Trapping para-Quinone Methide Intermediates with Ferrocene: Synthesis and Preliminary Biological Evaluation of New Phenol-Ferrocene Conjugates

The reaction of para-hydroxybenzyl alcohols with ferrocene in the presence of a catalytic amount of InCl3 provided ferrocenyl phenol derivatives, an interesting class of organometallic compounds with potential applications in medicinal chemistry. This transformation exhibited a reasonable substrate scope delivering the desired products in synthetically useful yields. Evidence of involvement of a para-quinone methide intermediate in this coupling process was also provided. Preliminary biological evaluation demonstrated that some of the ferrocene derivatives available by this methodology exhibit significant cytotoxicity against several cancer cell lines with IC50 values within the range of 1.07–4.89 μM.


Introduction
Since the discovery of ferrocene in the 1950s [1,2], the interest in this organometallic compound has not declined. In fact, its chemistry remains one of the most active areas of research. Very likely, this enduring interest resides in the fact that many functionalized ferrocene derivatives display a wide number of applications in a diverse range of fields [3][4][5][6][7][8][9]. For example, recent investigations have demonstrated the potential of some ferrocene derivatives in medicinal chemistry [10,11]. Particularly, some ferrocene-containing phenols have proved to be of great interest in cancer therapeutics because of their antitumoral activity [12][13][14][15][16]. Among them, a family of ferrocene analogues of hydroxytamoxifen, the so-called ferrocifens (Figure 1a), have been the subject of in-depth investigations showing exceptional cytotoxic activities against some types of breast cancer [17][18][19]. The mode of action of these organometallic drug candidates has been elucidated by electrochemical and chemical oxidation methods. According to these studies, ferrocenyl quinone methides have been suggested to play a key role in the antiproliferative activity [20][21][22][23][24]. The antitumoral activity of some unconjugated bisphenol derivatives of ferrocene ( Figure 1b) has also been evaluated [25,26].

Results and Discussion
The present study was carried out using easily available p-hydroxybenzyl alcohols 1a-l outlined in Figure 2. For our initial study, p-hydroxybenzyl alcohol 1a was chosen as the model substrate (Scheme 1). On the basis of our previous investigations in the ortho series, InCl3 in dichloroethane (DCE) was selected as the catalytic system. Pleasingly, we found that heating a mixture of 1a (1 equiv.), ferrocene (2, 3 equiv.), and InCl3 (10 mol%) in DCE at 60 °C led to complete disappearance of the starting p-hydroxybenzyl alcohol after 2 h (checked by thin layer chromatography, TLC). Chromatographic purification (SiO2, 5:1 hexane ethyl acetate) provided the desired functionalized ferrocene derivative 3a in a remarkable 75% yield. Interestingly, under otherwise similar conditions, benzydryl alcohol 1a' was found to be a fruitless reaction partner, thus demonstrating the key role of the phenolic OH group in the reaction course [31,32]. In connection with our studies on C-H bond functionalization of ferrocene based on the trapping of highly electrophilic species [27][28][29], we have recently described the trapping of ortho-quinone methide intermediates with ferrocene [30]. Interestingly, some of the ferrocene-containing monophenol derivatives available by this methodology (Figure 1c) display remarkable cytotoxic activity against various cancer cell lines.
In order to elucidate the structural requirements for cytotoxicity and, eventually, to identify more bioactive derivatives, we decided to develop a synthetic methodology for the synthesis of the isomeric para-substituted ferrocenylphenol analogues ( Figure 1d). Herein, we report the results of this study; specifically, we describe the generation of para-quinone methide intermediates and their trapping with ferrocene. Preliminary biological evaluation of some of the functionalized ferrocene derivatives prepared in this study is also disclosed.

Results and Discussion
The present study was carried out using easily available p-hydroxybenzyl alcohols 1a-l outlined in Figure 2.
Molecules 2018, 23, x FOR PEER REVIEW 2 of 11 monophenol derivatives available by this methodology (Figure 1c) display remarkable cytotoxic activity against various cancer cell lines.
In order to elucidate the structural requirements for cytotoxicity and, eventually, to identify more bioactive derivatives, we decided to develop a synthetic methodology for the synthesis of the isomeric para-substituted ferrocenylphenol analogues (Figure 1d). Herein, we report the results of this study; specifically, we describe the generation of para-quinone methide intermediates and their trapping with ferrocene. Preliminary biological evaluation of some of the functionalized ferrocene derivatives prepared in this study is also disclosed.

