Effect of Ephedra Herb on Erlotinib Resistance in c-Met-Overexpressing Non-Small-Cell Lung Cancer Cell Line, H1993, through Promotion of Endocytosis and Degradation of c-Met

Oriental Medicine Research Center of Kitasato University, 5-9-1 Shirokane, Minato-ku, Tokyo 108-8642, Japan National Institute of Health Sciences, 3-25-26 Tonomachi, Kawasaki-ku, Kawasaki, Kanagawa 210-9501, Japan Department of Pharmacognosy, College of Pharmaceutical Sciences, Matsuyama University, 4-2 Bunkyo-cho, Matsuyama, Ehime 790-8578, Japan Tokiwa Phytochemical Co., Ltd., 158 Kinoko, Sakura, Chiba 285-0801, Japan


Introduction
Lung cancer is the leading cause of cancer-related death in many countries around the world, including Japan. Lung cancer is categorized into small cell lung cancer and nonsmall-cell lung cancer (NSCLC). NSCLC is accountable for approximately 85% of lung cancers. EGFR-TKIs, epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors, are used to treat NSCLCs harboring EGFR-activating mutations. Although first-generation reversible EGFR-TKIs, including gefitinib and erlotinib, are highly effective in treating NSCLC cells harboring an EGFR mutation within codon 858 of exon 21 (L858 R mutation), acquired resistance to these treatments emerges after just a few years [1]. e T790 M mutation, situated in the ATP binding pocket of EGFR exon 20, was detected in up 50% of patients treated with the first-generation EGR-TKIs. Second-generation EGFR-TKIs such as afatinib bind irreversibly to the T790 M mutant EGFR receptor [1]. Recently, third-generation irreversible EGFR-TKIs, such as osimertinib, were developed, to target EGFR T790 M resistance. e third-generation of EGFR-TKIs were designed to covalently bind only to mutated forms of EGFR, thus lowering the incidence of EGFR-TKI-mediated toxicities at wild-type receptors [2]. Nonetheless, some patients have developed resistance to the firstto third-generation EGFR-TKIs, and these patients demonstrate a high level of c-Met amplification or c-Met protein overexpression/hyperactivation [1,3]. erefore, combination therapy with EGFR-TKIs and a c-Met inhibitor is required for patients with resistant NSCLC cells carrying c-Met amplification and/or protein hyperactivation.
Ephedra Herb is a crude drug and is used in Japan as a component in many Kampo formulae, including maoto, kakkonto, and shoseiryuto (http://mpdb.nibiohn.go.jp/ stork/). Ephedra Herb is defined in the sixteenth edition of the Japanese Pharmacopoeia (JP) as the terrestrial stem of Ephedra sinica Stapf., Ephedra intermedia Schrenk, and C.A. Meyer or Ephedra equisetina Bunge (Ephedraceae), which have stems with total alkaloids (ephedrine and pseudoephedrine) content greater than 0.7% [4]. We previously reported that Ephedra Herb extract (EHE) has c-Met inhibitory activity. EHE inhibits HGF-induced phosphorylation of c-Met by preventing c-Met tyrosine kinase activity [5][6][7]. us, we investigated the effect of combining EHE and an EGFR-TKI such as erlotinib on erlotinib resistance in NSCLC cells. In the present study, we used H1993 cells [8], an erlotinib resistance NSCLC cell line showing overexpression of c-Met.

Preparation of Ephedrine Alkaloids-Free Ephedra Herb Extract (EFE) and Four Fractions Fractionated from EFE.
EFE was prepared as described by Oshima et al. [9]. Ephedra Herb (200 g, E. sinica, Japanese pharmacopoeia grade) was added to water (2000 ml), extracted at 95°C for 1 h, and filtered, after which the residue was washed with water (200 ml). e extract was centrifuged at 1800 g for 10 min, and half of the supernatant was passed through 100 ml of DIAIONTM SK-1B ion-exchange resin (pretreated with 30 ml 1 M HCl/100 ml water and then washed with 100 ml water). e pH of the unadsorbed fraction (1100 ml) was adjusted to pH 5 using 5 % NaHCO 3 aq. (60 ml), and the solution was then evaporated under reduced pressure to obtain EFE (11.8 g

Cell Lines and Culture.
e human NSCLC cell line H1993 cells were obtained from the American Type Culture Collection (Manassas, VA, USA) and were cultured in RPMI 1640 medium (Sigma-Aldrich, Tokyo, Japan) containing 10% fetal calf serum (FCS) (Invitrogen Corp., Carlsbad, CA, USA) at 37°C in an atmosphere containing 5% CO 2 .

