*2.2. PSY Induces Cell Cycle Arrest in the G2/M Phase and Apoptosis in Human CRC Cells*

As cell viability is regulated by cell proliferation and death, we investigated the effect of PSY on the cell cycle. As shown in Figure 2a, PSY increased the number of human CRC cells in the G2/M phase while gradually decreasing the number of cells in the G0/G1 phase. Accordingly, PSY decreased the expression levels of CDK1 and cyclin B1 in HCT116 cells and increased the phosphorylation of CDK1 in SW480 cells (Figure 2b), indicating that the inactivation of CDK1 leads to G2/M arrest. At the same time, PSY increased the CRC cell population in the sub-G1 phase (Figure 2a), suggesting the induction of apoptotic cell death. To further confirm whether PSY induces apoptosis, we stained PSY-treated CRC cells with annexin V-FITC and PI, and the proportion of apoptotic cells was determined. As shown in Figure 3a, PSY increased the number of annexin V-positive cells, indicating apoptosis. These results were also confirmed by terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assays and examination of increased cleaved PARP, another apoptosis marker (Figure 3b,c). These results indicate that PSY decreases the viability of human CRC cells by inducing cell cycle arrest at the G2/M phase and apoptosis.

**Figure 2.** Effects of PSY on cell cycle in human CRC cells. CRC cells were treated with PSY EtOH-Ex for 48 h. (**a**) The treated cells were stained with PI, and the staining was analyzed using flow cytometry. The representative analysis is shown in the left panel. Every PI staining and analysis was performed at least three times in duplicate or triplicate. Cells were quantitated as a percentage of cells in each phase. Data in the right panel represent the mean ± SEM (\*, *p* < 0.05; \*\*, *p* < 0.01 and \*\*\*, *p* < 0.001 versus control). (**b**) Whole lysates of the treated cells were prepared, and Western blot analysis for CDK1, CDK4, CDK6, Cyclin B1, and phospho (p)-CDK1 was performed, and β-actin was used as an internal control. Data represent the mean ± standard deviation (SD) of three independent experiments (\*, *p* < 0.05; \*\*, *p* < 0.01 and \*\*\*, *p* < 0.001 versus control).

#### *2.3. PSY Inhibits the STAT3 Pathway in Human CRC Cells*

In the next step of the study, we attempted to determine the mechanism by which PSY induced cell cycle arrest and apoptosis in human CRC cells. The CRC development and progression are closely related to inflammation and the key molecule STAT3 [4]. Therefore, we prepared PSY with PR, MCR, and CS, each of which was traditionally used for treating inflammatory diseases. In this regard, we explored the effect of PSY on STAT3 expression. As shown in Figure 4a, PSY effectively reduced the phosphorylation of STAT3 in human CRC cells. PSY also inhibited the nuclear translocation of STAT3 (Figure 4b). Furthermore, PSY significantly decreased luciferase expression of the heterologous promoter system, which is regulated by STAT3 transcriptional activation in HCT116 cells (Figure 4c). STAT3 is constitutively activated in various types of human cancers and is associated with adverse clinical outcomes and poor prognosis in human CRC [18,19]. Recent studies have shown that PR and MCR induce apoptosis by inhibiting STAT3 in human multiple myeloma and CRC cells, respectively [15,16,20]. These results suggested that PSY induces apoptosis by inhibiting STAT3 in human CRC cells. Considering that JAK2, Src, and protein tyrosine phosphatases (PTPs), including SHP1 and SHP2, have been associated with STAT3 activation [6], we further examined the upstream signaling molecules of STAT3. As shown in Figure 5a, PSY decreased the phosphorylation of JAK2 and Src in HCT116 cells, while it only reduced the phosphorylation of JAK2 in SW480 cells. PSY also increased the expression of SHP1, a negative regulator of the JAK-STAT pathway, but not SHP2 in CRC cells (Figure 5b). In addition, pervanadate reversed PSY-mediated inhibition of STAT3 in HCT116 cells (Figure 5c), indicating a crucial role of SHP1 in the PSY mechanism of action. Accordingly, these results suggest that PSY suppresses human CRC cell proliferation by inducing cell cycle arrest and apoptosis via inhibiting the STAT3 signaling pathway.

