieto-Vila et al., 2017) and the diminished proliferation inhibition of tumor cells by chemotherapeutic agents, the direct inhibitory effect of TP5 on colon cancer stem cells suggests that the combination of TP5 with chemotherapeutic agents may enhance the therapeutic effect of these chemotherapeutic agents. Consistent with this speculation, we show that 25–100 μM of chemotherapeutic drug OXA produced inhibition on the proliferation of HCT116 in a concentration-dependent manner, while 100–800 μM TP5 significantly enhanced the anti-proliferative effect of OXA (Figure 6A).

To rule out the possibility that TP5 significantly enhanced the anti-proliferative effect of OXA on HCT116 because TP5 directly inhibited the proliferation of HCT116 cells, we treated HCT116 cells with TP5 for 48 h and observed its inhibitory effect on the proliferation of HCT116.100–800 μM TP5 did not cause significant visible changes in the cell viability and morphology of HCT116 cells compared with the control (Figures 6B,D); also, 100–800 μM TP5 did not induce apoptosis in HCT116 cells (Figure 6C). These results suggested that TP5 did not have a direct effect on the proliferation, morphology, and apoptosis of HCT116 cells.

To further verify whether TP5 promotes the antitumor effect of OXA by inhibiting colon cancer stem cells, we measured the expression of cancer stem cell markers CD24 and CD44 by flow cytometry after TP5 and OXA were co-administered to colon cancer HCT116 cells for 24 h. Compared with the control group, 5 μM OXA could not change the proportion of CD24-positive cells (28.6 ± 0.5% vs 27.8 ± 0.5%, p > 0.05; Figures 7A,B), while the proportion of CD24-positive cells decreased from 28.6 ± 0.5% to 18.3 ± 0.3% after 800 μM TP5 was administered to HCT116 cells for 24 h (p < 0.01, Figures 7A,B). Co-administration of TP5 (800 μM) with OXA (5 μM) further reduced the proportion of CD24-positive cells to 15.7 ± 0.4% (p < 0.01, Figures 7A,B). Similarly, 5 μM OXA could not change the proportion of CD44-positive cells, while the proportion of CD44-positive cells decreased from 46.6 ± 0.7% to 36.7 ± 0.7% after 24 h of 800 μM TP5 action on HCT116 cells, and the combined administration of TP5 (800 μM) with OXA (75 μM) further reduced the proportion of CD44-positive cells proportion to 33.6 ± 0.4% (p < 0.01, Figures 7C,D). It suggests that TP5 may inhibit colon cancer stem cells and therefore possibly sensitize the antitumor effect of OXA.

Our previous studies suggested that D-tubocurarine (TUB), an antagonist of nAchRs (Fan et al., 2006b), significantly antagonized the inhibitory effect of TP5 on the proliferation of leukemic HL60 cells, whereas the inhibitor of muscarinic acetylcholine receptors (mAchRs), atropine (Atr), did not have this antagonistic effect, suggesting that the inhibitory proliferative effect of TP5 on leukemic cells may be achieved through nAchRs (Fan et al., 2006a). We further investigated whether TP5 inhibited nAchRs via acetylcholine receptors in HCT116 colon cancer stem cells by using agonists and antagonists of nAchRs. TUB (10–100 μg/ml) significantly antagonized this effect of TP5 (Figures 8A–D). After 24 h of TP5 action on HCT116 cells with 800 μM, the percentage of CD24-positive cells decreased from 27.9 ± 0.5% to 19.3 ± 0.3% (p < 0.01) and the percentage of CD44-positive cells decreased from 48.0 ± 0.3% to 38.4 ± 0.7% (p < 0.001); whereas the early addition of 10 μg/ml or 100 μg/ml TUB, the proportion of CD24-positive cells reversed to 25.9 ± 0.2% (p < 0.01, vs. 19.3 ± 0.3%) and 24.9 ± 0.8% (p < 0.01, vs. 19.3 ± 0.3%); the proportion of CD44-positive cells reversed to 39.7 ± 1.4% (p > 0.05 vs. 38.4 ± 0.7%) and 45.0 ± 0.9% (p < 0.01 vs. 38.4 ± 0.7%).

In contrast, the cholinergic agonist the nicotine L-Nicotine (L-Nico) further promoted the inhibitory effect of TP5 on cancer stem cells in colon cancer cells (Figures 8E–H). Pretreatments of L-Nico (0.1 and 1 μg/ml) further reduced the proportion of CD24-positive cells to 14.0 ± 1.7%, 12.5 ± 0.9% (p < 0.01, vs. 19.3 ± 0.3%), and the proportion of CD44-positive cells to 34.3 ± 1.6%, and 33.5 ± 1.2% (p < 0.05 vs. 38.4 ± 0.7%).

