Comparison of a Tartaric Acid Derived Polymeric MRI Contrast Agent

Dec 20, 2007 - Contrast agents with high relaxivity are needed to increase the sensitivity .... Imaging: The Interplay between Size, Function, and Pha...
1 downloads 0 Views 105KB Size
24

Bioconjugate Chem. 2008, 19, 24–27

Comparison of a Tartaric Acid Derived Polymeric MRI Contrast Agent to a Small Molecule Model Chelate Robie L. Lucas, Michael Benjamin, and Theresa M. Reineke* Department of Chemistry, University of Cincinnati, Cincinnati, Ohio 45221-0172. Received October 6, 2007

Contrast agents with high relaxivity are needed to increase the sensitivity of magnetic resonance imaging (MRI) for novel clinical and research applications. For this reason, polymeric structures containing multiple Gd(III) chelates are of current interest. Described in this communication are the syntheses and characterization of a glycopolymer derived from L-tartaric acid, Gd4(H2O), as well as a low molecular weight compound, Gd10(H2O), that models the Gd(III) chelate structure in the repeat unit of polymer Gd4(H2O). Luminescence lifetime measurements in H2O and D2O for Eu(III) analogues of Gd4(H2O) and Gd10(H2O) [named Eu4(H2O) and Eu10(H2O)] reveal that the lanthanide in both structures likely has one water ligand in the primary coordination sphere. The relaxivity of the model chelate Gd10(H2O) at 400 MHz and 310 K was determined to be 4.7 mmol-1 · s-1, representing a nearly 50% increase over Magnevist (3.2 mmol-1 · s-1). Relaxivity values on a per Gd basis for the polymeric structure Gd4(H2O) prepared at two degrees of polymerization, n ) 12 and 19, are similar, but slightly lower than Gd10(H2O) (4.4 mmol-1 · s-1 and 4.5 mmol-1 · s-1, respectively). However, their molecular relaxivities of 51 mmol-1 · s-1 and 80 mmol-1 · s-1, respectively, provide a substantial increase over that of Magnevist.

The utility of magnetic resonance imaging (MRI) is currently being expanded beyond its role as the leading imaging modality in diagnostic medicine (1). MRI is noninvasive and provides real-time images, which make it an ideal research tool for visualizing drug delivery (2, 3), monitoring biological processes, and following functional changes in vivo (4). To reach the full potential of MRI for such advanced imaging applications requires that the inherently low sensitivity of MRI be modulated with contrast agents containing paramagnetic metal ions, such as Gd(III) (5, 6). However, current clinically used contrast agents based on low molecular weight Gd(III) chelates, such as Magnevist, do not sufficiently enhance water proton relaxation rates enough to function for these purposes (7), and their imageenhancing properties dramatically diminish at the increasingly higher magnetic fields of modern MRI instruments (8). In fact, the advanced applications described above will likely require macromolecular systems with multiple Gd(III) centers (6), which may retain their high relaxivities at high magnetic field strengths. Several design strategies have been used to incorporate multiple Gd(III) chelates within macromolecular structures, including linear polymers (9–12), dendritic systems (13–16), carbon nanotubes (17), and self-assembled bimetallic complexes (18, 19). For example, Kiessling and co-workers recently reported a linear polymer prepared via ring-opening metathesis polymerization that contained hydroxypyridonate (HOPO)-based Gd(III) chelates (12). Incorporating the high relaxivity HOPO-Gd(III) chelates (20) into a polymeric structure provided a contrast agent with per Gd and molecular relaxivities substantially higher than those of clinical contrast agents at 60 MHz. This example clearly demonstrates the ability of multivalent systems to significantly enhance relaxation properties. However, the influence of specific features of molecular structure in polymeric systems on relaxation properties has not been clearly established, particularly at higher magnetic field strengths (21). Our previous work has demonstrated that incorporating carbohydrate residues along a polymer backbone decreases the * Corresponding author E-mail: [email protected]

Figure 1. Target structures of the tartaramide-based polymer Gd4(H2O) and model chelate Gd10(H2O).