Animals.
Female 6-week-old BALB/c-nu/nu mice were obtained from Charles River Japan (Kanagawa, Japan). All mice were acclimatized for one week prior to the experiments. Animals were kept at constant temperature (25 ± 2°C) and humidity (50 ± 10%) under isolation in a specific pathogen-free (SPF) conditioning room with a 12 h light/12 h dark cycle. e protocol for animal experiments was approved by the Institutional Animal Care and Use Committee of Kitasato University. Experiments were performed in accordance with the Kitasato University guidelines for animal care, handling, and termination, which are in line with international and Japanese guidelines for animal care and welfare.

Xenograft
Model. H1993 cells (2.5 × 10 6 ) were inoculated subcutaneously into the left thigh of Balb/c-nu/nu mouse. After 7 days, the mice were randomly assigned into four groups (6 mice per group) and oral administration of the different drugs was started. In the erlotinib group, erlotinib dissolved in 0.5% methylcellulose 400 cp (FUJI-FILM Wako Pure Chemical Corporation, Osaka, Japan) and 0.05% Tween 80 (Nacalai Tesque, Kyoto, Japan) was orally administered to mice at 100 mg/kg/day. In the EHE group, EHE dissolved in pure water was orally administered to mice at 350 mg/kg/day. In the erlotinib and EHE group, erlotinib and EHE were orally administered to mice at 100 mg/kg/day and 350 mg/kg/day, respectively. In the vehicle group, 0.5% methylcellulose 400 cp and 0.05% Tween 80 were orally administered to mice. e weight of each mouse and the tumor size were measured twice a week. Tumor size (mm 3 ) � (length × width 2 )/2.
In a second set of experiments, H1993 cells (1 × 10 6 cells) were incubated in 4 ml of 10% FCS-RPMI-1640 for two days. After the cells were washed twice with 4 ml RPMI-1640, they were incubated for 4 h at 37°C in 4 ml RPMI-1640 or RPMI-1640 containing 25, 50, 100, and 200 µg/ml EHE. e cells were processed for western analysis as described above. e blots were incubated with antibodies against c-Met, pMet, EGFR, pEGFR, and GAPDH as described above.
e blots were incubated with antibodies against c-Met, pMet, EGFR, pEGFR, and GAPDH as described above.

2.8.
Clathrin-Mediated Endocytosis. H1993 cells (1 × 10 6 cells) were incubated in 4 ml of 10% FCS-RPMI-1640 for two days. After the cells were washed twice with 4 ml RPMI-1640, they were incubated for 5 min at 37°C in 4 ml RPMI-1640 or RPMI-1640 containing 10 and 20 µM Pitstop-2 (Abcam, Cambridge, UK). Next, 50 µg/ml EHE was added to the samples and the cells were incubated for 15 min. e cells were processed for western analysis as described above. e blots were incubated with antibodies against c-Met, pMet, EGFR, pEGFR, and GAPDH as described above.
e blots were incubated with antibodies against c-Met, pMet, EGFR, pEGFR, and GAPDH as described above.
2.13. Statistical Analysis. All data are expressed as mean-± standard deviation (S.D.). Data were analyzed by 1-way ANOVA and 2-way ANOVA. Significant differences between the control and treatment groups were determined by Dunnett's test using GraphPad Prism 7.03 (MDF Co., Ltd., Tokyo, Japan). Results of P < 0.05 were considered statistically significant.

