Met Kinase Inhibitor E7050 Reverses Three Different Mechanisms of Hepatocyte Growth Factor–Induced Tyrosine Kinase Inhibitor Resistance in EGFR Mutant Lung Cancer Free

The reversible epidermal growth factor receptor (EGFR)–tyrosine kinase inhibitors (TKI) gefitinib and erlotinib show dramatic therapeutic efficacy in patients with EGFR-activating mutations, such as in-frame deletions of exon 19 and the L858 point mutation in exon 21 (1, 2). Recent clinical trials have shown that these TKIs induced much higher response rates and longer progression-free survival than standard first-line cytotoxic chemotherapy in patients with EGFR mutant lung cancer (3, 4). Almost all patients, however, develop acquired resistance to EGFR-TKIs after varying periods of time (5). In addition, 20% to 30% of patients with EGFR-activating mutations show intrinsic resistance to EGFR-TKIs (5). Therefore, intrinsic and acquired resistances to EGFR-TKIs are major problems in the management of EGFR mutant lung cancer.

Three clinically relevant mechanisms have been reported to induce acquired resistance to EGFR-TKIs in EGFR mutant lung cancer—EGFR T790M secondary mutation (6, 7), Met gene amplification (8), and hepatocyte growth factor (HGF) overexpression (9). We found that HGF overexpression is involved not only in acquired but in intrinsic resistance to EGFR-TKIs (9). HGF has been shown to play at least 3 important roles in EGFR-TKI resistance in EGFR mutant lung cancer. First, HGF induces resistance to the reversible EGFR-TKIs gefitinib and erlotinib by restoring MetGab1/PI3K/Akt pathways (9, 10). Second, continuous exposure to HGF accelerates the expansion of preexisting Met-amplified cancer cells and facilitates Met amplification-mediated resistance during EGFR-TKI treatment (10). Third, after lung cancer cells acquire resistance to reversible EGFR-TKIs, HGF induces the resistance of cells with T790M secondary mutation to irreversible EGFR-TKIs (11). These findings indicate that HGF is an ideal target for overcoming EGFR-TKI resistance in EGFR mutant lung cancer.

There are several possible strategies for inhibiting HGF-Met signaling, including anti-HGF neutralizing antibody, HGF antagonist (NK4), Met tyrosine kinase inhibitors, and inhibitors of downstream molecules, such as phosphoinositide 3-kinase (PI3K), Akt, and mTOR (12). Previously, we showed that anti-HGF antibody (13), NK4 (13), and PI3K inhibitors (14) were effective in overcoming HGF-induced gefitinib resistance. Many Met-TKIs have therefore been developed and are expected to reverse HGF-induced resistance to EGFR-TKIs (10, 15).

E7050 is an orally active Met-TKI (16) that has been shown to inhibit the phosphorylation of Met, including amplified Met, and to suppress the growth of several types of cancer cells with Met amplification. On the basis of favorable preclinical data, a phase I clinical trial of E7050 is currently in progress. We have assessed whether E7050 can overcome the 3 HGF-induced resistance mechanisms to EGFR-TKIs.

The EGFR mutant human lung adenocarcinoma cell lines PC-9 and HCC827 were purchased from Immuno-Biological Laboratories Co. and the American Type Culture Collection, respectively. The human embryonic lung fibroblast cell line MRC-5 was purchased from Health Science Research Resources Bank. MRC-5 (P 30–35) cells were maintained in Dulbecco's modified Eagle's medium with 10% FBS. PC-9 and HCC827 cells were maintained in RPMI-1640 medium with 10% FBS.

E7050 was synthesized by Eisai Co., Ltd (16). Gefitinib was obtained from AstraZeneca. The irreversible EGFR-TKI, BIBW2992, and the mutant-selective EGFR-TKI, WZ4002, were purchased from Selleck. Recombinant HGF and anti-human HGF antibody were prepared as described (17).

Cell growth was measured using the MTT dye reduction method (18). Tumor cells were plated at a density of 2 × 10³ cells/100 μL/well into 96-well plates in RPMI-1640 medium with 10% FBS. After 24-hour incubation, various reagents were added to each well, and the cells incubated for a further 72 hours, followed by the addition of 50 μL of MTT solution (2 mg/mL; Sigma) to each well and further incubation for 2 hours. The media containing MTT solution were removed, and the dark blue crystals were dissolved by adding 100 μL of dimethyl sulfoxide. The absorbance of each well was measured with a microplate reader at test and reference wavelengths of 550 and 630 nm, respectively. The percentage of growth is shown relative to untreated controls. Each reagent and concentration was tested at least in triplicate during each experiment, and each experiment was conducted at least 3 times.