toxicity of gene delivery vectors (22). Here, we extend this strategy to create a system that contains repeated carbohydrates with Gd(III) chelates along the polymer backbone. Incorporating carbohydrate-based components within the repeat unit increases hydrophilicity and may serve to enhance biocompatibility. Herein, a linear polymer, Gd4(H2O), has been prepared that contains a repeat unit consisting of a tartaric acid diamide moiety, an ethylenediamine-derived linker, and a Gd(III)DTPA(bisamide) chelate (Figures 1 and 2). In addition, to assist our understanding of the coordination environment of this polymer, a low molecular weight chelate, Gd10(H2O), intended to model the Gd(III) chelate within the polymer repeat unit has been synthesized. These structures are allowing us to investigate the influence of polymer and repeat unit structure on the relaxivity properties of glycopolymer contrast agents. The syntheses, characterization, relaxivities, and water coordination properties of these macromolecular systems are presented. We have developed a methodology to prepare carbohydratediamine monomers that undergo step-growth polymerization with diethylenetriaminepentaacetic acid-bisanhydride (DTPABA) to yield novel glycopolymers that chelate Gd(III). To

10.1021/bc700375m CCC: $40.75  2008 American Chemical Society Published on Web 12/20/2007

Communications Scheme 1. Synthetic Protocol for the Polymer Gd4(H2O)a

a (i) Mono-N-Boc-ethylenediamine, MeOH, ∆; (ii) 4.0 M HCl in 1,4-dioxane; (iii) DTPA-BA, DMSO, 24 or 72 h; (iv) GdCl3 · 6H2O, NaOH, H2O, pH 6-7.

prepare the carbohydrate-diamine monomer 3, dimethyl Ltartrate 1 was refluxed with 2 equiv of mono-N-Boc-ethylenediamine in MeOH, which provided the Boc-protected tartaric acid diamide 2 in 67% yield (Scheme 1). Treatment of the diamide 2 with a solution of HCl in 1,4-dioxane to remove the Boc groups, followed by purification of the resulting diamine hydrochloride salt via ion exchange chromatography, yielded the pure monomer 3 in quantitative yield. The polymer precursors H34 were synthesized by condensing 3 with DTPABA in DMSO at room temperature, and the molecular weight was varied by increasing the polymerization time. After the polymerizations were complete, the polymers were purified by exhaustive dialysis against ultrapure water using 3500 MWCO membranes and were then lyophilized to dryness. Formation of Gd4(H2O) occurred by treatment of the polymers H34 with GdCl3 in aqueous solution, while maintaining a pH of 6–7 with aqueous NaOH. Following Gd(III) chelation, the polymers were again purified by exhaustive dialysis in 1000 MWCO membranes against ultrapure water to remove any unchelated Gd(III) ions and other salts. Characterization of Gd4(H2O) was accomplished by gel permeation chromatography (GPC) using a triple detection system of refractive index, viscometry, and static light scattering. Two polymers were synthesized with weightaveraged molecular weights (Mw) of 8.9 kDa and 14.0 kDa, which correspond to degrees of polymerization of n ) 12 and n ) 19, respectively. The polymers revealed relatively low polydispersity indices for a step-growth mechanism (1.4 and 1.6, respectively), which is likely an artifact of the dialysis step. The Mark–Houwink-Sakurada R values (0.6–0.7) obtained from viscometry data demonstrated the formation of linear, randomly coiled polymers. Synthesis of the small molecule chelate to model the polymer repeat unit required the preparation of an unsymmetrical tartaric