Combination Effects of EHE and Erlotinib on Growth of H1993 Xenograft Tumor. First, we examined the combination effect of EHE and erlotinib on growth of the H1993 xenograft tumor in vivo.
In the vehicle group, tumor sizes increased in a timedependent manner. In mice administered with a combination of 350 mg/kg/day EHE and 100 mg/kg/day erlotinib, tumor sizes were significantly smaller than the tumor sizes of mice administered with vehicle between the 16 th and 22 nd days. In mice administered with 350 mg/kg/day EHE or 100 mg/kg/day erlotinib, tumor sizes were significantly smaller than the tumor sizes of mice administered with vehicle on the 22 nd day (Figure 1(a)). e weight of mice administered with 100 mg/kg/day erlotinib was significantly decreased, but administration of 350 mg/kg/day EHE had no effect on weight (Figure 1(b)). us, EHE had almost no toxicity by oral administration for 22 days. Mice administered with a combination of 350 mg/ kg/day EHE and 100 mg/kg/day erlotinib showed a tendency to recover their weight between the 20 th and the 22 nd days ( Figure 1(b)).
ese results show that a combination of EHE and erlotinib is more effective in inhibiting the growth of H1993 xenograft tumor (as compared with either EHE or erlotinib alone) and suggest that EHE contributes to a recovery from resistance to erlotinib in vivo.

Effects of the Combination of EHE and Erlotinib or the Combination of SU11274 and Erlotinib on H1993 Cells In
Vitro.
e combination effects of EHE and erlotinib were analyzed in detail in vitro.
e inhibitory effects of a combination of EHE (100, 200, and 400 µg/ml) and erlotinib (2.5, 5, and 10 µM) on the viability of H1993 cells were investigated. Depending on the EHE and erlotinib concentrations used, the viability of the cells was significantly decreased (Figure 2(a)). Moreover, the 50% inhibitory concentration (IC 50 ) of erlotinib decreased depending on the concentration of EHE (Table 1). ese results are consistent with the proposal that EHE contributes to the recovery from resistance to erlotinib. e expression levels of c-Met, phosphorylated c-Met (pMet), EGFR, and phosphorylated EGFR (pEGFR) in H1993 cells at 4 h after treatment of EHE, erlotinib, or the combination of EHE and erlotinib were analyzed by western blotting. e control lane shows that EGFR and c-Met were both overexpressed and autophosphorylated in H1993 cells (Figure 3(a)). Erlotinib (5 and 10 µM) had no effect on the expression levels of c-Met and pMet, and inhibited the phosphorylation of EGFR in a dose-dependent manner. Remarkably, erlotinib increased the expression level of EGFR (Figure 3(a)). EHE (200 and 400 µg/ml) both inhibited phosphorylation of c-Met and EGFR and decreased the expression levels of c-Met and EGFR. Likewise, EHE and erlotinib combination treatment both inhibited phosphorylation of c-Met and EGFR and decreased the expression levels of c-Met and EGFR (Figure 3(a)). Surprisingly, these results show that EHE decreased the upregulation of EGFR induced by erlotinib.
SU11274, a c-Met inhibitor, was used as a positive control, and its effect on erlotinib resistance in H1993 cells was investigated. e viability of H1993 cells was significantly decreased by SU11274 (1.25 and 2.5 µM) (Figure 2(b)). e combined effect of SU11274 and erlotinib on the cell viability was not observed. SU11274 (1.25 and 2.5 µM) inhibited phosphorylation of c-Met and EGFR in a dose-dependent manner, and increased the expression level of EGFR (Figure 3(b)). Likewise, SU11274 and erlotinib combination treatment inhibited phosphorylation of c-Met and EGFR, and increased the expression levels of EGFR (Figure 3(b)).
ese results indicate that the pharmacological action of EHE is different from the action of typical tyrosine kinase inhibitors such as erlotinib and SU11274. Moreover, EHE not only inhibits the phosphorylation of c-Met and EGFR, Evidence-Based Complementary and Alternative Medicine but also induces downregulation of the overexpression of these receptors.