Results and Discussion
The present study was carried out using easily available p-hydroxybenzyl alcohols 1a-l outlined in Figure 2. For our initial study, p-hydroxybenzyl alcohol 1a was chosen as the model substrate (Scheme 1). On the basis of our previous investigations in the ortho series, InCl3 in dichloroethane (DCE) was selected as the catalytic system. Pleasingly, we found that heating a mixture of 1a (1 equiv.), ferrocene (2, 3 equiv.), and InCl3 (10 mol%) in DCE at 60 °C led to complete disappearance of the starting p-hydroxybenzyl alcohol after 2 h (checked by thin layer chromatography, TLC). Chromatographic purification (SiO2, 5:1 hexane ethyl acetate) provided the desired functionalized ferrocene derivative 3a in a remarkable 75% yield. Interestingly, under otherwise similar conditions, benzydryl alcohol 1a' was found to be a fruitless reaction partner, thus demonstrating the key role of the phenolic OH group in the reaction course [31,32]. For our initial study, p-hydroxybenzyl alcohol 1a was chosen as the model substrate (Scheme 1). On the basis of our previous investigations in the ortho series, InCl 3 in dichloroethane (DCE) was selected as the catalytic system. Pleasingly, we found that heating a mixture of 1a (1 equiv.), ferrocene (2, 3 equiv.), and InCl 3 (10 mol%) in DCE at 60 • C led to complete disappearance of the starting p-hydroxybenzyl alcohol after 2 h (checked by thin layer chromatography, TLC). Chromatographic purification (SiO 2 , 5:1 hexane ethyl acetate) provided the desired functionalized ferrocene derivative 3a in a remarkable 75% yield. Interestingly, under otherwise similar conditions, benzydryl alcohol 1a' was found to be a fruitless reaction partner, thus demonstrating the key role of the phenolic OH group in the reaction course [31,32]. Ferrocenyl phenol 3a was characterized by Nuclear Magnetic Resonance (NMR) spectroscopy. Moreover, crystals of compound 3a were obtained from diffusion of pentane into a dichloromethane solution at −20 °C and its molecular structure in the solid state has been determined by single-crystal X-ray diffraction (Figure 3 and Appendix A) [33,34]. The electrochemical behavior of compound 3a was investigated by cyclic voltammetry (Appendix B). With suitable reaction conditions in hand (10 mol% of InCl3, DCE as solvent, 60 °C), a variety of p-hydroxybenzyl alcohols were then evaluated for their suitability for this C-H bond functionalization process (Table 1). First, some p-hydroxybenzyl alcohols substituted with various aryl groups were investigated (entries 2-5). As shown, all three isomeric 4-[hydroxy(tolyl)methyl]phenol derivatives 1b-d (R = tolyl) served as suitable reaction partners for this process furnishing the desired functionalized ferrocene derivatives 3b-d in acceptable isolated yields (51-82%, entries 2-4). Similarly, para-methoxy substituted substrate 1e (R = p-MeOC6H4) delivered the corresponding product 3e in moderate isolated yield (48%, entry 5).
This transformation was compatible with unsaturated functional groups in the benzylic position as demonstrated by the synthesis of ferrocene derivative 3k in moderate yield when substrate 1k Ferrocenyl phenol 3a was characterized by Nuclear Magnetic Resonance (NMR) spectroscopy. Moreover, crystals of compound 3a were obtained from diffusion of pentane into a dichloromethane solution at −20 • C and its molecular structure in the solid state has been determined by single-crystal X-ray diffraction ( Figure 3 and Appendix A) [33,34]. The electrochemical behavior of compound 3a was investigated by cyclic voltammetry (Appendix B). Ferrocenyl phenol 3a was characterized by Nuclear Magnetic Resonance (NMR) spectroscopy. Moreover, crystals of compound 3a were obtained from diffusion of pentane into a dichloromethane solution at −20 °C and its molecular structure in the solid state has been determined by single-crystal X-ray diffraction ( Figure 3 and Appendix A) [33,34]. The electrochemical behavior of compound 3a was investigated by cyclic voltammetry (Appendix B). With suitable reaction conditions in hand (10 mol% of InCl3, DCE as solvent, 60 °C), a variety of p-hydroxybenzyl alcohols were then evaluated for their suitability for this C-H bond functionalization process (Table 1). First, some p-hydroxybenzyl alcohols substituted with various aryl groups were investigated (entries 2-5). As shown, all three isomeric 4-[hydroxy(tolyl)methyl]phenol derivatives 1b-d (R = tolyl) served as suitable reaction partners for this process furnishing the desired functionalized ferrocene derivatives 3b-d in acceptable isolated yields (51-82%, entries 2-4). Similarly, para-methoxy substituted substrate 1e (R = p-MeOC6H4) delivered the corresponding product 3e in moderate isolated yield (48%, entry 5).
This transformation was compatible with unsaturated functional groups in the benzylic position as demonstrated by the synthesis of ferrocene derivative 3k in moderate yield when substrate 1k With suitable reaction conditions in hand (10 mol% of InCl 3 , DCE as solvent, 60 • C), a variety of p-hydroxybenzyl alcohols were then evaluated for their suitability for this C-H bond functionalization process (Table 1). First, some p-hydroxybenzyl alcohols substituted with various aryl groups were investigated (entries 2-5). As shown, all three isomeric 4-[hydroxy(tolyl)methyl]phenol derivatives 1b-d (R = tolyl) served as suitable reaction partners for this process furnishing the desired functionalized ferrocene derivatives 3b-d in acceptable isolated yields (51-82%, entries 2-4). Similarly, para-methoxy substituted substrate 1e (R = p-MeOC 6 H 4 ) delivered the corresponding product 3e in moderate isolated yield (48%, entry 5). (R = allyl) was subjected to the standard reaction conditions. Finally, we found that the parent p-hydroxybenzyl alcohol 1l (R = H) is also a viable substrate affording the desired product 3l in 64% yield (entry 12). Next, to provide further evidence for the involvement of p-quinone methide intermediates in the present coupling, we performed an experiment with a stable p-quinone methide. Thus, 4-benzylidene-2,6-di-tert-butylcyclohexa-2,5-dienone (4) and ferrocene (2) were subjected to the standard reaction conditions (10 mol% of InCl3, DCE, 60 °C). However, a very low conversion was observed after 24 h very likely due to steric hindrance by the tert-butyl groups. Gratifyingly, performing the reaction in toluene at 100 °C enabled the preparation of ferrocene derivative 5 in 80% yield (Scheme 2). Notably, in the absence of InCl3, no reaction occurred at all. Based on these control experiments and on previous related literature precedents, a mechanistic proposal for the reaction of hydroxybenzyl alcohols 1 and ferrocene (2) is outlined in Scheme 3. In the present process, the Lewis acid is proposed to play a dual role. Firstly, it would promote the generation of the key quinone methide intermediate through dehydration of hydroxybenzyl alcohol Next, substrates 1f-j with alkyl groups in the benzylic position were tested (entries 6-10). As shown, primary, secondary and tertiary alkyl groups were well tolerated delivering functionalized ferrocene derivatives 3f-j in moderate yields (42-62%).
This transformation was compatible with unsaturated functional groups in the benzylic position as demonstrated by the synthesis of ferrocene derivative 3k in moderate yield when substrate 1k (R = allyl) was subjected to the standard reaction conditions. Finally, we found that the parent p-hydroxybenzyl alcohol 1l (R = H) is also a viable substrate affording the desired product 3l in 64% yield (entry 12).
Next, to provide further evidence for the involvement of p-quinone methide intermediates in the present coupling, we performed an experiment with a stable p-quinone methide. Thus, 4-benzylidene-2,6-di-tert-butylcyclohexa-2,5-dienone (4) and ferrocene (2) were subjected to the standard reaction conditions (10 mol% of InCl 3 , DCE, 60 • C). However, a very low conversion was observed after 24 h very likely due to steric hindrance by the tert-butyl groups. Gratifyingly, performing the reaction in toluene at 100 • C enabled the preparation of ferrocene derivative 5 in 80% yield (Scheme 2). Notably, in the absence of InCl 3 , no reaction occurred at all. (R = allyl) was subjected to the standard reaction conditions. Finally, we found that the parent p-hydroxybenzyl alcohol 1l (R = H) is also a viable substrate affording the desired product 3l in 64% yield (entry 12). Next, to provide further evidence for the involvement of p-quinone methide intermediates in the present coupling, we performed an experiment with a stable p-quinone methide. Thus, 4-benzylidene-2,6-di-tert-butylcyclohexa-2,5-dienone (4) and ferrocene (2) were subjected to the standard reaction conditions (10 mol% of InCl3, DCE, 60 °C). However, a very low conversion was observed after 24 h very likely due to steric hindrance by the tert-butyl groups. Gratifyingly, performing the reaction in toluene at 100 °C enabled the preparation of ferrocene derivative 5 in 80% yield (Scheme 2). Notably, in the absence of InCl3, no reaction occurred at all. Based on these control experiments and on previous related literature precedents, a mechanistic proposal for the reaction of hydroxybenzyl alcohols 1 and ferrocene (2) is outlined in Scheme 3. In the present process, the Lewis acid is proposed to play a dual role. Firstly, it would promote the generation of the key quinone methide intermediate through dehydration of hydroxybenzyl alcohol 1. Subsequent activation of the quinone methide by Lewis acid complexation would provide intermediate I. This intermediate, with a high electrophilic character at the exocyclic C=C bond, may be involved in a Friedel-Crafts type electrophilic aromatic substitution [35]. Indeed, 1,6-addition of ferrocene would provide the σ-complex intermediate II, which would evolve to the final product with regeneration of the catalyst [36]. Some of the functionalized ferrocene derivatives prepared were evaluated for their cytotoxicity against several cancer cell lines (Table 2). In this preliminary study, ferrocene derivatives 3a and 3g were identified as the most active ones [37]. For example, 3a displayed significant toxicity against A2780 ovarian cancer cell line (IC50 of 1.07 μM). Compared with the value previously reported for the ortho-isomer (IC50 of 2.68 μM), ferrocene derivative 3a has superior characteristics. Ferrocene 3g also displayed toxicity against this cell line (IC50 of 2.23 μM), although somewhat lower than that found for the ortho-analogue (IC50 of 1.86 μM). We have also studied the cytotoxicity profile of ferrocene derivatives 3a and 3g against A549 lung cancer cells. Both derivatives exhibited moderate cytotoxicity with IC50 values of 3.55 and 4.89 μM, respectively. These values are comparable to that previously measured for the ortho-isomers (IC50 of 2.77 and 5.96 μM, respectively).