Bioconjugate Chem., Vol. 19, No. 1, 2008 25

acid diamide with only a single primary amine (compound 9 in Scheme 2). To prepare this compound, we made use of the known di-O-acetyl protected tartaric anhydride 6, which was prepared from L-tartaric acid 5 in 81% yield by refluxing in acetic anhydride in the presence of sulfuric acid (23). The anhydride 6 was subsequently converted to the N-methyltartaramic acid 7 in 75% yield by treatment of the anhydride with a solution of methylamine in THF. This sequence was deemed more practical than attempting a selective monoamidation of L-tartaric acid directly. Standard coupling conditions using N,N′dicyclohexylcarbodiimde (DCC) were used for amide bond formation between tartaramic acid 7 and mono-N-Boc-ethylenediamine to provide the diamide 8 in 47% yield. Simultaneous removal of the Boc (24) and acetyl (25) protecting groups was then achieved with 3 equiv of p-toluenesulfonic acid monohydrate in a CH2Cl2-MeOH solvent system. The crude product resulting from this reaction was purified and neutralized by ion exchange chromatography, giving the unsymmetrical tartaric acid diamide 9 in 91% yield. The ligand of the model chelate H310 was then obtained in 65% yield by reacting 2 equiv of 9 with DTPA-BA in DMF at 65 °C. The ESI mass spectrum of H310 showed only two significant peaks at m/z ) 768 and m/z ) 385, corresponding to the proton adduct [H310 + H]+ of the desired product and its doubly charged derivative [H310 + 2H]2+, respectively. The model chelate Gd10(H2O) was prepared by treating an aqueous solution of the ligand H310 with GdCl3 at a pH of 6–7 and purifying by exhaustive dialysis in a 1000 MWCO membrane against ultrapure water. The major peak in the ESI mass spectrum corresponded to the proton adduct of the anhydrous chelated product {[Gd10 + H]+, m/z ) 923}. Complete loading of Gd(III) was also evidenced by the absence of a peak at m/z ) 768 (the proton adduct of the unchelated ligand H310). Knowledge of the structure of a Gd(III) chelate in aqueous solution is important for understanding the factors that dictate its capacity to enhance proton relaxation. Of particular importance is knowing the number of water molecules bound in the primary coordination sphere, q, of the Gd(III) center. The isostructural nature of Ln(III) complexes with a common chelating ligand is useful for probing the solution structure of Gd(III) chelates (7). Replacement of Gd(III) with Eu(III) makes it possible to determine q by using the proportionality of their corresponding luminescence decay rates in H2O and D2O (26). With this in mind, the Eu(III) analogues of the model chelate and the polymer were synthesized in a fashion similar to the related Gd(III) structures. Analysis of Eu4(H2O) via GPC indicated a polymer with a Mw value of 18.6 kDa (25 repeat units). Luminescence lifetimes were measured for both Eu(III) structures in H2O and D2O, and q values were calculated using an equation developed by Horrocks et al., [q ) 1.11(kH2O kD2O - 0.31 + 0.075nO)CNH)] (27). These experiments revealed the same q values of 1.2 for both the Eu(III) model chelate, Eu10(H2O), and for the Eu(III) polymer, Eu4(H2O). It should be noted that these results (Table 1) are in good agreement with other Ln(III)-DTPA derivatives with a single water molecule in the primary coordination sphere (27). Relaxivity values for the model chelate Gd10(H2O) and the polymers Gd4(H2O) were determined at 400 MHz (9.4 T) and 310 K. For comparison to a clinical contrast agent at this magnetic field, we measured the relaxivity, r1, of Magnevist and obtained a value of 3.2 mmol-1 · s-1. The relaxivity value of the model chelate Gd10(H2O) was found to be 4.7 mmol-1 · s-1, which represents an approximately 50% increase over Magnevist. The relaxivity values for the polymeric chelates Gd4(H2O) on a per Gd basis are slightly lower than that obtained for the small molecule model chelate, with values of 4.4 mmol-1 · s-1 (for n ) 12) and 4.5 mmol-1 · s-1 (for n ) 19).

26 Bioconjugate Chem., Vol. 19, No. 1, 2008

Communications

Figure 2. Per Gd(III) ion (A) and per molecule (B) relaxivities of Magnevist, the small molecule model chelate Gd10(H2O), and the polymeric system Gd4(H2O) with 12 and 19 repeat units. Scheme 2. Synthesis of the Model Chelate Gd10(H2O)a

a

(i) Acetic anhydride, H2SO4, ∆; (ii) 2.0 M methylamine in THF; (iii) N,N′-dicyclohexylcarbodiimide, CH2Cl2, 0 °C to rt; (iv) ptoluenesulfonic acid monohydrate, MeOH-CH2Cl2 (1:9); (v) DTPA-BA, DMF, 65 °C; (vi) GdCl3 · 6H2O, NaOH, H2O, pH 6-7. Table 1. Luminescence Lifetime Data for the Polymer Eu4(H2O) and Model Chelate Eu10(H2O)a compound

τH2O (ms)

τD2O (ms)

qa

Eu4(H2O) Eu10(H2O)

0.60 0.62

2.3 2.5

1.2 1.2

a The polymer Eu4(H2O) consists of 25 repeat units based on Mw from GPC analysis. Values of q were determined using the equation developed by Horrocks et al. (27).

These relaxivity values suggest that the relatively short polymers have considerable internal flexibility, despite incorporating the DTPA-based Gd(III) chelates directly into the polymer backbone. In addition, the lack of rigidity may facilitate electronic relaxation via Gd-Gd interactions and decrease relaxivity (28). Merbach et al. previously synthesized a series of related linear polymers prepared by polymerizing DTPA-BA with simple