We further found that the inhibitor TUB (1–100 μg/ml) significantly blocked the TP5-promoted proliferation inhibition of HCT116 by OXA, whereas the agonist L-Nico promoted this effect of TP5 (Figure 9A, B), implying that nAchRs might be involved in TP5-mediated inhibition of HCT116 colon cancer stem cells inhibition, thereby interrupting the TP5-promoted proliferation inhibition of HCT116 by oxaliplatin. Certainly, since restoration of cell viability in the presence of TP5 is only partial, there may be other mechanisms for TP5 to act on HCT116 cells in addition to nAchRs.

TP5 was purchased from ChemBest Research Laboratories Ltd. (China). Oxaliplatin (OXA), D-tufosine pentahydrate (TUB) and L-nicotine (L-Nico) were purchased from MedChemExpress (USA). HCT116 cells were cultured in RPMI-1640 medium (C11875500BT, Gibco) supplemented with 10% fetal bovine serum (10099–141, Gibco), penicillin (100 IU/ml) and streptomycin (100 μg/ml, C0222, Beyotime). LoVo cells were cultured in F-12K Nutrient Mixture (21127–022, Gibco) supplemented with 10% fetal bovine serum (10099–141, Gibco), penicillin (100 IU/ml) and streptomycin (100 μg/ml, C0222, Beyotime). Cells were incubated in a humidified incubator containing 5% CO₂ at 37°C.

The serum-free medium for spheroid culture consisted of DMEM/F12 (SH30023.01, HyClone) medium supplemented with 0.4% BSA (B2064-100G, Sigma), 20 ng/ml human recombinant epidermal growth factor (AF-100-15-100UG, PeproTech), 20 ng/ml human recombinant basic fibroblast growth factor (100–18 B-10UG, PeproTech), 25 μg/ml human insulin (HY-P0035, MCE), 2% B27 supplement (50×) (17504–044, Gibco), and 0.1% 2-mercaptoethanol (21985023, Gibco). Cells were plated in Corning Ultra-Low Attachment 24-well plates (3473, Costar) at a density of 5,000 viable cells/well and grown in serum-free medium. Cells were grown for 6 days and the number of spheroids was counted under the microscope (DMI3000 B, Leica) as previously described (Tominaga et al., 2019).

HCT116 cells or LoVo cells (1.5 × 10⁵ cells/well) were seeded in 12-well plates and exposed to different concentrations of TP5 for 24 h. The supernatant was then removed and cells were digested with 0.25% trypsin (25200-072, Gibco). After washing, cells were resuspended in 100 ul of PBS (ST476, Beyotime). For surface staining, cells were labeled with PE-CD24 (555428, BD Pharmingen), FITC-CD44 (555478, BD Pharmingen), and PE-CD133 (12-1338-41, eBiosciences) (He et al., 2021). Cell suspensions were incubated with the appropriate antibodies for 30 min at room temperature in the dark, followed by a washing step to remove unlabeled antibodies. Flow cytometry analysis was performed using BD FACS Jazz™ (BD Biosciences) and analyzed by FlowJo software.

Total RNA was isolated and reverse transcribed using EZ-press RNA purification kit (B0004-plus, EZBioscience) and PrimeScript RT kit (RR047A, TaKaRa), respectively, according to the manufacturer’s instructions (Wang et al., 2013), and qPCR was performed using TB Green Premix Ex Taq (RR420A. TaKaRa) for qPCR. relative changes in gene expression were determined using the 2-ΔΔCt method, and relative mRNA expression was normalized to GAPDH. Following primer sets were used for analyzing the expression of specific genes including ALDH1, forward: 5′-ACG​CCA​GAC​TTA​CCT​GTC​CTA​CTC-3′ and reverse: 5′- TCT​TGC​CAC​TCA​CTG​AAT​CAT​GCC-3′; SOX-2, forward: 5′- GCT​CGC​AGA​CCT​ACA​TGA​ACG​G-3′, and reverse: 5′-AGC​TGG​CCT​CGG​ACT​TGA​CC -3′; OCT-4, forward: 5′- GATGTGGTCCGAGTGTGGTTCTG-3′and reverse: 5′-CGA​GGA​GTA​CAG​TGC​AGT​GAA​GTG -3′; Nanog, forward: 5′- AGC​AAT​GGT​GTG​ACG​CAG​AAG​G-3′, and reverse: 5′-ACC​AGG​TCT​GAG​TGT​TCC​AGG​AG -3′; CD24, forward: 5′- CTC​CTA​CCC​ACG​CAG​ATT​TAT​TC-3′, and reverse: 5′-AGA​GTG​AGA​CCA​CGA​AGA​GAC-3′; CD44, forward: 5′- CTG​CCG​CTT​TGC​AGG​TGT​A-3′, and reverse: 5′-CAT​TGT​GGG​CAA​GGT​GCT​ATT-3′; CD133, forward: 5′- AGT​CGG​AAA​CTG​GCA​GAT​AGC-3′, and reverse: 5′-GGT​AGT​GTT​GTA​CTG​GGC​CAA​T-3′; GAPDH, forward: 5′-TTG​GTA​TCG​TGG​AAG​GAC​T-3′, and reverse: 5′-GGA​TGA​TGT​TCT​GGA​GAG​C-3′.

