Synthesis and Characterization of Lipophilic Salts of Metformin to Improve Its Repurposing for Cancer Therapy.

Epidemiological evidence has accentuated the repurposing of metformin hydrochloride for cancer treatment. However, the extreme hydrophilicity and poor permeability of metformin hydrochloride are responsible for its poor anticancer activity in vitro and in vivo. Here, we report the synthesis and characterization of several lipophilic metformin salts containing bulky anionic permeation enhancers such as caprate, laurate, oleate, cholate, and docusate as counterions. Of various counterions tested, only docusate was able to significantly improve the lipophilicity and lipid solubility of metformin. To evaluate the impact of the association of anionic permeation enhancers with metformin, we checked the in vitro anticancer activity of various lipophilic salts of metformin using drug-sensitive (MYCN-2) and drug-resistant (SK-N-Be2c) neuroblastoma cells as model cancer cells. Metformin hydrochloride showed a very low potency (IC50 ≈ >100 mM) against MYCN-2 and SK-N-Be2c cells. Anionic permeation enhancers showed a considerably higher activity (IC50 ≈ 125 μM to 1.6 mM) against MYCN-2 and SK-N-Be2c cells than metformin. The association of metformin with most of the bulky anionic agents negatively impacted the anticancer activity against MYCN-2 and SK-N-Be2c cells. However, metformin docusate showed 700- to 4300-fold improvement in anticancer potency compared to metformin hydrochloride and four- to five-fold higher in vitro anticancer activity compared to sodium docusate, indicating a synergistic association between metformin and docusate. A similar trend was observed when we tested the in vitro activity of metformin docusate, sodium docusate, and metformin hydrochloride against hepatocellular carcinoma (HepG2) and triple-negative breast cancer (MDA-MB-231) cells.

Introduction

Metformin hydrochloride is an antidiabetic biguanide which was approved for the treatment of type 2 diabetes mellitus (T2DM) almost 60 years ago.[1−3] Metformin hydrochloride is typically the first choice for the treatment of T2DM because of its well-established efficacy, safety, and relatively low cost. Repurposing metformin for cancer therapy has been of interest after epidemiological studies suggested that diabetic cancer patients on metformin therapy have higher survival rates.[1−5] Several in vitro and preclinical studies have been carried out to validate the anticancer activity of metformin.[5−11] The anticancer activity of metformin is mediated via the inhibition of the mammalian target of rapamycin, activation of adenosine monophosphate-activated protein kinase, and p53, and inhibition of mitochondrial complex I.[1−3] Studies also demonstrate that metformin is capable of selectively killing cancer stem cells without affecting the normal stem cells.[8]

A vast majority of observational retrospective studies and recent meta-analyses have shown a correlation between metformin treatment and reduction in the incidence of various cancers such as colon cancer, breast cancer, prostate cancer, and non-small cell lung cancer.[1,2,4,5,8,12−15] At the same time, some retrospective and prospective trials have reported lack of beneficial effects of metformin in cancer patients.[16−19] At present, several prospective, controlled clinical trials are being carried out to demonstrate the efficacy of metformin, either alone or in combination with established anticancer agents, in various cancers.

Although metformin has been used in the clinic for almost six decades, it does not have optimal biopharmaceutical properties. The oral bioavailability of metformin ranges from 50–60%, and almost 30% of the drug is excreted unchanged via feces because of poor absorption. Furthermore, metformin shows considerable intra- and interindividual variations in absorption.[20] Metformin is a highly basic and hydrophilic (BCS class III) drug which is completely ionized (protonated) in the gastrointestinal tract. These properties coupled with metformin’s poor membrane permeability are some of the key reasons for the low oral bioavailability of metformin. Incidentally, metformin shows anticancer activity at very high concentrations (IC50 > 1 mM), and the lack of efficacy of metformin in the clinical trials could partially be attributed to the inability of metformin to reach a concentration in the tumor that is high enough to exert an anticancer effect.[8] Thus, improvement in the oral delivery of metformin is necessary to achieve effective concentrations for cancer treatment.

Lipophilic salt synthesis is an emerging strategy to improve the apparent lipophilicity, lipid vehicle solubility, permeability, and bioavailability of ionizable drugs belonging to BCS class II and BCS class III.[21] In view of this, we hypothesized that the use of bulky anionic permeation enhancers with a proven ability to improve the permeation of hydrophilic molecules with low membrane permeability as counterions could improve the apparent lipophilicity, lipid vehicle solubility, and cell permeability of metformin. Hence, we focused our studies on the development of lipophilic salts of metformin using various anionic permeation enhancers such as sodium caprate, sodium laurate, sodium oleate, sodium docusate, and sodium cholate as counterions.