**Figure 3.** Effects of PSY on apoptosis in human CRC cells. CRC cells were treated with PSY EtOH-Ex for 48 h. (**a**) The cells were double-stained with FITC-Annexin V and PI and analyzed using flow cytometry (left panel). (**b**) The TUNEL and DAPI staining of the PSY (50 μg/mL)-treated cells were analyzed by confocal microscopy (left panel). Scale bar, 50 μm. Apoptotic cells were quantified as a percentage of Annexin V-positive and TUNEL-positive cells (right panel in **a**,**b**, respectively). Data present the mean ± SEM of three independent experiments (\*, *p* < 0.05 and \*\*\*, *p* < 0.001 versus control). (**c**) Whole lysates of the treated cells were subjected to Western blotting for PARP and cleaved PARP (c-PARP). β-actin was used as an internal control.

**Figure 4.** Effects of PSY on STAT3 in human CRC cells. (**a**) CRC cells were treated with PSY EtOH-Ex for 24 h, and cell lysates were subjected to Western blot analysis for STAT3 and p-STAT3. β-actin was used as an internal control. Data in the graphs represent the mean ± SD to that of the mock-treated cells taken as 1 (\*\*, *p* < 0.01 and \*\*\*, *p* < 0.001 versus control). (**b**) Immunofluorescence staining of p-STAT3 and DAPI in the PSY (50 μg/mL)-treated CRC cells were analyzed by confocal microscopy. Scale bar, 12.5 μm. (**c**) HCT116 cells transfected with pSTAT3-Luc plasmid were treated with PSY for 24 h to analyze the transcriptional activity of STAT3. Data in the graphs are presented as the mean of fold-normalized luciferase (Luc) activities ± SEM to that of the untreated cells taken as 100% (\*\*, *p* < 0.01 and \*\*\*, *p* < 0.001 versus control).

**Figure 5.** Effects of PSY on the STAT3 signaling pathway in human CRC cells. Whole lysates of CRC cells treated with PSY EtOH-Ex for 24 h were subjected to Western blot analysis for JAK2, p-JAK2, Src, and p-Src (**a**), and SHP1 and SHP2 (**b**). In addition, lysates of HCT116 cells treated with pervanadate (5 μM) and PSY (50 μg/mL) for 24 h were subjected to Western blot analysis for STAT3 and p-STAT3 (**c**). β-actin was used as an internal control. Data in graphs present the mean ± SD to that of the untreated cells taken as 1 (\*, *p* < 0.05 and \*\*, *p* < 0.01 versus control).

#### *2.4. Chemical Identification in PSY Extracts*

The chemical compositions of PSY EtOH-Ex and Water-Ex were characterized using UPLC-ESI-QTOF MS/MS in positive and negative ion modes. Representative base peak chromatograms (BPCs) of PSY Water-Ex and EtOH-Ex are shown in Figure 6a,b, and the identified minor or overlapping peaks on the BPCs are divided in the extracted-ion chromatograms (XICs) (Figure 6c,d). Detailed chemical and chromatographic information on the identified peaks (Figure 6) are summarized in Table 1, and their chemical profiles

revealed 38 compounds of PSY, 7 organic acids, 8 iridoids, 2 lignans, 12 prenylflavonoids, 8 fatty acids, and 1 carbohydrate.

**Figure 6.** Chemical identification in PSY extracts. Representative BPCs of the Water-Ex (**a**) and EtOH-Ex (**b**) and expanded XICs from the Water-Ex (**c**) and EtOH-Ex (**d**) are presented. The scale of Y axis for intensity (CPS) on the chromatograms indicate that 1.6e5 equals 1.6 <sup>×</sup> 105.