General
NMR spectra were recorded at room temperature in CDCl3 on a Bruker DPX-300 or Bruker AVANCE-300 MHz instruments (Bruker, Billerica, MA, USA). Chemical shifts are given in ppm relative to TMS ( 1 H, 0.0 ppm) or CDCl3 ( 13 C, 77.0 ppm). High-resolution mass spectra were determined on a VG Autospec M mass spectrometer (Waters Corporation, Milford, MA, USA). Cyclic voltammetric studies were performed using a μ-AutoLab type II equipped with GPES 4.9 software (EcoChemie, Utrecht, The Netherlands). All measurements were carried out using a conventional three electrode system in phosphate saline buffer (pH 7.4). A modified carbon paste acted as the Some of the functionalized ferrocene derivatives prepared were evaluated for their cytotoxicity against several cancer cell lines (Table 2). In this preliminary study, ferrocene derivatives 3a and 3g were identified as the most active ones [37]. For example, 3a displayed significant toxicity against A2780 ovarian cancer cell line (IC 50 of 1.07 µM). Compared with the value previously reported for the ortho-isomer (IC 50 of 2.68 µM), ferrocene derivative 3a has superior characteristics. Ferrocene 3g also displayed toxicity against this cell line (IC 50 of 2.23 µM), although somewhat lower than that found for the ortho-analogue (IC 50 of 1.86 µM). We have also studied the cytotoxicity profile of ferrocene derivatives 3a and 3g against A549 lung cancer cells. Both derivatives exhibited moderate cytotoxicity with IC 50 values of 3.55 and 4.89 µM, respectively. These values are comparable to that previously measured for the ortho-isomers (IC 50 of 2.77 and 5.96 µM, respectively).