Here, we report the synthesis and characterization of metformin caprate, metformin laurate, metformin cholate, metformin oleate, and metformin docusate (Figure ). Our studies show that only metformin docusate is capable of sufficiently improving the lipophilicity of metformin to enable its miscibility with bioavailability-enhancing lipid excipients. Further, using drug-sensitive (MYCN-2) and drug-resistant (SK-N-Be2c) neuroblastoma cell lines as model cancer cells, we demonstrate that metformin docusate is capable of dramatically improving the anticancer activity of metformin, which is an indirect indicator of improved cell permeability.

Figure 1

Structures of various lipophilic salts of metformin synthesized in this study.

Results

Synthesis of Lipophilic Salts of Metformin

Metformin lipophilic salts were prepared by a simple metathesis reaction between the hydrochloride salt of metformin and the sodium salt of docusate acid, capric acid, lauric acid, oleic acid, or cholic acid. Metformin lipophilic salts containing caprate, laurate, oleate, or cholate as a counterion could be synthesized using methanol as a solvent for the metathesis reaction, whereas a biphasic mixture of dichloromethane (DCM) and water was used for the preparation of metformin docusate. Metformin caprate, metformin laurate, metformin oleate, and metformin cholate were obtained as solids, whereas metformin docusate was obtained as an ionic liquid.

Characterization of Metformin Lipophilic Salts

The successful formation of metformin lipophilic salts (Figure ) was confirmed using several characterization techniques such as 1H and 13C NMR spectroscopy analyses (Supporting Information; Figures S1–S17), melting point determination, Fourier transform infrared (FTIR) (Supporting Information; Figures S18–S22), and high-performance liquid chromatography (HPLC). The 1H NMR spectrum of the metformin lipophilic salt was compared with both metformin hydrochloride and the counterion. The ratios of signal integrations from metformin hydrochloride and the counterion confirmed were used to confirm the puristy of the metformin lipophilic salts (Supporting Information; Figures S1–S17). The proton shifts in the metformin lipophilic salt with respect to metformin and the counterions confirmed the interaction between the anion and cation.[22,23] In the case of metformin caprate, metformin laurate, metformin oleate, metformin cholate, the protons adjacent to the carboxylate anion were found to be more shifted compared to the pure sodium salts (Supporting Information; Tables S1–S4). However, in the case of metformin docusate, the protons near the sulfonyl group were found to be shifted because of the strong electrostatic interaction (Supporting Information; Table S5). The synthesized metformin lipophilic salts showed a considerably different melting point than metformin hydrochloride and the respective counterion.

The FTIR spectrum of metformin hydrochloride showed characteristic bands at 3368 and 3294 cm–1 that correspond to the N–H stretching of the primary amine, whereas the bands at 3150 and 1544 cm–1 are due to the N–H stretching and bending of the secondary amine. The FTIR spectrum also exhibited a notable band at 1622 cm–1 for the C=N imine stretch (Table ; supplementary Figures S18–S22). Almost all metformin lipophilic salts exhibited characteristic bands of metformin with a slight shifting (Table ; Supplementary Figures S18–S22). The shifting and appearance of new bands in the FTIR spectrum of metformin in the presence of counterions show the physical interaction of metformin with the bulky anionic counterions. The ionic liquid metformin docusate showed a band at 1734 cm–1 (C=O ester stretch) as well as a broad band at 1204 cm–1 (S=O stretch) from the docusate counterion. The caprate, laurate, oleate, and cholate salts of metformin showed characteristic bands at 1737, 1731, 1737, and 1736 cm–1, respectively, which are attributed to the C=O stretching of carbonyl of fatty acids. In the case of the metformin–cholate ion pair, the broad band at 3368–3262 cm–1 corresponds to the O–H stretch of cholate, which got shifted and merged in the bands of the metformin N–H stretch.

Table 1

Summary of the Characteristic FTIR Peaks Representing the Key Functional Groups in Metformin Hydrochloride and Various Lipophilic Metformin Saltsa

metformin/metformin salts	metformin–NH2	carbonyl group C=O	imine C=N stretch	N–H bending	C–N stretch
metformin	3368.17 and 3294.06		1622.87	1544.76	1059.73
metformin–docusate	3346.99	1734.72	1639.26	1558.26	1035.63
metformin–oleate	3368.20	1737.62	1623.43	1560.19	1040.22
metformin–caprate	3361.46	1737.62	1641.19	1553.44	1047.20
metformin–laurate	3369.17	1731.83	1640.23	1554.41	1052.98
metformin–cholate	3368.20	1736.65	1651.80	1579.48	1079.98

The characteristic FTIR peaks of metformin show considerable shifts in the lipophilic salts of metformin, indicating the formation of metformin salts.

The HPLC analysis of metformin lipophilic salts showed the presence of a metformin peak at 4.13 min. The HPLC analysis also confirmed the absence of any impurities in metformin lipophilic salts.