Overexpression of c-Met and EGFR Was Downregulated by EHE.
We investigated the effects of EHE on overexpression of c-Met and EGFR in detail. e expression levels of c-Met, pMet, EGFR, and pEGFR in H1993 cells at 4 h after treatment with various concentrations of EHE (25-200 µg/ml) were analyzed by western blotting. EHE downregulated the expression levels of c-Met, pMet, EGFR, and pEGFR depending on its concentration (Figure 4(a)). e expression levels of c-Met and pMet were almost completely downregulated by 100 and 200 µg/ml EHE, while the expression levels of EGFR and pEGFR were strongly downregulated by 200 µg/ml EHE, indicating that EHE acts on c-Met at a lower concentration than on EGFR. e expression levels of c-Met, pMet, EGFR, and pEGFR in H1993 cells from 0 to 24 h after treatment with 100 µg/ml EHE were followed by western blotting. e expression levels of these receptors decreased in a time-dependent manner from 0.25 h to 8 h after treatment with 100 µg/ml EHE, though a tendency for expression levels to recover by 24 h was observed (Figure 4(b)). e downregulation activity of EHE appeared at only 15 min after addition of EHE to H1993 cells, suggesting that EHE does not affect gene expression of c-Met and EGFR. e effect of EHE was reversible.    Evidence-Based Complementary and Alternative Medicine 2, it is possible that this was due to a nonspecific action of Pitstop-2. Pitstop-2 had no effect on the expression level of EGFR and pEGFR in H1993 cells treated with EHE (Supplemental Figure 2). e effect of methyl-β-cyclodextrin, an inhibitor of caveolae-mediated endocytosis [11], on the EHE-induced downregulation of c-Met and EGFR, was investigated. Methyl-β-cyclodextrin had no effect on the expression levels of c-Met, pMet, EGFR, and pEGFR in H1993 cells treated EHE ( Figure 6 and Supplemental Figure 3). ese results suggest that the EHE had no effect on the caveolae-mediated endocytosis.

Promotion of Lysosomal Degradation of c-Met by EHE.
EHE may promote lysosomal degradation of c-Met and EGFR in H1993 cells. To test this proposal, we examined whether bafilomycin A1, an inhibitor of protein degradation in lysosome [12], suppresses the EHE-promoted downregulation of c-Met and EGFR. e expression level of c-Met in H1993 cells treated with 50 µg/ml EHE was increased by bafilomycin A1 in a concentration-dependent manner. e expression level of c-Met in cells treated with both 100 µg/ml bafilomycin A1 and 50 µg/ml EHE was significantly increased compared to cells treated only with 50 µg/ml EHE (Figures 8(a) and 8(c)). Bafilomycin A1 had no effects on the expression level of c-Met in untreated H1993 cells. ese results suggest that EHE promotes downregulation of c-Met by stimulating lysosomal degradation. In contrast, bafilomycin A1 had no effects on the expression level of pMet in H1993 cells treated with 50 µg/ml EHE or in untreated H1993 cells (Figures 7(a)  and 7(b)). Moreover, bafilomycin A1 had no effect on the expression levels of EGFR and pEGFR in H1993 cells treated with EHE or in untreated H1993 cells (Supplemental Figure 4).

Effects of EHE on HGF-c-Met Downstream Signaling
Molecules. Next, we investigated the effects of EHE on HGFc-Met downstream signaling molecules (Akt, ERK, GRB2, and Gab1) in H1993 cells. First, the effect of EHE on the expression of pMet and c-Met in the presence of 50 ng/ml HGF was examined. EHE decreased the expression levels of pMet and c-Met in a concentration-dependent manner (Figure 8(a)). e expression levels of Akt, ERK, GRB2, and Gab1, from 0 to 6 h after treatment with 50 µg/ml EHE and 50 ng/ml HGF, were analyzed by western blotting. e expression levels of c-Met, pMet, and Gab1 in H1993 cells treated with 50 ng/ml HGF and 50 µg/ml EHE were decreased in a time-dependent manner, while the expression levels of c-Met, pMet, and Gab1 in H1993 cells treated only with 50 ng/ ml HGF (control) were essentially unchanged (Figure 8(b)). While Akt was phosphorylated after 1 h to 6 h treatment with 50 ng/ml HGF (control), the phosphorylation level of Akt decreased in a time-dependent manner in cells treated with 50 µg/ml EHE and 50 ng/ml HGF. ERK was phosphorylated after 1 h to 2 h treatment with 50 ng/ml HGF (control), though the phosphorylation level subsequently decreased. In contrast, the phosphorylation levels of ERK in cells after 1 h to 2 h treatment with 50 µg/ml EHE and 50 ng/ml HGF were lower than those of control (Figure 8(b)). ese results suggest that EHE suppressed the HGF-c-Met downstream signal via downregulation of c-Met, pMet, and Gab1.

