Synthesis of dialkylaminoaryl phosphonate bioisosteres of tyrosine and tyramine: a novel application of allene phosphonate chemistry for the synthesis of false substrates of tyrosinase -

Tyrosinase is the rate-limiting oxidase in the synthesis of melanin, making it an obvious target for the treatment of melanotic melanomas. Tyrosine and tyramine are its natural substrates, but many of their derivatives are inhibitors or false substrates, and are therefore prime candidates for melanoma chemotherapy. A series of dialkylphosphonate derivatives of tyramine have now been synthesized in order to extend the chemical diversity of tyrosinase substrates. The known reactivity between alkenephosphonates and nucleophiles was exploited by the addition of 4-(2-aminoethyl)phenol (tyramine) across the 2,3-double bond of dialkyl 1,2-alkadiene phosphonates, to obtain the desired bisphosphonate derivatives. These reactions were highly chemoselective and regioselective but not stereoselective. Five of the reported novel dialkylphosphonate aminophenols were substrates for mushroom tyrosinase in vitro: dimethyl 2-[2(4-hydroxyphenyl)ethylamino]-3-methyl-1-butene phosphonate (3); diethyl 2-[2(4-hydroxyphenyl)ethylamino]-3-methyl-1-butene phosphonate (4); dimethyl 2-[2-(4-hydroxyphenyl)ethylamino]-2-cyclohexyl-1-ethene phosphonate (5); diethyl 2-[2-(4-hydroxyphenyl)ethylamino]-2-cyclohexyl-1-ethene phosphonate (6); diethyl 2-[2-(4-hydroxyphenyl)ethylamino]ethane phosphonate (7). Compound 3 blocked the pigmentation of anagen hair in vivo in a murine animal model, a further demonstration that these compounds are able to enter and disrupt the melanogenic pathway.