Kinetic Solubility Studies on Metformin Lipophilic Salts Using Lipid Vehicles

The results of kinetic solubility of metformin hydrochloride and metformin lipophilic salts are shown in Table . Metformin hydrochloride, metformin caprate, metformin laurate, metformin oleate, and metformin cholate showed low solubility (<5 mg/g) in lipid vehicles (Miglyol 812N and Capryol 90). On the contrary, metformin docusate demonstrated a significantly high solubility (>200 mg/g) in Miglyol 812N and Capryol 90 (Table ).

Table 2

Kinetic Solubility Study of Metformin Salts in Lipid Excipients

	solubility (mg/g) of lipid excipient
lipid excipient	metformin hydrochloride	metformin oleate	metformin laurate	metformin caprate	metformin cholate	metformin docusate
medium-chain triglyceride (Miglyol 812N)	<5	<5	<5	<5	<5	>200
propylene glycol monocaprylate (Capryol 90)	<5	<5	<5	<5	<5	>200

Evaluation of In Vitro Cytotoxicity against Neuroblastoma Cell Lines

We evaluated the in vitro cytotoxicity of metformin lipophilic salts to understand the impact of the association of the bulky permeation enhancers as counterions. For this purpose, we used MYCN nonamplified drug-sensitive (MYCN-2) and MYCN gene-amplified drug-resistant (SK-N-Be2c) neuroblastoma cells as models. We also evaluated the cytotoxicity of metformin hydrochloride and individual bulky counterions (without metformin) in neuroblastoma cells as controls. We used the sulforhodamine B (SRB) assay to measure the in vitro dose-dependent effects of the different treatments.[24] As anticipated, metformin hydrochloride solution exhibited a very high IC50 in MYCN-2 as well as Be2c cells (Figure ). In our studies, the anionic permeation enhancers also showed cytotoxicity against MYCN-2 as well as Be2c cells, but the extent of activity was different (Figures –4). The IC50 values for metformin caprate, metformin laurate, metformin oleate, and metformin cholate were significantly lower than that of metformin (Figures and 4). However, these metformin salts were generally less potent than the sodium salts of their respective counterions (Figures and 4). On the contrary, docusate alone exhibited cytotoxicity in NB cells, and the association of metformin with the docusate counterion significantly potentiated the anticancer activity of metformin. Metformin docusate showed approximately 700- to 4300-fold lower IC50 values compared to the metformin solution (Figure ). Furthermore, metformin docusate showed an approximately four- to five-fold lower IC50 value compared to sodium docusate in MYCN2 and SKNBe2c cells, respectively (Figure ).

Figure 2

Metformin docusate shows significantly higher in vitro anticancer activity than metformin hydrochloride as well as docusate sodium in drug-sensitive (MYCN-2) and drug-resistant (SK-N-Be2c) neuroblastoma cell lines. The in vitro cytotoxicity was evaluated using the SRB assay. Data expressed as mean ± SEM (n = 3).

Figure 4

In vitro cytotoxicity of metformin caprate, metformin laurate, metformin oleate, and metformin cholate and their counterions against SK-N-Be2c cells. Data expressed as mean ± SEM (n = 3).

Figure 3

In vitro cytotoxicity of metformin caprate, metformin laurate, metformin oleate, and metformin cholate and their counterions against MYCN-2 cells. Data expressed as mean ± SEM (n = 3).

Evaluation of In Vitro Cytotoxicity of Metformin Docusate against HepG2 and MDA-MB-231 Cells

To evaluate the synergistic interaction between metformin and docusate in metformin docusate, we carried out in vitro cytotoxicity experiments using HepG2 and MDA-MB-231 cells as model cancer cells. We used metformin hydrochloride and sodium docusate as controls. Interestingly, the IC50 value of metformin docusate was significantly lower than that of metformin hydrochloride and slightly lower than that of sodium docusate (Figures and 6). The difference between the IC50 value of metformin docusate and sodium docusate was not as pronounced as seen in drug-sensitive and drug-resistant neuroblastoma cells.

Figure 5

In vitro cytotoxicity of (A) metformin hydrochloride, (B) sodium docusate, and (C) metformin docusate HepG2 cells. Metformin docusate showed a lower IC50 value than metformin hydrochloride and sodium docusate. Data expressed as mean ± SEM (n = 3).

Figure 6

In vitro cytotoxicity of (A) metformin hydrochloride, (B) sodium docusate, and (C) metformin docusate MDA-MB-231 cells. Metformin docusate showed a lower IC50 value than metformin hydrochloride and sodium docusate. Data expressed as mean ± SEM (n = 3).