Effects of EFE and EFE Fractions on the Expression of c-Met, pMet, EGFR, and pEGFR.
e effects of EFE and the four EFE fractions (Water Fr, 20% MeOH Fr, 40% MeOH Fr, and MeOH Fr) on the expression levels of c-Met, pMet, EGFR, and pEGFR were examined. e expression levels of these receptors at 4 h after treatment of EFE or each EFE fraction were analyzed by western blotting. EFE (100 µg/ml) decreased the expression levels of c-Met, pMet, EGFR, and pEGFR, and components other than ephedrine alkaloids in EHE may contribute to these effects (Figure 9). Water Fr (50 µg/ml) had no effect on the expression levels of these receptors. 20% MeOH Fr (25 µg/ml) and MeOH Fr (25 µg/ ml) decreased the expression levels of c-Met, pMet, and pEGFR and slightly decreased the expression level of EGFR. In contrast, 40% MeOH Fr (25 µg/ml) decreased the expression levels of c-Met, pMet, EGFR, and pEGFR to almost the same level as observed with 100 µg/ml EFE (Figure 9), showing that the active ingredients are enriched in this fraction.

Discussion
Erlotinib resistance in H1993 cells is thought to be the leading cause of c-Met overexpression (Supplemental Figure 1) [8]. EHE was expected to contribute to recovery from erlotinib resistance in these cells, because it inhibits the phosphorylation of c-Met [5]. We initially evaluated the combination effect of oral administration of EHE and erlotinib on growth of a H1993 xenograft tumor. It was important to investigate the effect of oral administration of EHE, because EHE is a component of Kampo formula, and only an oral route of administration of this traditional medicine is observed in Japan. A combination of EHE and erlotinib proved to be a more effective treatment for suppression of the growth of the H1993 xenograft tumor than either EHE or erlotinib alone (Figure 1(a)). Furthermore, the oral administration of EHE for 22 days showed little toxicity (Figure 1(b)). ese results suggest that EHE is effective in promoting recovery from resistance to erlotinib. Next, we compared the combination effect of EHE and erlotinib on the growth of H1993 cells with the combination effect of a c-Met inhibitor, SU11274, and erlotinib in vitro.   Table 1), suggesting that EHE contributes to the recovery from resistance to erlotinib. SU11274 (1.25-2.5 µM) suppressed the viability of H1993 cells and this c-Met inhibitor did not show a combination effect with erlotinib (Figure 2(b)). Moreover, because the c-Met inhibitory effect of 1.25-2.5 µM SU11274 was weak ( Figure 3(b)), the antiproliferative effect of SU11274 was likely to be independent of this activity.
Unexpectedly, EHE not only inhibited the phosphorylation of c-Met, but also downregulated the expression of c-Met. In addition, EHE inhibited the phosphorylation of EGFR and downregulated the expression of EGFR (Figure 3(a)). In contrast, erlotinib induced the overexpression of EGFR (Figures 3(a) and 3(b)). Generally, growth factor receptors are downregulated after activation by phosphorylation [13]. It is possible that downregulation of EGFR was arrested via inhibition of EGFR phosphorylation by erlotinib and that EGFR accumulation occurred as a consequence. Surprisingly, EHE downregulated the overexpression of EGFR induced by erlotinib (Figure 3(a)). SU11274 did not downregulate the expression of c-Met and EGFR, and it induced overexpression of EGFR (Figure 3(b)).
ese results indicate that EHE activity is different from that of a typical c-Met tyrosine kinase inhibitor such as SU11274.
EHE decreased the expression and phosphorylation levels of c-Met and EGFR in a concentration-dependent manner (Figure 4(a)). Moreover, the expression and phosphorylation levels of c-Met and EGFR were decreased in a time-dependent manner from 0 h to 8 h after treatment of EHE, though the expression and phosphorylation levels showed a tendency to recover after 24 h (Figure 4(b)). ese  results reveal that EHE promotes downregulation of nonphosphorylated c-Met and EGFR and that these effects are reversible.
Previously, we reported that EHE stimulates HGF-induced MET and p-MET endocytosis by the early/late endocytic pathways [14]. We investigated the effect of EHE on endocytosis of these receptors using inhibitors such as Pitstop2 and methyl-β-cyclodextrin. According to the results, EHE facilitated the clathrin-mediated endocytosis of c-Met (Figures 5 and 6). In line with these results, c-Met was previously reported to be internalized by clathrin-dependent endocytosis [15].