Quinic acid (peak **2**) was identified using molecular networking (MN) analysis through Global Natural Products Social Molecular Networking (GNPS), and its derivatives, neochlorogenic acid (**3**), chlorogenic acid (**5**), cryptochlorogenic acid (**6**), 1,4-dicaffeoylquinic acid (**15**), 1,3-dicaffeoylquinic acid (**16**), and 4,5-dicaffeoylquinic acid (**17**), were identified by comparison with the retention time and fragmentation patterns of the reference standard in both positive and negative ion modes. Peaks **2**, **5**, **15**, **16**, and **17** have been reported as constituents of *P. scabiosaefolia* and *Partinia* [21,22].

Eight iridoids were identified in the PSY extracts. Loganic acid (**4**) and loganin (**8**) were identified using the reference standard and MN analysis on the GNPS. The loganic acid yielded its quasi-molecular ion [M−H]<sup>−</sup> at *m*/*z* 375.1287 (mass error = −1.72), and its fragment ions were *m*/*z* 213.0762 [M−H−Glc]−, 169.0864 [M−H−Glc−CO2], 151.0757 [M−H−Glc−CO2−H2O], 113.0243 [M−H−Glc−C3O4], and 69.0369 (C5H9, isoprenyl moiety). The precursor ion at *m*/*z* 359.1344 was putatively identified as a deoxyloganic acid isomer (**9** and **11**) in MN analysis (Figure 7a). The mass difference (15.9943 Da, oxygen) of the precursor ion between loganic acid and deoxyloganic acid was found at *m*/*z* 197.0823, 135.0823, and 153.0915 from the fragment ions of **9** and **11**, respectively. The precursor ions of deoxyloganic acid were detected at peaks **9** (RT 11.1 min) and **11** (RT 12.9 min) on BPC and XIC (Figure 6). Although compounds **9** and **11** were identified as deoxyloganic acid isomers, they could not be clearly classified as 7-deoxyloganic acid or 8-epideoxyloganic acid. Geniposide (**7**) was putatively identified through MN analysis on the GNPS (Figure 7b). Patrinalloside (**10**) and patrinoside isomers (**13** and **14**) have been identified in the literature [23–26]. Some iridoids (**4**, **8**, **9**, **10**, and **11**) are representative components of *P. scabiosaefolia* [21,22].

**Figure 7.** *Cont*.

**Figure 7.** The representative MS/MS spectral network of identified phytochemical compounds in PSY extracts. Spectral nodes indicating identified compounds are noted with numbers (yellow nodes with peak numbers in Table 1). Annotated candidates for selected spectral nodes (yellow nodes without numbers) show compound name, GNPS library spectrum matching (in bold), and chemical class and subclass (in italic with blue). Molecular networks of representative chemical classes or biosynthetic pathways of natural products in PSY extracts. (**a**) iridoid class, (**b**) shikimate pathway including cinnamic (or quinic) acid derivatives, iridoid, furofuranoid lignan subclass in lignan, and stilbene class, (**c**) lignan class, (**d**) prenylflavone subclass in prenylflavonoid class, and (**e**) prenylisoflavanone, prenylflavanone, and prenylflavone subclasses in prenylflanoid class. Further information can be found at the GNPS job link (https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=e4406e69e619467b9 a347a905d6c383a accessed on 31 March 2022).

Eleutheroside E (**12**) and nortrachelogenin (**18**) were detected and putatively identified as lignans by matching the MS/MS fragment ions with the MN analysis on the GNPS (Figure 7b,c). Nortrachelogenin has been previously isolated from PR [27].