Discussion

In recent years, metformin hydrochloride has been of great interest in the “drug repurposing” field.[1−4] Metformin hydrochloride is being explored for several repurposing applications including cancer therapy. Because of the excellent clinical tolerability, potential to modulate several molecular pathways that can benefit cancer therapy, ease of availability, and low cost of metformin, several clinical trials are being carried out to evaluate its anticancer effect.[1−4,12−15] All the hitherto reported clinical trials utilize the existing formulations of metformin hydrochloride. However, the low potency of metformin hydrochloride against cancer cells and its suboptimal biopharmaceutical properties (low permeability and erratic oral bioavailability) are the major hurdles for its repurposing for oral chemotherapy.

Recent studies have shown that metformin salts such as metformin butyrate, metformin gamma-aminobutyric acid salt, metformin pregabalin salt, and metformin gabapentin salt can improve the potency of metformin by 2- to 10-fold against various breast cancer cells in vitro.[25,26] Furthermore, metformin butyrate salt showed a significantly higher anticancer activity than metformin in the mouse model of triple-negative breast cancer. However, a relatively high dose of metformin butyrate (250 mg/kg) was required to demonstrate an in vivo effect.[26] We hypothesized that the synthesis of lipophilic metformin salts that contain bulky anionic permeation enhancers as counterions will have improved cell permeability, which could improve the chemotherapeutic potential of metformin. Thus far, the syntheses of metformin docusate and metformin stearate have been reported in the literature.[27,28] However, their potential to improve the anticancer activity of metformin has not been explored. We decided to evaluate various bulky anionic permeation enhancers such as sodium caprate, sodium laurate, sodium oleate, sodium cholate, and sodium docusate which have a proven capability to improve the permeability of hydrophilic small molecules as well as macromolecules.[29−33] We also envisaged that the development of lipophilic metformin salts could further facilitate their incorporation into lipid-based drug delivery systems which could further improve the oral bioavailability and oral chemotherapeutic effect of metformin.

Typically, for a successful salt formation, the required difference between the pKa values of an acid and a base should be >2.[34] The pKa value of metformin is 10.59,[35] and the reported pKa values for caprate, laurate, oleate, cholate, and docusate are 4.9, 4.95, 5, 4.98, and −0.75, respectively.[36] Thus, the difference in the pKa values of metformin and the counterions was large enough to allow for the salt formation. We successfully developed various lipophilic salts of metformin using the metathesis reaction. We could synthesize metformin caprate, metformin laurate, metformin oleate, and metformin cholate using the reported methods with suitable modifications. However, our attempts to synthesize pure metformin docusate using the reported methods were unsuccessful.[27,37] The reported methods utilize methanol or isopropanol for the metathesis reaction. We found that the rate of reaction of metformin hydrochloride and sodium docusate was slower in methanol or isopropanol because of their ability to donate a proton.[38,39] Also, the quality of the end product was compromised. Hence, we completely changed the conditions for the metathesis reaction to synthesize metformin docusate. Metformin hydrochloride has minimal solubility in several polar organic solvents, whereas it is highly soluble in water. Sodium docusate, because of its amphiphilic nature, has solubility in water as well as various organic solvents. After several trials, we found that the DCM and water biphasic mixture was quite appropriate for the synthesis of metformin docusate with a high yield (∼97%) and purity. Our characterization studies confirmed the formation of metformin lipophilic salts. It was noteworthy that metformin caprate, metformin laurate, metformin oleate, and metformin cholate were all obtained as solids, whereas metformin docusate was obtained as an ionic liquid. Our kinetic solubility screening studies showed that metformin caprate, metformin laurate, metformin oleate, and metformin cholate did not have appreciable solubility in the commonly used lipid excipients such as Miglyol 812N and Capryol 90, indicating their lack of suitability for the oral lipid-based formulations. Previous studies have shown that metformin stearate showed solubility in water up to 3 mg/mL,[28] and our studies showed that metformin caprate, metformin laurate, metformin oleate, and metformin cholate have considerable water solubility (>1 mg/ml; data not shown). Metformin docusate was the only metformin salt that showed high lipid solubility. Williams et al. reported the solubility of metformin docusate in medium-chain triglycerides and propylene glycol fatty acid esters to be >75 mg/g.[27] We continued evaluating the lipid solubility of metformin docusate in lipid vehicles and found that the solubility of metformin docusate is >200 mg/g. A recent study by Ford et al., indicates that ionic liquids/lipophilic salts based on branched alkyl sulfonates (such as docusate sodium) as counterions exhibit more lipid solubility than those based on linear alkyl sulfonates as counterions.[40] This may be the reason behind the higher lipid solubility of metformin docusate compared to the other metformin salts with linear fatty acids as counterions. Metformin docusate is the only lipophilic salt of metformin suitable for the development of oral lipid-based formulations.