INTRODUCTION
Malignant melanoma is among the most aggressive and drug resistant cancers.It arises via a poorly understood interplay of host-environment interactions and risk factors.Sunlight, mainly UV B (280-320 nm), appears to be one of the most important environmental causative factors, although the relationship between sun exposure and the development of melanoma is not simple (Etzkorn et al., 2013;Braeuer et al., 2014).Localized melanoma can be managed and most cases even cured by early surgical ablation (Urist, 1996) but treatment for the disseminated disease remains elusive.An extensive literature describes clinical and experimental approaches to the .chemotherapy, immunotherapy, endocrine therapy and molecular approaches to control melanoma (Garbe et al., 2011;An et al., 2009).Chemotherapy based on alkylating agents (e.g., dacarbazine; DTIC), nitrosoureas and platinum complexes all have low response rates, are moderately to severely toxic, and have serious side-effects.In short, there is no accepted standard regimen for the treatment of metastatic melanoma (Voskoboynik and Arkenau, 2014;NCI, 2013).One of the defining characteristics of most melanomas is an increase in melanin pigmentation, making melanin an obvious target for anti-melanoma therapy (Slomonski et al., 2015;Farmer et al., 2003).The production and distribution of melanin (melanogenesis) by melanocytes in skin and hair follicles involves packaging melanin in specialized lysosome-like organelles (melanosomes) which contain several enzymes that mediate the production of melanin.Tyrosinase (TYR) initiates melanin synthesis by catalyzing the hydroxylation of L-tyrosine to L-3,4-dihydroxyphenylalanine (DOPA) and oxidation of DOPA to to DOPA quinone.The pathway diverges to either eumelanin (red and amber) or pheomelanin (black), depending on whether dopaquinone interacts with cysteine or other related sulfhydrylcontaining moieties.Because the melanogenic pathway is found only in melanocytes, it is an attractive, selective target not only for therapeutic intervention in cancer but also for cosmetic control of skin pigmentation (Speeckaert et al., 2014;Regad, 2013).Tyrosinase inhibitors or false substrates of tyrosinase represent one approach to the development of anti-melanin drugs (Jimbow et al., 1993).Several classes of tyrosinase inhibitors / false substrates with potential antimelanotic activity have been identified, including 2-and 4-sulfur-substituted phenols (Riley et al, 1997;Jimbow et al., 2011).
The literature is replete with reports of the use of bisphosphonate drugs for treating osteoporosis (Reid, 2011) and to a lesser extent, of nucleotide phosphonate analogues for the treatment of viral disease (Yim and Hwang, 2013).However, there are very few other reported applications of phosphonates, or indeed phosphorous-containing drugs in modern medicine.As early as the 1960's, it was reported that acetylene phosphines obtained through the reaction of phosphorus trichloride with acetylenic alcohols spontaneously isomerized under mild conditions to form allenephosphonates (Ignat'ev et al., 1969;Angelov and Neilson, 2992;Neilson and Angelov, 2009).Given that phosphorylated allenes are quite reactive and interact with both electrophilic and nucleophilic reagents, this simple method provided a synthetic route to many compounds with a range of substituents on phosphorus.The reactions with electrophiles have been studied in detail (Khusainova and Pudovik, 1987;Angelov, 1983;Khusainova and Pudovik, 1978;Alabigin and Brel, 1997), while interactions with nucleophiles have been less thoroughly explored.One of the historical reasons for the limited nucleophilic chemistry with these compounds is that most nucleophilic reagents require the use of a catalyst to drive the reaction.An exception to this rule is found in the reactions with amines, where allene phosphorus compounds react directly.The first example of the reaction of amines with substituted allenephosphonic acid derivatives was published in 1966 (Pudovik and Khusainova, 1966) when it was shown that diethylamine and piperidine add to dialkyl 3-methyl-1,2-butadienephosphonates on the -double bond to afford 1,2-aducts (Scheme 1).
More recently, nucleophilic additions to allene phosphonium salts, phosphine oxides and diethyl allene phosphonate have been reported to yield a wide variety of primary amines.For example, addition of aminoethanol occurs at the double bond, with isolation of the 2,3-adducts in all cases (Scheme 2) (Palacios et al, 1996).The current work is based on the premise that incorporation of phosphonate moieties into the alkylamino portion of the tyrosinase substrates could potentially open the door to novel antimelanotic compounds through changes in target enzyme selectivity, pharmacokinetics, target-selective delivery and processing along the melanogenic pathway.The synthesis of the novel phosphonate derivatives now reported is based on the reactivity between amines, specifically tyramine analogues, and dialkylallene phosphonates.The design and synthesis of several structural phosphonate analogues of tyrosine is now reported, along with preliminary evidence for their oxidation by tyrosinase and their inhibition of melanogenesis.

Chemistry
The objective of the current investigations was to synthesize N-substituted tyramine phosphonates, which in theory could proceed via either or both the amino and hydroxyl groups.The high acidity of the phenolic hydroxyl could potentially promote reaction with the allenic system of 1,2-alkadiene phosphonates via an electrophilic mechanism (Alabugin and Brel, 1997), but current experimental data from reactions between several 1,2-alkadiene phosphonates and 4-(2-aminoethyl)phenol (tyramine) show that the reaction is highly chemo-and regioselective, but not stereoselective, and proceeds only by addition of the amino group to the 2,3-double bond (Scheme 3).
Experimentally, 1,2-alkadiene phosphonates were heated with amines under reflux in acetonitrile (Table 1).Reaction progress was monitored with 31 P-NMR.After removing the solvent, crude product could be recrystallized as a solid compound.
In all cases, the 2,3-adduct was obtained as a mixture of E and Z isomers (Scheme3), but it was not possible to separate either isomer in pure form by recrystallization.Thus, in a representative case, the 31 P-NMR spectrum of crude reaction product for 4, for example, included two peaks at  P 29.07 (Eisomer) and 29.6 (Z-isomer) ppm (Palacios et al., 1996), in an approximate 2:1 isomeric ratio.After repeated recrystallization, the ratio was approximately 6:1 in favor of the E-isomer.The 1 H and 13 C NMR spectra also displayed two groups of signals, confirming the structures as E-and Z-isomer mixtures.