Twelve prenylflavonoids originating from MCR were detected, and **23** and **31** were identified as kuwanon G and morusin [28,29], representative components in MCR with each reference standard, respectively. Peak **23** is the Diels-Alder (DA)-type adduct of a chalcone and prenylflavone and exhibited an [M−H]<sup>−</sup> ion at *m*/*z* 691.2182. In negative ion mode, fragment ions were observed at *m*/*z* 581.1820 [M−H−resorcinol (C6H6O2, 110 Da)]−, 353.1029 [M−H−resorcinol−C14H12O3] <sup>−</sup>, 419.1501, 379.1189 [M-H-2resorcinol−C4H8O−H2O]−, and 539.1719 [M−H−resorcinol−C3H6] −. Morusin exhibited characteristic fragment ions at *m*/*z* 297.1134, 191.0716, 309.1139 [M−H−resorcinol]−, and 350.04804 [M−H−isoprenyl]−. The fragment ion at *m*/*z* 191.0716 lost one CO and received two hydrogens at C-10 from *m*/*z* 217.0507, produced by retro-DA cleavage [30]. Peak **25** was detected and putatively identified as Kuwanon C through MN analysis and clustered with **31** as prenylflavones (Figure 7d). The mass difference (2 Da, two hydrogens) between the precursor ions of peaks **31** and **25** was found in the characteristic fragmentation patterns (Table 1). Peaks **28** and **30** were putatively identified as Kuwanon E and Kuwanon T using PeakView 2.2, using exact mass and isotope patterns. Kuwanon E was further identified by fragmentation patterns in FooDB. Additionally, **28** and **30** (prenylflavones) were clustered by MN analysis, with two isoflavanones (not shown in Table 1) identified by library matching (Figure 7e). Mulberrofuran G, kuwanon G, kuwanon T, sanggenon F, and morusin, found in MCR, are known to exert anti-inflammatory and anti-cancer activities [15,31]. Kuwanon G also ameliorates lipopolysaccharide-induced disruption of the gut epithelial barrier [32,33]. In addition, morusin induces autophagy and apoptosis by regulating various signaling molecules, including STAT3, in several types of cancer cells [31,32].

Peaks **27**, **29**, and **32**–**37** were identified as fatty acids (Table 1), and linoleic (**34**) and oleic (**37**) acids are known to be present in CS [34]. Moreover, Linolenic, linoleic, and oleic acids are known to have various pharmacological activities, including anti-inflammatory, anti-cancer, and anti-microbial activities [35,36].

#### *2.5. Identification of Potential Bioactive Phytochemicals in PSY EtOH-Ex*

To investigate the potential bioactive candidates in EtOH-Ex, due to its potent anticancer activity compared to that of Water-Ex, the fold change in the peak area of every compound in EtOH-Ex versus Water-Ex was calculated and compared because the fold change reflects the degree of differences in chemical concentration. As shown in Figure 8, the comparison of ploidy area variation was significant for 21 phytochemicals (fold change > 2), including 12 prenylflavonoids, two iridoids, and seven fatty acids. All prenylflavonoids were significantly enriched in EtOH-Ex. Intriguingly, four fatty acids (**32**, **35**, **36**, and **37**) and one iridoid (**13**) were found only in the EtOH-Ex. These results suggest that the EtOH-Ex-enriched compounds could be potential bioactive molecules, indicating the PSY effect to suppress CRC cell proliferation, and could be a useful source for developing anti-cancer drugs. Additionally, other phytochemicals remain unknown. Since these candidates may also contribute to anti-cancer effects, further identification of novel compounds, evaluation of their anti-cancer effects, and further investigation of their mechanisms of action seem to be necessary.



*Int. J. Mol. Sci.* **2022**, *23*, 14826

**Table 1.** *Cont.*

**Figure 8.** Identification of PSY EtOH-Ex-enriched phytochemicals. Peak area fold change of 38 compounds from PSY EtOH-Ex and Water-Ex. Fold change > 0 means that the relative content of chemical compounds in EtOH-Ex is higher than Water-Ex. The fold changes in peak area were calculated using the formula log2 for the ratio of peak area values of the phytochemicals extracted by different solvents.