Owing to previous reports on the improved in vitro and/or in vivo anticancer activity of metformin salts compared to metformin hydrochloride,[25,26] we decided to evaluate the potential of lipophilic metformin salts developed in this investigation against model cancer cell lines. It is noteworthy that the hitherto reported papers on the improved anticancer activity of metformin salts do not provide adequate information about the in vitro and/or in vivo anticancer activity of the counterions used to for the preparation of the metformin salts . For example, Lee et al., demonstrated an enhanced anticancer activity of metformin butyrate in vitro and in vivo.[26] However, the authors did not use butyrate, a natural histone deacetylase inhibitor with a known anticancer activity,[41] as a control for in vitro and in vivo studies. Hence, we sought to evaluate the in vitro cytotoxicity of the lipophilic salts of metformin, their respective counterions, and metformin hydrochloride. We used drug-sensitive (MYCN-2) and MYCN gene-amplified drug-resistant (SK-N-Be2c) neuroblastoma cells as models for this purpose.

Our in vitro studies showed that metformin hydrochloride had low potency against MYCN-2 as well as Be2c cells (IC50 > 100 mM). Interestingly, our studies showed that anionic permeation enhancers used for the synthesis of lipophilic metformin salts showed significantly higher in vitro cytotoxicity against MYCN-2 and Be2c cells compared to metformin hydrochloride, although the extent of activity was different (Figures –4).

The surface-active nature of the anionic permeation enhancer could be one of the reasons for the effects against cancer cells. Furthermore, the literature reports indicate that the anionic permeation enhancers used in this study can cause toxicity to cancer cells by modulating several molecular pathways.[42−44] The anticancer activity of capric acid is mediated by the downregulation of the genes essential for cell cycle progression and division and upregulation of genes responsible for cell cycle arrest and apoptosis. Lauric acid has been reported to exert the anticancer activity by downregulating the epidermal growth factor receptor and by augmenting the reactive oxygen species levels and phosphorylation of extracellular signal-regulated kinase. A recent study has established that sodium cholate can induce apoptosis in the cancer cells and sustain the activation of p38 and Akt. Jiang et al., have reported that oleic acid’s anticancer activity is mediated by the induction of the G0/G1 cell cycle arrest, reduction in the expression of cyclinD1 and Bcl-2, and enhanced expression of p53 and cleaved caspase-3.[45−47] Thus, the in vitro cytotoxicity observed with anionic permeation enhancers could be due to their inherent chemotherapeutic potential, and our investigation underscores the necessity of evaluating in vitro and using appropriate controls for the in vitro and in vivo anticancer activity of counterions used to enhance the activity of ionizable anticancer drugs.

It is well known that the counterions have a considerable impact on the physicochemical properties as well as biological properties. Several examples in the literature describe the alteration in the inherent biological activity of the drug molecule after association with different counterions.[48−50] Our observation is in agreement with the reported data.

Our in vitro cytotoxicity studies showed that metformin caprate, metformin laurate, metformin oleate, and metformin cholate showed a significantly higher anticancer activity than metformin hydrochloride, but generally they exhibited lesser activity than the respective counterions. Thus, the association of metformin with the caprate, laurate, oleate, and cholate counterions passivated the anticancer activity of the counterions. On the other hand, the association of metformin with docusate appeared to be beneficial in terms of anticancer activity. The in vitro chemotherapeutic effect of metformin docusate was greater than that of metformin as well as docusate, indicating the synergy between metformin and docusate in addition to the enhanced permeability. Interestingly, the inhibitory effect of metformin docusate on NB cell proliferation was increased in the MYCN-amplified SK-N-Be2c cell line when compared to the effects on the MYCN2 cell line which has a single copy of the MYCN gene amplification (36 μM vs 150 μM, metformin–docusate). The enhanced antiproliferative effect may be explained by the ability of metformin to destabilize MYCN in neuroblastoma.[51] MYCN amplification is the strongest prognostic marker for poor prognosis and survival in neuroblastoma.[52] The destabilization of MYCN may therefore enhance the proliferative effect of metformin in SKNBe2c cells. Metformin docusate was also more active than metformin hydrochloride and sodium docusate when tested in HepG2 and MDA-MB-231 cells, indicating its potential to improve the repurposing of metformin for cancer therapy. It is noteworthy that metformin caprate, metformin laurate, metformin oleate, and metformin cholate exhibit considerable solubility in water despite the presence of bulky counterions, whereas metformin docusate is a hydrophobic ionic liquid. It is also possible that metformin docusate, because of its amphiphilic nature, improved the passive permeability of metformin, leading to a higher intracellular concentration of metformin and higher anticancer activity. Our study indicated that the association of metformin and docusate is synergistic, and our future studies would explore this aspect further. Recent studies have shown that docusate-containing ionic liquids or hydrophobic ion pair complexes can be incorporated into lipid-based drug delivery systems to dramatically improve the oral bioavailability of various small molecules as well as peptides.[27,31,53,54] Our future efforts will focus on the development of metformin docusate-containing lipid-based formulations.