Cells were lysed in cell lysis buffer containing phosphatase and proteinase inhibitor cocktails (Sigma), and protein concentrations were determined using a BCA Protein Assay Kit (Pierce Biotechnology). For the detection of phosphorylated Met in subcutaneous tumors, 10 mg tumor lysates were immunoprecipitated with anti-Met (25H2) antibody. Total protein (40 μg per lane) was resolved by SDS-PAGE, and the proteins were transferred onto polyvinylidene difluoride membranes (Bio-Rad). After washing 4 times, the membranes were incubated with Blocking One (Nacalai Tesque Inc.) for 1 hour at room temperature, followed by overnight incubation at 4°C with primary antibodies to Met (25H2), phospho-Met (Y1234/Y1235; 3D7), phospho EGFR (Y1068), ErbB3 (1B2), phospho-ErbB3 (Tyr1289; 21D3), Gab1 (#3232), phospho-Gab1 (Y627; C32H2), Akt, and phospho-Akt (Ser473; 736E11; 1:1,000 each; Cell Signaling Technology); and anti-human EGFR (1 μg/mL) antibody (R&D Systems). After washing 3 times, the membranes were incubated for 1 hour at room temperature with species-specific horseradish peroxidase–conjugated secondary antibodies. Immunoreactive bands were visualized using SuperSignal West Dura Extended Duration Substrate Enhanced Chemiluminescent Substrate (Pierce Biotechnology). Each experiment was conducted at least 3 times independently.

Cells (2 × 10⁵) were cultured in RPMI-1640 medium with 10% FBS for 24 hours, washed with PBS, and incubated for 48 hours in 2 mL of the same medium. The culture medium was harvested and centrifuged, and the supernatant was stored at −70°C until analysis. HGF concentrations were measured by IMMUNIS HGF EIA (Institute of Immunology, Tokyo, Japan), with a detection limit of 100 pg/mL, according to the manufacturer's instructions. All culture supernatants were tested in duplicate. Color intensity was measured at 450 nm using a spectrophotometric plate reader. Growth factor concentrations were determined by comparison with standard curves.

One day before transfection, aliquots of 1 × 10⁵ HCC827 cells in 1 mL of antibiotic-free medium were plated on 6-well plates. Full-length HGF cDNA cloned into the BCMGSneo expression vector (19) was transfected using Lipofectamine 2000 in accordance with the manufacturer's instructions. After 24-hour incubation, the cells were washed with PBS and incubated for an additional 72 hours in antibiotic-containing medium, followed by selection in G418 sulfate (Calbiochem). After limiting dilution, HGF-producing cells, HCC827/HGF, were established. HGF production by HCC827/HGF cells was confirmed by ELISA.

Duplexed Stealth RNAi (Invitrogen) against MET, ErbB3, and Gab1, and Stealth RNAi Negative Control Low GC Duplex #3 (Invitrogen) were used for RNA interference assays. One day before transfection, aliquots of 2 × 10⁴ tumor cells in 400 μL of antibiotic-free medium were plated on 24-well plates. After incubation for 24 hours, the cells were transfected with siRNA (50 pmol) or scrambled RNA using Lipofectamine 2000 (1 μL) in accordance with the manufacturer's instructions. After 24-hour incubation, the cells were washed with PBS and incubated with or without various reagents for an additional 72 hours in antibiotic-containing medium. Cell growth was measured using a Cell Counting Kit-8 (Dojin) in accordance with the manufacturer's instructions. Knockdown of MET, ErbB3, Gab1, and, Shc1 was confirmed by Western blotting. Each reagent and concentration was tested at least in triplicate during each experiment, and each experiment was conducted at least 3 times.

Cell block sections (4-μm thick) were subjected to dual-color FISH using a MET/CEP7 probe cocktail (Kreatech Diagnostics) according to the manufacturer's instructions. Staining was evaluated as described (20).