General
NMR spectra were recorded at room temperature in CDCl 3 on a Bruker DPX-300 or Bruker AVANCE-300 MHz instruments (Bruker, Billerica, MA, USA). Chemical shifts are given in ppm relative to TMS ( 1 H, 0.0 ppm) or CDCl 3 ( 13 C, 77.0 ppm). High-resolution mass spectra were determined on a VG Autospec M mass spectrometer (Waters Corporation, Milford, MA, USA). Cyclic voltammetric studies were performed using a µ-AutoLab type II equipped with GPES 4.9 software (EcoChemie, Utrecht, The Netherlands). All measurements were carried out using a conventional three electrode system in phosphate saline buffer (pH 7.4). A modified carbon paste acted as the working electrode and a Pt wire as a counter electrode. All potentials were referred to a Ag|AgCl|KCl (sat) reference electrode.
Experiments were carried out under nitrogen using standard Schlenck techniques. 1,2-Dichloroethane was distilled from CaH 2 . Toluene was distilled from sodium-benzophenone ketyl prior to use. TLC was performed on aluminum-backed plates coated with silica gel 60 with F 254 indicator. Flash column chromatography was carried out on silica gel (230-240 mesh). The solvents used in column chromatography, hexane and ethyl acetate, were obtained from commercial suppliers and used without further purification.