In vitro cytotoxicity
MTT cell survival assays were implemented using test compounds at concentrations from 1x10 -1 to 1x10 -7 mM.There was no evidence of growth inhibition at concentrations below 0.1 mM (data not shown).The low cost and convenience of the MTT assay is offset by caveats including its tendency to overestimate survival, its relative insensitivity to small changes and, most importantly, it measures mitochondrial activity (i.e.metabolism) .rather than viability per se.Although the MTT assay is a convenient screening procedure, interested readers are referred to Cobb's review on cell viability testing (Cobb, 2013).

Oxidation by mushroom tyrosinase
A number of antimicrobial and anticancer αaminophosphonates (Dake et al., 2011;Gu and Jin, 2012;Abdel-Megeed et al., 2012) andβ-keto-andβ-hydroxyphosphonates (Cui et al., 2008) have been reported, but none have been tested against tyrosinase.Nine of the currently-reported phosphorylated amines were incubated with tyrosinase, and of these, five (3-7) were mushroom tyrosinase substrates in vitro as judged by visual inspection and UV-vis spectrophotometry.In this screen, active compounds developed a light brownish color (Plate 1A) and display an additional UV absorbance peak at 289-292 nm (data not shown).As anticipated, amines which contained a blocked aromatic hydroxyl group (8, 9), or no phenolic hydroxyl (1, 2), were inactive in these tests.This finding demonstrates the vital nature of the aromatic hydroxyl group in the design of tyrosinase substrates and false substrates and shows that the phosphonate moiety does not interfere with tyrosinase-mediated oxidation of the phenolic component of these molecules.The natural tyrosinase substrates (tyramine; tyrosine) served as positive controls in these experiments, and solvent and unreacted diene phosphonates provided negative results, thereby confirming that the observed effects were due to test compound-enzyme interaction (Table 1).

Inhibition of in vivo melanin biosynthesis
A preliminary study of compound 3 in a C57BL/6J mouse provided a clear indication that these compounds are false substrates of tyrosinase once they enter the melanogenic pathway.Black hair were manually plucked from a small region on the backs of mice in order to stimulate new follicular melanocytes and increase tyrosinase activity.This allows visualization of the blockage of melanin synthesis through the regrowth of unpigmented (white) new (anagen) hair (Plate 1B).In this test, .the mouse was given daily intraperitoneal (i.p.) injections of 3 (0.878 mmol/ 300 mg/kg) daily for 14 days, a dosage regimen based on literature data for other putative false substrates (Ito et al., 1987).
None of the amino-or thiol adducts were toxic to either G361 or SK-MEL-24 melanoma cells in cell culture at inhibitory concentrations (IC 50 ) below 0.5x10 -4 M.This low toxicity was also evident in the in vivo study, although the in vivo experiment was not intended to provide toxicity data.

S. No. Compound
Precursor Amine Precursor Phosphonate Tyrosinase assay In conclusion, the facile synthesis of dialkyl aminoarylphosphonates as potential substrates of tyrosinase proceeded by reaction of the appropriate amine (i.e., tyramine, 4methoxyphenethyl amine orbenzylamine) with dialkyl 1,2alkadienephosphonates.In contrast, the alkyl amines afforded only 2,3-adducts and the thiophenols produced only 1,2-adducts under similar reaction conditions.Five of the amino adduct tyramine derivatives, 3, 4, 6, 7 and 8 tested positive for oxidation by mushroom tyrosinase, and compound 3, in a preliminary study inhibited melanogenesis in vivo in a murine model.