Suspensions of PC-9 cells (5 × 10⁶) mixed with MRC-5 cells (5 × 10⁶) were injected subcutaneously into the backs of 5-week-old female severe combined immunodeficient (SCID) mice (Clea), as described (13). After 4 days (tumor diameter >5 mm), mice were randomly allocated into groups of 6 animals, each to receive E7050 (50 mg/kg/d) and/or gefitinib (25 mg/kg/d) by oral gavage. Tumor volume was calculated as mm³ = width² × length/2. All animal experiments were carried out in compliance with the Guidelines for the Institute for Experimental Animals, Kanazawa University Advanced Science Research Center (Approval number: AP-081088).

Frozen sections (5-μm thick) of xenograft tumors were fixed with cold acetone and washed with PBS. After blocking endogenous peroxidase activity with 3% aqueous H₂O₂ solution for 10 minutes, the sections were incubated with 5% normal horse serum, followed by overnight incubation at 4°C with anti–phospho-Akt antibody (Ser473; 736E11, 1:100 dilution). The sections were washed with PBS, incubated with biotin-conjugated anti-rabbit IgG (1:200 dilution) for 30 minutes at room temperature, and incubated for 30 minutes with avidin–biotin–peroxidase complex (ABC) using a Vectastain ABC Kit (Vector Laboratories). Staining was detected using the DAB (3,3′-diaminobenzidine tetrahydrochloride) Liquid System (DakoCytomation). Samples from which primary antibodies had been omitted served as negative controls.

Between-group differences were analyzed by one-way ANOVA, with P values less than 0.05 for overall comparisons tested by post hoc pairwise comparisons using the Newman–Keuls multiple comparison test. All statistical analyses were carried out using GraphPad Prism Ver. 4.01 (GraphPad Software, Inc.).

PC-9 and HCC827 cells were highly sensitive to gefitinib (Fig. 1A), whereas exogenously added HGF induced resistance to gefitinib in both cell lines (9, 13, 14). Although E7050 did not affect the growth of PC-9 or HCC827 cells at concentrations less than 3 μmol/L, the combination of E7050 with gefitinib reversed HGF-induced resistance of both cell lines in a concentration-dependent manner (Fig. 1B).

We previously reported that stromal fibroblasts are a source of exogenous HGF for EGFR-TKI naive non–small cell lung carcinoma (NSCLC) and that fibroblast-derived HGF induces resistance to gefitinib and erlotinib in PC-9 and HCC827 cells (13). Although E7050 had no effect on the growth or production of HGF or VEGF by MRC-5 cells (HGF-high producing fibroblasts) or PC-9 cells (data not shown), it reversed the gefitinib resistance of PC-9 cells induced by coculturing with MRC-5 cells (Fig. 1C), indicating that E7050 can reverse the EGFR-TKI resistance induced by exogenous HGF in vitro.

We have shown that HGF is present in tumor cells of NSCLC patients with acquired resistance to EGFR-TKIs, and that transient HGF gene transfection into PC-9 cells resulted in resistance to EGFR-TKIs (9). We therefore generated a stable HGF gene transfectant in HCC827 cells (HCC827/HGF) and assessed the effects of continuously produced endogenous HGF. HCC827/HGF, but not HCC827 or the vector control HCC827/Vec, cells secreted high levels of HGF and became resistant to gefitinib (Fig. 2A and B). Anti-HGF antibody reversed the gefitinib resistance of HCC827/HGF cells (Supplementary Fig. S1), indicating that endogenously produced HGF induced gefitinib resistance in this cell line. Although the combination of E7050 plus gefitinib successfully reversed the resistance of HCC827/HGF cells, E7050 alone did not inhibit the proliferation of HCC827/HGF cells (Fig. 2B).

Using Western blotting, we examined the effects of E7050 on signal transduction in HCC827/Vec and HCC827/HGF cells. We found that gefitinib inhibited the phosphorylation of EGFR and ErbB3 in HCC827/Vec cells, thereby inhibiting the phosphorylation of Akt and ERK1/2. However, gefitinib failed to inhibit phosphorylation of Akt in the presence of HGF. E7050 suppressed the constitutive phosphorylation of Met, but not of EGFR, ErbB3, and downstream Akt and ERK1/2. Whereas HGF stimulated the phosphorylation of Met, E7050 plus gefitinib inhibited this HGF-induced Met phosphorylation and strongly suppressed the phosphorylation of Gab1, Akt, and ERK1/2 (Fig. 2C).