Using a lysosomal enzyme inhibitor, bafilomycin A1, EHE was also shown to promote the lysosomal degradation of c-Met. It is significant that EHE not only facilitates the clathrin-mediated endocytosis of c-Met, but also promotes the lysosomal degradation of c-Met. If EHE only induced the clathrin-mediated endocytosis of c-Met, it may not suppress c-Met signaling as effectively. Joffre et al. reported that c-Met mutants increased endocytosis/recycling activity and decreased levels of degradation, leading to c-Met accumulation in endosomes. e endosome represents a platform for integrating signaling pathways both temporally and spatially [16]. It should be noted that the expression level of pMet was almost unchanged after treatment with Pitstop2 or bafilomycin A1 (Figures 5−7), suggesting that the observed reduction in the expression level of pMet was affected via inhibition of tyrosine phosphorylation by EHE. e expression level of EGFR was almost unchanged after treatment with these inhibitors (Supplemental Figures 2−4), and it is unclear how EHE promotes the downregulation of EGFR. Because these inhibitors show cytotoxicity, it was necessary to shorten their incubation times with H1993 cells. e expression level of c-Met was downregulated at 15 min after treatment with EHE. However, the expression level of EGFR was unchanged at 15 to 30 min, and EGFR downregulation was only observed after 2 h of treatment (Figure 4(b)). erefore, it was difficult to reveal the mechanism of action of EHE on EGFR using these inhibitors. EGFR was previously reported to be downregulated via clathrin-mediated endocytosis [13,17]. It is possible that EHE facilities clathrin-mediated endocytosis of EGFR. In the future, we plan to investigate the effect of EHE on the expression of EGFR using clathrin knockdown cells.
Next, we analyzed the effects of EHE on HGF-c-Met downstream signaling molecules. EHE decreased the expression levels of pMet, c-Met, pAkt, pERK, and Gab1. e expression level of Gab1 was correlated with the expression levels of pMet and c-Met. Binding of HGF leads to dimerization of c-Met, autophosphorylation of tyrosine residues (Y1234 and Y1235) in the kinase domain, and transautophosphorylation of tyrosine residues (Y1349 and Y1356) in the C-terminal region of c-Met. Gab1 binds to activated c-Met on either Y1349 or Y1356 through its specific c-Met binding site or indirectly through the adaptor Grb2. c-Met activation leads to downstream activation of multiple signaling molecules, including Akt and ERK, through the adaptors Gab1 and Grb2 [15]. It is possible that HGF-c-Met signaling was suppressed via downregulation of pMet, c-Met, and Gab1 by EHE. It is also possible that Gab1 bound to activated c-Met is digested along with c-Met in the lysosome.
We previously demonstrated that the c-Met inhibitory activity of EHE is independent of ephedrine alkaloids, which are the principle components of EHE. Indeed, EFE inhibited the HGF-induced phosphorylation of c-Met at the same level as EHE [7]. In the present study, EFE promoted the downregulation of c-Met, pMet, EGFR, and pEGFR. Moreover, the 40% MeOH Fr fractionated from EFE was equally as effective at promoting the downregulation of these receptors (Figure 9). Recently, we have shown that the 40% MeOH Fr contained mainly high-molecular mass condensed tannins. ese tannins were procyanidin B-type and partly procyanidin A-type, including pyrogallol-and catechol-type flavan-3-ols as extension and terminal units, and the weightaverage molecular weight of the fraction was >45,000 [10]. We hypothesize that the high-molecular mass condensed tannins play an important role in the activity of EHE and EFE. e high-molecular mass condensed tannins may bind to the extracellular domains of c-Met and EGFR, and these receptors may be cross-linked by the tannins and thus downregulated ( Figure 10).
Sym004 is a novel synergistic anti-EGFR Ab mixture. e two Abs in Sym004 are directed against distinct epitopes on the EGFR extracellular domain III [18], which is the EGF biding site [19]. Sym004 stimulated EGFR internalization and degradation in cancer cells. EGFR internalization is thought to be induced by receptor cross-linking with Sym004 [18]. It is possible that EHE and EFE act on c-Met and EGFR in a similar manner to Sym004. However, EHE and EFE differ from Sym004 in inducing the downregulation of not only EGFR but also c-Met. Furthermore, EHE-induced downregulation of c-Met was faster than that of EGFR. e downregulation of c-Met started at 15 min after treatment with EHE, but the downregulation of EGFR only started at 2 h after treatment with EHE (Figure 4(b)). ese results suggest that EHE shows specificity towards c-Met over EGFR. In the future, we plan to compare the binding affinity between EHE and c-Met with the binding affinity between EHE and EGFR using the Biacore surface plasmon resonance detector.
EHE is applicable to the treatment of c-Met-overexpressing NSCLC with resistance to EGFR-TKIs. EFE may be better suited for clinical treatment than EHE, because ephedrine alkaloids may cause side-effects. Recently, we reported that EFE treatment has none of the adverse sideeffects, such as excitation, insomnia, and arrhythmia, observed with ephedrine alkaloids [20], and we revealed the clinical safety of EFE in a double-blind, randomized, crossover comparative study [21]. In the future, we hope to obtain licensing approval for the therapeutic use of EFE. Because the tannins are unstable, isolated high-molecular mass condensed tannins in EFE are unsuitable for the development of drugs. However, different components in EFE may protect the high-molecular mass condensed tannins in this preparation.