General Procedure for the Synthesis of Ferrocene Derivatives 3a-l
InCl 3 (4.4 mg, 0.02 mmol, 10 mol%) was added to a solution of p-hydroxybenzyl alcohols 1 (0.2 mmol) and ferrocene 2 (111.6 mg, 0.6 mmol) in 1,2-dichloroethane (2 mL). The mixture was stirred at 60 • C for 2-14 h (disappearance of 1 checked by TLC). The solvent was removed under reduced pressure and the resulting residue was purified by flash chromatography (silica gel, mixtures of hexanes/ethyl acetate). Two fractions were collected. The first fraction was unreacted ferrocene and the second one was the corresponding functionalized ferrocene derivative 3. Crystals of compound 3a suitable for X-ray analysis were obtained from diffusion of pentane into dichloromethane at −20 • C. Copies of 1 H-and 13 C-NMR spectra are provided in the Supplementary Materials.

Cytotoxic Assays
Cell Counting Kit-8 (CCK-8) from Sigma-Aldrich (Madrid, Spain) was used according to the protocol provided by the company. The A2780 and A549 cell lines were used in this preliminary study. First, cell lines were cultured for 7 days in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS). Then, cells were seeded into a 96-well flat-bottom culture plate at a cell density of 500-2000 cells/well and incubated for 24 h in the same medium (DMEM/10% FBS). After that, 10 µL of a solution of the corresponding ferrocene derivative at different concentrations were added and the cells were incubated for 72 h. Then, 10 µL of the CCK-8 solution were added to each well of the plate. After 2 h of incubation the absorbance at 450 nm was recorded using a BioTek ELx800 Absorbance Microplate Reader (BioTek, Bad Friedrichshall, Germany). Measurements were performed in triplicate, and each experiment was repeated three times. The IC 50 values (µm) were estimated by treatment of the data obtained with the statistical program GraphPad Prism5 (version 5.04).

Conclusions
Guided by earlier work from our group, we have developed a convenient synthesis of para-substituted phenol derivatives containing a ferrocenyl moiety. Salient features of our protocol include (i) easy availability of the required starting materials, (ii) synthetically useful yields, and (iii) mild reaction conditions. This C-H bond functionalization of ferrocene relies on the generation of a para-quinone methide intermediate that, activated by Lewis acid complexation, would serve as electrophilic partner in an aromatic electrophilic substitution. Preliminary biological evaluation revealed that some of the ferrocene derivatives available by this protocol display significant cytotoxicity against ovarian and lung cancer cell lines. Further studies aimed at the preparation of new ferrocene derivatives with enhanced antiproliferative properties are being pursued in our laboratory. rising anodic current) probably associated to phenolic compounds. Only one reduction process is observed as a result of Ic and IIc overlapping ( Figure A1c). The potential of both oxidation and reduction peaks shift to more extreme potentials which indicates that the process is irreversible. The origin of the irreversibility might be the formation of non-conducting products on the electrode surface that hinder the electron transfer. The appearance of resistance affecting the CV shape strongly supports this explanation. Figure A1. Cyclic voltammetry of compound 3a incorporated to the carbon paste electrode in phosphate saline buffer (pH 7.4). Scan rate = 50 mV/s. Potential scan from 0 V to +0.5 V; (a) +0.6 (first scan, black; second scan, blue) and +0.7 (first scan, green; second scan, red) (b) and +1.3 V (c).