The amount of Met protein was decreased in HCC827/HGF cells, compared with HCC827/Vec cells. This could be a result of Met downregulation by persistent HGF stimulation, as also observed in a previous report (21). In contrast, the degree of Met phosphorylation was higher in HCC827/HGF than in HCC827/Vec cells. Gefitinib inhibited the phosphorylation of EGFR and ErbB3, but not of Akt in HCC827/HGF cells. The combination of E7050 and gefitinib inhibited the phosphorylation of both Met and Akt (Fig. 2C). These results suggested that E7050 reversed HGF-induced gefitinib resistance by inhibiting the Met/Gab1/PI3K/Akt pathway.

To confirm that the E7050 reversal of gefitinib resistance in HCC827/HGF cells was due to the inhibition of Met/Gab1, we transfected cells with siRNA specific for Met or Gab1. Transfection of ErbB3, Met, or Gab1 siRNA successfully knocked down the expression of the corresponding protein (Fig. 2D). Although scrambled or ErbB3 siRNA did not reverse the gefitinib resistance of HCC827/HGF cells, siRNAs for Met and Gab1 sensitized these cells to gefitinib (Fig. 2D), indicating that E7050 reverses gefitinib resistance in HCC827/HGF cells by inhibiting the Met/Gab1 pathway.

Next-generation EGFR-TKIs, irreversible TKIs (22–24), and mutant EGFR-selective TKIs (25) have been developed to treat gefitinib-resistant tumors caused by the EGFR T790M secondary mutation. H1975 cells with the EGFR mutations L858R and T790M mutations were resistant to reversible EGFR-TKIs, gefitinib, and erlotinib (data not shown), but were sensitive to BIBW2992, an irreversible EGFR-TKI, and WZ4002, a mutant-selective EGFR-TKI (Fig. 3). HGF markedly induced resistance to BIBW2992 and WZ4002, whereas E7050 efficiently reversed the HGF-induced resistance to both BIBW2992 and WZ4002. These results indicated that E7050 can overcome HGF-induced resistance not only to gefitinib but to next-generation EGFR-TKIs, including irreversible and mutant-selective EGFR-TKIs.

As HGF has been reported to accelerate the expansion of preexisting Met-amplified HCC827 cells and to facilitate Met amplification–mediated resistance during EGFR-TKI treatment (10), we examined the effects of E7050 on these phenomena. Although HCC827 cells did not produce viable colonies after 30 days of continuous exposure to gefitinib alone (Fig. 4A and B), these cells produced many colonies after exposure to both HGF and gefitinib. In contrast to previous findings (10), the percentage of cells with Met amplification was not increased when compared with parental HCC827 cells. The reason for this discrepancy remains unclear. Western blot analyses revealed that although the resultant cells expressed the same level of Met and Gab1 proteins compared with parental HCC827 cells, they expressed much higher levels of phosphorylated Met and Gab1 (Supplementary Fig. S2).

Importantly, E7050 prevented the emergence of viable clones even under conditions of continuous exposure to gefitinib and HGF (Fig. 4B). These results suggested the potential of E7050 to abrogate the effects resulting from continuous exposure to HGF.

To investigate the therapeutic efficacy of E7050 in vivo, we used the gefitinib resistance model previously described (13). We mixed PC-9 cells with the HGF-high producing fibroblast cell line, MRC-5, and inoculated SCID mice subcutaneously with this mixture. Oral treatment with gefitinib and/or E7050 was started after the establishment of solid tumors on day 4. Consistent with previous observations, we found that treatment with gefitinib alone prevented the enlargement of tumors produced by the mixture of PC-9 and MRC-5 cells, but did not cause tumor regression. As gefitinib induces shrinkage of PC-9 tumors (13, 14), our results suggested that MRC-5 cells induced gefitinib resistance in vivo. Under these experimental conditions, treatment with E7050 alone did not inhibit tumor growth, whereas the combination of E7050 and gefitinib induced marked tumor regression (Fig. 5A and B).