Conclusions

In summary, we showed that it is possible to synthesize salts of metformin containing bulky anionic permeation enhancers as counterions. Metformin caprate, metformin laurate, metformin oleate, and metformin cholate did not show appreciable solubility in lipid vehicles despite the bulky nature of the counterions, whereas metformin docusate showed a significantly high solubility in lipid vehicles. Among various lipophilic salts of metformin, only metformin docusate showed a dramatic improvement in the chemotherapeutic potential of metformin, at least in vitro.

Materials and Methods

Materials

Metformin hydrochloride, sodium docusate, sodium oleate, sodium caprate, sodium laurate, and sodium cholate were purchased from Carbosynth US LLC (CA, USA). Methanol (AR grade), acetonitrile (HPLC grade), triethylamine (AR grade), sodium dihydrogen phosphate (AR grade), O-phosphoric acid (AR grade), acetone (AR grade), isopropanol (AR grade), ethyl acetate (AR grade), sodium lauryl sulfate (AR grade), tributyrin (AR grade), and 0.45 μm membrane filters were purchased from VWR International (PA, USA). Miglyol 812 N (IOI Oleo GmbH, Hamburg, Germany) and Capryol 90 (Gattefosse USA, NJ, USA) were received as gift samples. All other chemicals used were of analytical grade, unless otherwise indicated.

Cell Lines

The human NB cell lines SK-N-Be2c (provided by Michael D. Hogarty, PA, USA) and MYCN2 (provided by Jason Shohet, TX, USA) were maintained in RPMI-1640 (Mediatech, Inc., Manassas, VA, USA) containing 10% (v/v) heat-inactivated fetal bovine serum (FBS) (Atlanta Biologicals, Lawrenceville, GA, USA). The MYCN2 cells are SHEP-1 cells with doxycycline-inducible MYCN overexpression. The HepG2 and MDA-MB-231 cell lines were maintained in Dulbecco’s modified Eagle’s medium (DMEM) (Fisher Scientific, Waltham, MA, USA) containing 10% heat-inactivated FBS.

NMR Spectroscopy

Nuclear magnetic resonance (NMR) spectra were obtained using a Bruker AVANCE Digital 400 MHz NMR spectrometer coupled to a BACS 1 automatic sample changer. The spectrometer is equipped with a 5 mm PABBO BB-1H/D Z-GRD probe. The 1H spectra of the purified products (6–8 mg) were recorded in (500–600 μL) methanol-d4 (Alfa Aesar, 99.8% D) or dimethyl sulfoxide-d6 (Acros Organics, 99.5% D) with average 16 scans, and for 13C spectra, average 1000 scans were performed for each sample. Chemical shifts were reported in parts per million (ppm), with the residual undeuterated solvent peaks as an internal reference for 1H NMR: CD3OD (3.31 and 4.78 ppm) and DMSO (2.50 ppm). Multiplicities were reported as singlet (s), doublet (d), triplet (t), or multiplet (m), and coupling constants (J) are given in Hertz (Hz).

Determination of Melting Point

The melting point of compounds was determined by the capillary method. Briefly, an appropriate amount of compound was filled in a capillary tube (1.5–1.8 mm OD) fused at one end, and the melting point was determined using a BUCHI melting point B-540 apparatus. The instrument was set at 400 °C with a gradient of 10 °C.

Synthesis of Metformin Oleate, Metformin Laurate, Metformin Caprate, and Metformin Cholate

The procedure reported by Koh et al. was modified to synthesize various metformin salts.[25] An equimolar quantity of metformin hydrochloride (100 mg, 0.60 mM) was dissolved in methanol (10 mL) along with the counterion sodium salt (0.60 mmol). The mixture was stirred at 50 °C for 3 h, and then methanol was evaporated (Scheme ). DCM (30 mL) was added into the resulting white solid residue and was washed gently with a small quantity of distilled water. The organic phase was separated and then dried against sodium sulfate, filtered, and evaporated to obtain the desired product (white solid: ∼65–72%).

Scheme 1

Synthetic Scheme for Metformin Salts Containing Bulky Carboxylate Counterions

Metformin Oleate

White solid. Melting point: 187–190 °C. 1H NMR (400 MHz, CD3OD): δH 5.36 (t, J = 5.7 Hz, 2H, CH-g), 3.06 (3.056) [s, 6H, N(CH3)2-f], 2.17 (t, J = 8.0 Hz, 4H, 2CH2-e), 2.05 (m, 2H, CH2-d), 1.62 (t, J = 1.1 Hz, 2H, CH2-c), 1.35 (d, J = 14.5 Hz, 20H, CH2-b), 0.92 (t, J = 7.4 Hz, 3H, CH3-a). 13C NMR (CD3OD): δ 183.1, 161.4, 160.4, 130.9, 39.3, 38.0, 33.0, 30.9, 30.8, 30.6, 30.5, 30.4, 30.3, 28.2, 28.1, 27.8, 23.7, 14.6.