Data Availability
(1) Previously reported preparation and characterization of EFE data were used to support this study and are available at DOI: 10.1007/s11418-016-0977-1 and DOI: 10.1007/s11418-016-0979-z. ese prior studies are cited at relevant places within the text as [7,9]. (2) e preparation and characterization of fractions from EFE used to support the findings of this study are available at M. Yoshimura et al., "Quality Evaluation and Characterization of Fractions with Biological Activity from Ephedra Herb Extract and Ephedrine Alkaloids-Free Ephedra Herb Extract," Chem. Pharm. Bull., 68, 140-149, 2020. e paper is cited at relevant places within the text as [10]. (3) e effect of erlotinib on the viability of erlotinib-resistant H1993 cells used to support the findings of this study is included in Supplemental Figure 1 within the Supplementary information file. (4) Expression of EGFR and pEGFR in H1993 cells after treatment with Pitstop-2 and EHE used to support the findings of this study is included in Supplemental Figure 2 within the Supplementary information file. (5) Expression of EGFR and pEGFR in H1993 cells after treatment with methyl-β-cyclodextrin and EHE used to support the findings of this study is included in Supplemental Figure 3 within the Supplementary information file. (6) Expression of EGFR and pEGFR in H1993 cells after treatment with bafilomycin A1 and EHE used to support the findings of this study is included in Supplemental Figure 4 within the Supplementary information file.
Ethical Approval e protocol for animal experiments was approved by the Institutional Animal Care and Use Committee of Kitasato University. Experiments were performed in accordance with the Kitasato University guidelines for animal care, handling, and termination, which are in line with international and Japanese guidelines for animal care and welfare.