Harnessing Chemoselective Imine Ligation for Tethering Bioactive Molecules to Platinum(IV) Prodrugs

Abstract

Platinum(II)-based drugs, namely cisplatin, carboplatin and oxaliplatin, are the first line of treatment for many types of malignancies, including testicular, ovarian and lung cancer (Kelland in Nat Rev Cancer 7:573–584, [1]; Abu-Surrah and Kettunen in Curr Med Chem 13:1337–1357, [2]). However, their efficacy is severely limited by adverse side-effects due to high toxicities and incidences of drug resistance, either inherent or acquired (Fuertes et al. in Chem Rev 103:645–662, [3]).

Reproduced with permission from the Royal Society of Chemistry.

Supplementary Information

¹H NMR spectra of Pt-tyrosine hydrazide 2c. The asymmetry of the aromatic protons observed in the ¹H NMR spectra of 2c was initially puzzling as the protons H_a/H ¹_a , H_b/H ¹_b and H_d/H ¹_d exhibited greater magnetic inequivalence than was observed with the other Pt(IV)-imine conjugates where the same protons (though magnetically inequivalent) had closely overlapping chemical shifts. Consequently, in order to rule out the possibility of asymmetry of the axial ligands (eg. one side having E stereoisomerism with the other side being Z), we synthesized the purely organic hydrazone conjugate between 4-carboxylbenzaldehyde and tyrosine hydrazide for comparison. As shown in Fig. S2.6, the organic hydrazone displayed the same asymmetry of the aromatic protons. We postulated that this asymmetry arose due to slow rotation of the N=CH–Ph double bond in the NMR timescale, resulting in more distinctive magnetic inequivalence of H_b versus H ¹_b .

Fig. S2.1

¹H NMR spectra of platinum(IV)-benzaldehyde complex 1 in DMSO-d₆

Fig. S2.2

¹H NMR spectra of complex 1 in acetone-d₆ and ESI-MS characterization

Fig. S2.3

RP-HPLC assessment of purity of Pt-benzaldehyde 1 dissolved in DMF/H₂O. Elution conditions for both spectra (a) and (b): 20–80% gradient elution system with aq. NH₄OAc buffer (10 mM, pH 5.5) (solvent A) and MeCN (solvent B) over 15 min at 1.0 mL/min. Columns used are: a Phenomenex Luna C18(2) (250 × 4.60 mm i.d), b Shimpack VP-ODS column (150 × 4.60 mm i.d)

Fig. S2.4

Similarities between ¹H NMR spectrum of: a Pt-tyrosine hydrazide (2c) and b the purely organic hydrazone ligation between 4-carboxylbenzaldehyde and tyrosine hydrazide. The 3D visual illustration is generated by CORINA [46]

Synthesis of hydrazone conjugate between 4-carboxylbenzaldehyde and l-tyrosine hydrazide as NMR reference. 4-carboxylbenzaldehyde (21.5 mg, 0.143 mmol) was added to a solution of l-tyrosine hydrazide (140 mg, 0.716 mmol) in 50% DMF/H₂O (3 mL). The reaction mixture was lyophilized after 24 h and washed with H₂O to yield a white precipitate. See Fig. S2.4 for ¹H NMR spectra; Purity (HPLC): 92% at 254 nm.

Fig. S2.5

Only slight hydrolysis of Pt-Girard’s reagent T 2e was observed at pH 7.4 and 37 °C over 24 h; the λ_max of the spectra was at 303 nm, attributable to the hydrazone bond

Catalysis at physiological pH by p-anisidine. The reaction between Pt-benzaldehyde 1 (A) and benzhydrazide to form the mono-ligated Pt-benzhydrazide (B) and bis-ligated Pt-benzhydrazide 2a (C) product follows pseudo 1st order reaction kinetics in the presence of excess hydrazide and may be described by the following chemical equations.

The consumption of 1 and formation of the bis-conjugated product 2a at hourly intervals was quantified by integration at 254 and 280 nm. In order to obtain a smooth plot of [1] and [2a] as a function of time, the experimental data was curve-fitted to model the chemical equilibriums as shown in Fig. S2.6 and Table S2.1 using the chemical reactions module of Berkeley Madonna, a commercial graphical differential equation solver.

Fig. S2.6

Illustration of imine ligation between Pt-benzaldehyde 1 (a) and benzhydrazide to yield the mono-ligated product (b) and the bis-ligated product 2a (c)

In order to account for differences in molar absorptivity, the fraction of Pt-benzaldehyde (A) at any one point in time (t) was calculated as the current amount of A over the initial amount of A as quantified by HPLC integration. Similarly, the fraction of bis-ligated product (C) at any one point in time was calculated as the current amount of C over the maximum amount of C produced at the end of the reaction. Since it was difficult to extrapolate the maximum amount of mono-ligated product (B) from the experimental data, the fraction of B at any one point in time was calculated by substracting from the fraction of A and C. This is summarized as the follows (Table S2.1):

Table S2.1 Representative summary of experimental data for curve fitting

$$\begin{aligned} {\text{Fraction of A}} & = {\text{Average of}}\frac{Absorbance\,of\,A\,at\,t}{Inital\,Absorbance\,of\,A\,at\,t = 0}{\text{at}}\,254\,{\text{and}}\,280\;{\text{nm}} \\ {\text{Fraction of C}} & = {\text{Average of}}\frac{\text{Absorbance of C at t}}{{{\text{Absorbance of C at t}} = \infty }}{\text{at}}\,254\,{\text{and}}\,280\;{\text{nm}} \\ {\text{Fraction of B}} & = 1{-}\left( {{\text{Fraction of A}} + {\text{Fraction of C}}} \right) \\ \end{aligned}$$