Metformin Laurate

White solid. Melting point: 178–181 °C. 1H NMR (400 MHz, CD3OD): δH 3.04 [s, 6H, N(CH3)2-a], 2.17 (t, J = 7.25 Hz, 2H, CH2-b), 1.59 (t, J = 7.4 Hz, 2H, CH2-c), 1.29 (m, 16H, CH2-d), 0.90 (t, J = 6.45 Hz, 3H, CH3-e). 13C NMR (400 MHz, CD3OD): δ 180.8, 159.8, 159.9, 37.2, 36.6, 31.7, 29.5, 29.3, 29.2, 29.1, 26.1, 22.3, 13.0.

Metformin Caprate

White solid. Melting point: 140–143 °C. 1H NMR (400 MHz, CD3OD): δH 3.04 [s, 6H, N(CH3)2-a], 2.15 (t, J = 8.0 Hz, 2H, CH2-b), 1.57 (t, J = 7.4 Hz, 2H, CH2-c), 1.30 (m, 14H, CH2-d), 0.90 (t, J = 6.48 Hz, 3H, CH3-e). 13C NMR (400 MHz, CD3OD): δ 181.6, 160.0, 159.0, 38.0, 36.6, 31.6, 29.5, 29.3, 29.2, 29.1, 26.4, 22.3, 13.0.

Metformin Cholate

White solid. Melting point: 201–204 °C. 1H NMR (400 MHz, CD3OD): δH 3.98 (m, 1H, CH-a), 3.82 (m, 1H, CH-b), 3.38 (m, 1H, CH-c), 3.05 [s, 6H, N(CH3)2-d], 1.37–2.35 (m, all CH2, CH aliphatic, 23H-e), 1.13 (m, 2H, CH2-f), 1.05 (d, J = 5.35 Hz, 3H, CH3-g), 0.93 (s, 3H, CH3-h), 0.73 (s, 3H, CH3-i). 13C NMR (400 MHz, CD3OD): δ 182.3, 160.0, 159.0, 72.7, 71.5, 67.7, 41.8, 41.5, 39.7, 39.1, 36.6, 35.0, 34.5, 34.4, 32.7, 29.8, 28.1, 27.3, 26.5, 22.9, 21.8, 16.5, 11.6.

Synthesis of Metformin Docusate

Equimolar quantities of metformin hydrochloride (100 mg, 0.60 mmol) and sodium docusate (267 mg, 0.60 mmol) were dissolved in the DCM:water (12 mL, 1:1) mixture and stirred at room temperature for 4 h (Scheme ). Further, the DCM layer was separated and concentrated in vacuo. Acetonitrile (20 mL) was added to the resulting viscous liquid residue to remove traces of salt (NaCl) and unreacted sodium docusate, and the mixture was stirred for 30 min. The mixture was passed through a 0.22 μm pore size nylon membrane filter to remove solid residues and concentrated in vacuo to give a clear viscous oil (97%).

Scheme 2

Synthetic Scheme for Metformin–Docusate Ion Pair

Metformin Docusate

Clear viscous liquid. 1H NMR (400 MHz, DMSO-d6): δH 7.17 (s, 1H, NH-a), 6.55 (s, 2H, NH-b,c), 3.90 (m, 1H, CH-d), 3.68–3.64 (dd, J = 3.6 Hz, 4H, CH2-e), 2.93 [s, 6H, N(CH3)2-f], 2.82–2.78 (dd, J = 3.6 Hz, 2H, OCH2-g), 1.50 (m, 2H, CH-h), 1.24 (m, 16H, CH2-i), 0.85 (m, 12H, CH3-j). 13C NMR (400 MHz, CD3OD): δ 171.2, 168.4, 159.8, 158.9, 67.4, 66.8, 61.9, 28.7, 36.6, 33.5, 30.1, 30.0, 28.7, 23.4, 23.3, 22.6, 13.1, 10.0.

Characterization of Metformin Salts Using FT-IR and HPLC

FTIR spectroscopy measurements were performed using a FTIR spectrophotometer (Thermo Scientific Nicolet iS10) equipped with a diamond attenuated total reflection unit. FTIR spectra were obtained in transmission mode from 4000 to 500 cm–1 and with an average 32 scans for background and each sample. Metformin hydrochloride and metformin salts were identified by using the previously reported HPLC method with suitable modifications.[55] The HPLC apparatus consisted of a HPLC binary pump (Shimadzu, USA), a Shimadzu SPD-20A UV–vis detector (Shimadzu, USA), prominence autosampler injector and Lab solution integrator software 5.87 SP1. Briefly, a stock solution of metformin hydrochloride or metformin salt (100 μg/mL) was prepared in water and/or methanol, and the stock solution was further diluted with methanol to obtain a concentration of 10 μg/mL. The chromatographic separation was carried out on a Gemini C18 reversed-phase column (150 × 4.6 mm, 3 μ particle size). The mobile phase was composed of methanol and a buffer containing 10 mM sodium dihydrogen phosphate in the ratio of 40:60 (v/v). The flow rate of the mobile phase was 0.4 mL/min. The column oven temperature was set at 35 °C. Metformin and metformin ion pairs were monitored at 232 nm.