Materials and methods
All chemical manipulations were performed under an atmosphere of dry argon.The 1,2-alkadienephosphonate precursors were prepared by a known procedure (Ignat'ev et al., all other starting compounds were obtained from commercial sources (Aldrich) and were distilled or recrystallized before use.Solvents were dried and distilled before use and stored over molecular sieves. 1 H, 13 C and 31 P NMR spectra were recorded on a Bruker AM 300 instrument operating at 300.13, 75.47 and 121.53 MHz for the respective nuclei using 85 % H 3 PO 4 and Si(CH 3 ) 4 as external standards.
In all spectroscopic studies CDCl 3 was used as both a solvent and as an internal lock.Positive shifts lie downfield of the standard in all cases.

Chemical syntheses General Procedure
A suitable round bottom flask equipped with a stir bar, rubber septum and argon inlet was charged with distilled anhydrous solvent (stored over molecular sieves) and the respective dialkyl phosphonate and amineprecursors.
The reaction mixture was heated under reflux, usually overnight, until 31 P NMR showed the reaction was complete.The solvent was removed under reduced pressure and the residue was worked up by direct recrystallization, chromatography followed by recrystallization, or distillation under vacuum followed by crystallization to obtain pure product.

Cytotoxicity
The MTT (methyltetrazolium; 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyl tetrazolium bromide) colorimetric assay was used to estimate the in vitro toxicity of the test compounds.G361 and SK-Mel-24human melanoma cell lines used in the drug challenges were cultured in RPMI 1640 culture medium supplemented with 10% FBS and 1% PSF at 37 ºC in a humidified atmosphere of 5 % CO 2 in air.The studies were carried out in 96 well microtitre plates inoculated on day 1 with 1x10 4 cells per well in culture medium (100 L).Twenty-four hours after plating serial dilutions of test compounds were added and the cells were incubated for an additional 72 hours under identical culture conditions.At the end of this period the culture medium containing the test compound was removed from the wells and replaced with RPMI 1640 culture medium (100 L) containing 0.5 mg/mL MTT.The plates were incubated for 4 hours as above, then the MTT-containing medium was removed and replaced with isopropyl alcohol (100 L) for 30 min to dissolve any MTT crystals in the cells.The plates were read at 540 nm on a Beckman Coulter DU640 Spectrophotometer ELISA Plate Reader.The number of viable cells per well were calculated by interpolation from the individual optical density readings, which are proportional to the number of surviving cell numbers (Mosmann, 1983).

Mushroom tyrosinase assay
Oxidation of the novel dialkyl phosphonate compounds by mushroom tyrosinase was determined using a screening assay adapted from the literature (Kahn and Andrawis, 1986).Substrate concentrations on the order of 1 x 10 -4 M were incubated with mushroom tyrosinase (4.04 x 10 -4 g/mL) in PBS (pH 6.8) at 37 o C for 15 min in a total volume of 2 mL PBS (pH 6.8).Incubation controls included PBS (pH 6.8; negative control), tyramine and DL-tyrosine in PBS (pH 6.8; positive controls), and two nonphenolic dialkyl phosphonates, bis(methylazeridinyl) 3-methyl-1,2-butadiene phosphonate and dimethyl 2-cyclohexylidene ethane phosphonate.The reaction mixtures were cooled with ice, post incubation to stop reaction and then visualized to determine color changeand UV spectrophotometric analysis from 200-600 nm to check for presence of a chromophore in the 289-292 nm region.

In vivo inhibition of melanogenesis
Animal studies were approved by the University of Alberta Animal Care Committee in accordance with the tenets of the Canadian Council on Animal Care.
Preliminary testing of model compound 3 was carried out on a single C57BL/6J black mouse (University of Alberta Animal Services).Using a literature method to test for inhibition of melanin synthesis in anagen hair, the black hair were manually plucked from a small area on the back of the mouse under light fluothane anesthesia; this process initiates new anagen growth and activates follicular melanocytes to increase tyrosinase activity (Alena et al, 1990).Compound 3, dissolved in a solution of tween and P.E.G (1:1), was administered daily for 14 days by intraperitoneal (i.p.) injection at a dose of 300 mg/kg body weight.