Although many early stage CRC patients undergo surgery such as colectomy or proctectomy without chemotherapy, most patients, especially those suffering from metastatic CRC require chemotherapy with or without surgery and radiotherapy [37]. Conventional combinatorial strategies, FOLFOX (folinic acid, 5-fluorouracil (5-FU), and oxaliplatin) or FOLFIRI (folinic acid, 5-FU, and irinotecan) have been the most appropriate first-line chemotherapy options for CRC patients since their development in the early 2000s [38,39]. The recent addition of molecular-targeted drugs such as cetuximab and encorafenib has been a major improvement with more effective therapeutic options [39]. These therapeutic

advances have resulted in clinically relevant survival improvements, but the five-year survival rate is still less than 15% for stage IV disease and most survivors suffer from severe side effects including neuropathy, bowel and bladder dysfunction, and sexual dysfunction [1,38,39]. Therefore, novel therapeutic strategies are still required for this fatal disease. We, thus, have focused on medicinal plants which have been traditionally used for thousands of years and are edible, expecting at least fewer adverse effects, and found the combination of PSY in the present study.

Altogether, these results demonstrate that PSY likely suppresses CRC cell proliferation by inducing cell cycle arrest and apoptosis via inhibiting oncogenic STAT3 signals. Further, we found 21 EtOH-Ex-enriched PSY phytochemicals, suggesting further studies investigating the potential anti-cancer activities of these compounds and their mechanisms of action involved in STAT3 (Figure 9). In conclusion, these results suggest that PSY could be a potentially effective therapeutic for patients with CRC. More research about the potential combination of PSY or an EtOH-Ex-enriched PSY phytochemical with any conventional drug including 5-FU, oxaliplatin, and irinotecan would be also interesting and facilitate the development of a more effective and therapeutic strategy with less adverse effects for CRC patients.

**Figure 9.** A schematic representation of the molecular mechanism by which PSY suppresses human CRC cell proliferation. CTK: cytokine, CTKR: cytokine receptor, GF: growth factor, GFR: growth factor receptor, PTP: protein tyrosine phosphatase, CRC: colorectal cancer.

#### **3. Materials and Methods**

#### *3.1. PSY Preparation*

Dried PR, MCR, and CS were purchased from Omniherb (Daegu, Republic of Korea), regulated in herbal Good Manufacturing Practices (hGMP), and each voucher specimen was deposited in the herbarium of the Korean Medicine Clinical Trial Center, Kyung Hee University Korean Medicine Hospital (Seoul, Republic of Korea). For EtOH-Ex preparation, 37.5 g of PR, 22.5 g of MCR, and 37.5 g of CS were extracted with 877.5 mL of ethanol by refluxing at room temperature sonication. Similarly, for the preparation of PSY Water-Ex, the amount of PR, MCR, and CS was extracted with 877.5 mL of distilled water by refluxing at 100 ◦C for 1.5 h. Each extract was filtered through vacuum filter paper and concentrated in a rotary evaporator (Eyela N-1200 system, Tokyo, Japan) at 40 ◦C under lower pressure. The extracts were lyophilized, and the yields of dried EtOH-Ex and Water-Ex were 3.21% and 10.13%, respectively. Each extract was dissolved in dimethyl sulfoxide (DMSO) to obtain a 500 mg/mL stock solution.

#### *3.2. Chemicals and Reagents*

Standard references, including neochlorogenic acid, loganic acid, cryptochlorogenic acid, loganin, kuwanon G, and 3-Oxo-olean-12-en-28-oic acid, were purchased from Chem-Faces (Wuhan, China). Linoleic and palmitic acids were obtained from Sigma-Aldrich (St. Louis, MO, USA). Morusin and chlorogenic acid were obtained from Biopurity Phytochemicals, Ltd. (Chengdu, China). Further, mixed references were prepared by dissolving the references in a 50% methanol in water (*v*/*v*) solution for qualitative analysis using UPLC-ESI-QTOF MS/MS.

#### *3.3. Cell Lines and Culture Condition*

HCT116, HT-29, and SW480 human CRC cell lines were acquired from the Korean Cell Line Bank (Seoul, Republic of Korea) and were cultured according to the manufacturer's instructions.