Kinetic Solubility Study of Metformin Salts in Lipid Excipients

The kinetic solubility of the commercial salt metformin hydrochloride and its lipophilic salts was determined in a medium-chain triglyceride (Miglyol 812N) and propylene glycol monocaprylate (Capryol 90) using a previously reported method.[27] Miglyol 812N and Capryol 90 (1 g) were weighed and transferred to a scintillation vial. Metformin hydrochloride or metformin lipophilic salt (5 mg) was added to Miglyol 812N or Capryol 90. The mixture was then sonicated (Fisherbrand sonicator unit–model FB11201, USA) at a frequency of 45 kHz for 15 min and was visually inspected for solubilization/precipitation. If the first portion of the metformin salt was solubilized in the oils, then the metformin salt(s) was added in 5 mg increments, followed by 15 min of sonication until precipitation is observed.

Evaluation of In Vitro Cytotoxicity of Metformin Hydrochloride, Metformin Lipophilic Salts, and Anionic Permeation Enhancers

The impact of association metformin with bulky counterions was evaluated by determining the in vitro cytotoxicity of metformin hydrochloride, metformin lipophilic salts, and anionic permeation enhancers against the drug-sensitive (MYCN-2) and drug-resistant (SK-N-Be2c) neuroblastoma cell lines. Metformin hydrochloride, sodium caprate, metformin caprate, sodium laurate, metformin laurate, sodium cholate, metformin cholate, sodium oleate, and metformin oleate were dissolved in water to obtain a stock solution. Docusate sodium and metformin docusate were dissolved in DMSO. The SRB colorimetric assay was used to determine cell proliferation, following the protocol previously described.[24] Briefly, the cells were seeded at a density of 10,000 cells/well on a transparent, flat-bottom, 96-well plate and allowed to settle overnight. At the initiation of each experiment (t = 0), and after drug treatments, 100 μL of 10% (w/v) trichloroacetic acid was added to each well, incubated for 1 h at 4 °C, washed with deionized water, and dried at room temperature. A volume of 100 μL of 0.057% (w/v) SRB solution was added to each well, incubated for 30 min at room temperature, rinsed four times with 1% (v/v) acetic acid, and allowed to dry at room temperature. Finally, 200 μL of 10 mM Tris base solution (pH 10.5) was added to each well, and after shaking for 5 min at room temperature, the absorbance was measured at 510 nm in a microplate reader. The absorbance at t = 0 was compared with the absorbance at the end of the experiment to determine the cell growth in treated cells compared with control cells.

In Vitro Cytotoxicity Evaluation of Metformin Docusate against Hepatocellular Carcinoma (HepG2) and Triple-Negative Breast Cancer (MDA-MB-231) Cells

The in vitro cytotoxicity of metformin docusate, sodium docusate, and metformin hydrochloride against HepG2 and MDA-MB-231 cells was evaluated using the MTT colorimetric procedure.[56] For in vitro cytotoxicity tests, cells were seeded in flat-bottom 96-well plates (Greiner Bio, NC, USA) at a density of 7.0 × 103 cells/well and kept at 37 °C in a humidified atmosphere of 5% CO2 for 24 h to attach the cells to the bottom of the plate. Metformin docusate and sodium docusate were dissolved in DMSO to obtain a stock solution, whereas the metformin hydrochloride stock solution was prepared in distilled water. The stock solutions were diluted with the cell culture medium (DMEM) to obtain different concentrations. Considering the high tolerability of metformin hydrochloride, a higher concentration (mM) was used for the cytotoxic assay. The cells were then treated with different concentrations of metformin docusate, sodium docusate, or metformin hydrochloride for 48 h. After 48 h, the old media were removed, and the cells were washed twice with PBS. The MTT solution (5 mg/mL) was diluted with fresh DMEM medium to obtain a final concentration of 0.5 mg/mL. The MTT reagent prepared in DMEM (100 μL) was added to the cells and incubated at 37 °C for 2 h. After 2 h of incubation, the medium was discarded, and the blue formazan crystals formed in the well were dissolved in DMSO, and the absorbance was measured at 570 nm using a microplate reader (Synergy H1, BioTek, Vermont, USA). The absorbance values obtained for the cells treated with metformin or metformin salts and control cells were used to calculate percent cell viability.