To confirm that E7050 inhibits Met/PI3K/Akt signaling in vivo, we assessed expression of phosphorylated Met and Akt in the xenograft tumors. Immunoprecipitation revealed that phosphorylated Met was detected in control tumors and gefitinib-treated tumors, but not in tumors treated with E7050 monotherapy or E7050 plus gefitinib (Fig. 5C), indicating efficacy of E7050 as a Met kinase inhibitor. Moreover, we observed higher levels of phosphorylated Akt in control cancer cells, with this phosphorylation slightly decreased by either E7050 or gefitinib alone and markedly inhibited by the combination of E7050 and gefitinib (Fig. 5D). In addition, there were no discernible differences in HGF concentrations between control and treated groups, when HGF protein concentrations were determined by EIA using lysates of tumors obtained after 5 days of treatment (Supplementary Fig. S3). These results suggested that E7050 overcame the gefitinib resistance associated with inhibition of the Met/Akt pathway.

HGF is a multifunctional cytokine that can be produced not only by cancer cells but also by stromal cells, such as fibroblasts. The HGF receptor, Met, and EGFR interact with each other and mediate redundant signaling (26). Elevated serum concentrations of EGFR ligands and HGF were detected in patients with NSCLC, and HGF expression has been associated with poor prognosis in patients resected for NSCLC (27, 28). Although the role of HGF in EGFR mutant lung cancer remained unclear, we observed HGF-induced EGFR-TKI resistance in EGFR mutant lung cancers (9). Moreover, many studies have shown the important roles of HGF in sensitivity to molecular targeted drugs. Our observations with regard to EGFR-TKI in lung cancer were confirmed by subsequent studies (10, 29), and the concentrations of HGF in peripheral blood were found to be inversely correlated with clinical responses to EGFR-TKIs, in both EGFR mutant and wild-type lung cancer (30, 31). HGF was also found to cause resistance to sunitinib, a multikinase inhibitor, in renal cell carcinoma by compensating for inhibited angiogenesis (32). Taken together, these findings indicate the importance of HGF as a therapeutic target for drug resistance in cancer.

We have shown here that a new Met-TKI, E7050, reversed 3 HGF-induced resistance mechanisms in EGFR mutant lung cancer. First, E7050 reversed HGF-induced gefitinib resistance by inhibiting Met phosphorylation and thereby suppressing the downstream PI3K/Akt pathway. Second, E7050 inhibited the HGF-induced resistance to next-generation EGFR-TKIs, irreversible EGFR-TKIs, and mutant-selective EGFR-TKIs. Third, E7050 prevented the emergence of resistant clones induced by continuous exposure to HGF.

An interaction between HGF and Met amplification has been associated with EGFR-TKI resistance in lung cancer (10). In the presence of gefitinib, continuous exposure to HGF accelerated the expansion of preexisting Met amplified HCC827 cells. Unexpectedly, when we cultured HCC827 cells with gefitinib and HGF for 30 days, we found that the percentage of cells with Met amplification was not increased. The reason we failed to detect expansion of clones with Met amplification, however, remains unclear. Transfection of the HGF gene into HCC827 cells produced HCC827/HGF cells, which constitutively produce HGF. These cells, however, were selected in the presence of geneticin but not gefitinib, with several clones showing amplification of Met (data not shown). Therefore, this phenomenon may be unique to a population of EGFR mutant lung cancer cells observed only under selection pressure with gefitinib plus an as yet unknown concentration of HGF.

Met was shown to be constitutively phosphorylated in human lung cancer cell lines, with the degree of phosphorylation not always correlated with susceptibility to EGFR-TKIs (33). Indeed, previous studies reported that the level of Met phosphorylation was higher in HCC827 cells than in other EGFR mutant cell lines (9, 10, 13, 29). Similar to these results, we also observed that the level of Met phosphorylation was higher in HCC827 cells than in PC-9 and Ma-1 cells (Supplementary Fig. S4). Although the bands for pMet in our study seem to be weaker than those in a previous study (34), ours and previous studies constantly showed that Met phosphorylation in HCC827 cells was higher than that in other EGFR mutant cells. Although the difference in the intensity of pMet bands between our study and the previous is unclear, it might be due to minor differences in experimental conditions, including the exposure time at Western blot and the cell culture conditions. With regard to HGF-triggered EGFR-TKI resistance, previous studies also support our findings that although HCC827 cells were highly sensitive to EGFR-TKIs, further Met activation or phosphorylation resulted in inducing resistance to EGFR-TKIs (10, 29, 35). We confirmed that knockdown of Met by siRNA canceled HGF-induced resistance in HCC827 cells (9). Moreover, it was reported that Met amplification resulted in increased level of Met phosphorylation and caused resistance to EGFR-TKIs in HCC827 cells (8). This accumulating evidence indicates that constitutive Met phosphorylation is insufficient and further activation by HGF or Met amplification may be necessary to induce EGFR-TKI resistance in HCC827 cells. Therefore, there may be a threshold level for Met phosphorylation to sufficiently cause EGFR-TKI resistance.