ALDH activity was assessed using an ALDH activity detection kit (BC0755, Solarbio) according to the manufacturer’s instructions (Jiang et al., 2019). HCT116 cells (3.5 × 10⁵ cells/well) were seeded in 6-well plates with different treatments of TP5 (0, 100, 200, 400, and 800 μM) for 48 h at 37°C. Cells were lysed with ultrasound, centrifuged and the supernatant (100 μL/well) was transferred to UV-Star 96-well plates (655801, Greiner Bio-one) and incubated at 37°C for 10 min. ALDH activity was measured at 340 nm using a microplate reader (VarioskanTM LUX, Thermo).

After treatment with TP5 (100, 200, 400 and 800 µM) for 48 h, total proteins from HCT116 cells were lysed in lysis buffer (P0013, Beyotime) containing 1% cocktail (B14001, Bimake) (Li et al., 2012; Wang et al., 2017). Cell lysates were cleared by centrifugation at 12,000 rpm for 15 min at 4°C. The concentration of total protein was determined using the Enhanced BCA Protein Assay Kit (P0010, Beyotime). Protein sample buffer was added and samples were boiled at 100°C for 10 min. Briefly, 20 µg of protein was loaded into each lane and transferred to a polyvinylidene difluoride (PVDF) membrane. The membranes were incubated overnight at 4°C with primary antibody, and secondary antibodies anti-rabbit IgG HRP-conjugated (1: 3,000, 7074S, Cell Signaling Technology) and anti-mouse IgG HRP-conjugated (1: 3,000, 7076S, Cell Signaling Technology) were incubated at 25°C for incubate for 2 h. The primary antibodies used are Beta Catenin Rabbit Polyclonal Antibody (1: 1,000, 51067-2-AP, Proteintech), WNT1 Rabbit Polyclonal Antibody (1: 1,000, 27935-1-AP, Proteintech), Phospho-beta Catenin (Ser33/37/Thr41) antibody (1: 1,000, 9561, Cell Signaling Technology), AKT1 (Phospho-Ser473 + Tyr474) antibody (1: 1,000, 12669, Abcam), PI3K p110 (beta) polyclonal antibody (1: 1,000, 20584-1-AP, Proteintech), AKT1 polyclonal antibody (1: 1,000, 10176-2-AP, Proteintech), and GAPDH (1: 10,000, 60004-I-Ig, Proteintech). Phosphorylated protein bands were categorized as their total protein level, and other bands were categorized as GAPDH.

HCT116 cells were incubated in 12-well plates with different treatments of TP5 (0, 100, 200, 400, 800 μM) for 24 h at 37°C. To determine the cellular calcium concentration (He et al., 2021), the cell membrane Ca²⁺-free concentration was assessed using the fluorescent Ca²⁺-probe Fluo-4 AM (S1060, Beyotime). Briefly, cells were incubated in PBS at 37°C with 2 μM Fluo-4 AM for 30 min in the dark, followed by three washes with PBS. Binding buffer (500 μL) containing 1% FBS (Gibco, Australia) was added and incubated for 30 min in the dark at room temperature. The results were then evaluated by flow cytometry. The excitation wavelength was 488 nm and the emission wavelength was 520 nm. The experiment was repeated at least three times.

Cell viability was assessed using the MTT assay kit (C0009, Beyotime) according to the manufacturer’s protocol (Foldbjerg et al., 2011). HCT116 cells were seeded in 96-well plates at a density of 15 × 103 cells/well and incubated overnight at 37°C. Next, cells were treated with TP5 and OXA and incubated for 48 h 10 μl MTT was added to each well and incubation was continued for 4 h. The absorbance was measured at 570 nm using a microplate reader (VarioskanTM LUX, Thermo). Cell viability was estimated by comparing the relative absorbance values with those of the untreated samples.

The extent of apoptosis in treated HCT116 cells was assessed using the Annexin V-FITC/PI Apoptosis Assay Kit (556547, BD Pharmingen) according to the manufacturer’s instructions (Park et al., 2017). HCT116 cells (1.5 × 10⁵ cells/well) were seeded in 12-well plates, exposed to different concentrations of TP5 for 48 h, and stained with 5 μl of FITC-labeled Annexin V and 5 μl of PI simultaneously at room temperature, protected from light. Stained cells were analyzed using a fluorescence-activated cell sorter flow cytometer (BD FACS Jazz™, BD Biosciences). A minimum of 10,000 cells were used for each analysis, and experiments were performed in triplicate.

Data are expressed as mean ± SD. All experiments were performed independently, at least 3 times. Statistical analyses were performed as described in each corresponding legend. Differences between two groups were assessed by unpaired two-sided Student’s t-test, and differences between multiple groups were assessed by one- or two-way ANOVA and Dunnett’s post hoc test. p less than 0.05 was considered statistically significant.