E7050 inhibits both Met and VEGFR2 kinases (16). In vitro, PC-9 and HCC827 cells express little VEGFR2 (data not shown). E7050 did not significantly inhibit the growth of these cell lines, and the anti-VEGF antibody bevacizumab did not augment the susceptibility of these cell lines to gefitinib (data not shown). These results suggest that the in vitro antitumor effects of E7050, when combined with gefitinib and HGF, may be largely due to Met inhibition. In vivo, we found that very high concentrations of HGF, obtained by HGF gene transfection into cancer cells, increased intratumor vessel density (submitted for publication elsewhere). However, HGF concentrations were lower in our xenograft model of mixed PC-9 and MRC-5 cells (fibroblasts) than in xenograft tumors produced by HGF gene–transfected lung cancer cells. We observed no difference in intratumor vessel density between tumors induced by PC-9 cells alone and tumors induced by PC-9 and MRC-5 cells (Supplementary Fig. S5). In addition, E7050 did not affect significantly the vessel density in tumors induced by PC-9 and MRC-5 cells. Collectively, these observations suggest that the antitumor effects of E7050 in this resistance model may not be predominantly due to angiogenesis inhibition.

The secondary T790M mutation in EGFR is the most prominent mechanism of acquired resistance to EGFR-TKIs in EGFR mutant lung cancer, with this mutation detected in about 50% of these patients (4). The T790M mutation increases the affinity of EGFR for ATP, decreasing the binding of EGFR to EGFR-TKIs and inducing resistance to the latter agents (36). EGFR mutant lung cancer cells with the T790M secondary mutation, however, remain susceptible to EGFR-mediating signaling and are thought to be manageable by inhibition of EGFR-mediated signaling (37). Preclinical studies have shown that next-generation EGFR-TKIs, irreversible TKIs, and mutant EGFR-selective TKIs have activity against gefitinib-resistant tumors with EGFR T790M secondary mutation (21–23). However, several irreversible EGFR-TKIs, including BIBW2992 (38) and HKI-272 (39), failed to meet primary endpoints in clinical trials of patients with EGFR-TKI–refractory lung cancer. High concentrations of HGF have been frequently detected in tumors with EGFR-T790M secondary mutations showing acquired resistance (10, 40, 41). In addition, we found previously (11) and confirmed here that HGF induces resistance to irreversible EGFR-TKIs in EGFR mutant lung cancer cells. Taken together, these observations suggest that HGF expressed in tumors with acquired resistance and EGFR T790M secondary mutations induce resistance to irreversible EGFR-TKI. As E7050 reversed the resistance to irreversible and mutant-selective EGFR-TKIs, it may augment the therapeutic efficacy of next-generation EGFR-TKIs in EGFR mutant lung cancer patients with acquired resistance to the EGFR T790M secondary mutation. These ideas further illustrate the necessity of methods to select patients who develop EGFR-TKI resistance due to HGF.

In conclusion, we have presented preclinical evidence showing that a new Met kinase inhibitor, E7050, may overcome HGF-induced resistance in EGFR mutant lung cancer. Further evaluation of E7050 in clinical trials is warranted to improve the outcomes of patients with EGFR mutant lung cancer.

T. Uenaka and T. Nakagawa are employees of Eisai Co. S. Yano has received a commercial research grant from Eisai Co. and honoraria from speaker's bureau from Chugai Pharma.

The authors thank Mr. Kenji Kita for technical assistance and fruitful discussion.

This study was supported by grants-in-aid for Cancer Research (S. Yano, 21390256) and Scientific Research on Innovative Areas “Integrative Research on Cancer Microenvironment Network” (S. Yano, 22112010A01) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.