straight-chain alkyldiamines, where the number of methylenes between the amino groups was 4, 6, 10, or 12 (9). The high relaxivities exhibited by the polymers formed from diamines with 10 and 12 methylene groups was attributed to the formation of micelle structures with strong hydrophobic interactions, a consequence of using lipophilic alkyldiamine monomers. The polar diamine monomer 3 used in the present study is hydrophilic and would not form micelle-like structures in aqueous solution. Besides the influence of rotational properties, the limited relaxivity gains for Gd4(H2O) and Gd10(H2O) versus Magnevist may also be due to the well-established slow water exchange rate of Gd(III)-DTPA(bisamides) (29). Despite the fact that the per Gd relaxivity does not increase for the polymers of the present study versus the model chelate, the molecular relaxivities are very high for the polymers of n ) 12 and n ) 19, with values of 51 mmol-1 · s-1 and 80 mmol-1 · s-1, which represent 16- and 25-fold increases, respectively, over the molecular relaxivity of Magnevist. Molecular relaxivities for polymeric systems are substantial, as they demonstrate the potential for higher Gd concentration to increase local proton relaxation in the vicinity of the contrast agent. Advanced biomedical and molecular imaging applications using MRI contrast agents will likely require incorporating multiple Gd(III) chelates into macromolecular structures (6, 21, 30). We have presented the synthesis, characterization, and properties of two novel contrast agents, a small molecule model chelate, Gd10(H2O), and a polymeric structure, Gd4(H2O), which were formed by coupling tartaric acid derived amines with DTPA-BA. The relaxivity properties of the Gd(III) chelates within the polymeric structures are similar to that of their low molecular weight analogue Gd10(H2O) on a per Gd basis; however on a molecular basis, the polymers clearly yielded substantially higher relaxivities. The generality of the synthetic protocol for these polymers is currently being utilized to generate a series of related polymeric materials using different carbohydrate-based monomers. The influence of carbohydrate structural variation within this series is currently under investigation.

ACKNOWLEDGMENT The authors thank Drs. Jing-Huei Lee, Wen-Jang Chu, and Elwood Brooks for assistance with inversion–recovery experiments. This work was supported by the National Institutes of Health (1-R21-EB007244-01), the University of Cincinnati Department of Chemistry, and University of Cincinnati Chemical Sensors Group. Supporting Information Available: Synthetic procedures and analytical data for 2–9, Gd4(H2O), Eu4(H2O), H310,

Communications

Gd10(H2O), and Eu10(H2O). Experimental procedures for relaxivity and luminescence lifetime measurements. This material is available free of charge via the Internet at http:// pubs.acs.org.

LITERATURE CITED (1) Rudin, M., and Weissleder, R. (2003) Molecular imaging in drug discovery and development. Nat. ReV. 2, 123–131. (2) Richardson, J. C., Bowtell, R. W., Mäder, K., and Melia, C. D. (2005) Pharmaceutical applications of magnetic resonance imaging (MRI). AdV. Drug DeliVery ReV. 57, 1191–1209. (3) Ye, F., Ke, T., Jeong, E.-K., Wang, X., Sun, Y., Johnson, M., and Lu, Z.-R. (2006) Noninvasive visualization of poly(Lglutamic acid) using contrast-enhanced MRI. Mol. Pharm. 3, 507–515. (4) Louie, A. Y., Hüber, M. M., Ahrens, E. T., Rothbächer, U., Moats, R., Jacobs, R. E., Fraser, S. E., and Meade, T. J. (2000) In vivo visualization of gene expression using magnetic resonance imaging. Nat. Biotechnol. 18, 321–325. (5) Aime, S., Barge, A., Cabella, C., Crich, S. G., and Gianolio, E. (2004) Targeting cells with MR imaging probes based on paramagnetic Gd(III) chelates. Curr. Pharm. Biotechnol. 5, 509– 518. (6) Caravan, P. (2006) Strategies for increasing the sensitivity of gadolinium based MRI contrast agents. Chem. Soc. ReV. 35, 512– 523. (7) Caravan, P., Ellison, J. J., McMurry, T. J., and Lauffer, R. B. (1999) Gadolinium(III) chelates as MRI contrast agents: structure, dynamics, and applications. Chem. ReV. 99, 2293–2352. (8) Tóth, É., Helm, L., and Merbach, A. E. (2002) Relaxivity of MRI contrast agents. Top. Curr. Chem. 221, 61–101. (9) Tóth, É., Helm, L., Kellar, K. E., and Merbach, A. E. (1999) Gd(DTPA-bisamide)alkyl copolymers: a hint for the formation of MRI contrast agents with very high relaxivity. Chem. Eur. J. 5, 1202–1211. (10) Ladd, D. L., Hollister, R., Peng, X., Wei, D., Wu, G., Delecki, D., Snow, R. A., Toner, J. L., Kellar, K., Eck, J., Desai, V. C., Raymond, G., Kinter, L. B., Desser, T. S., and Rubin, D. L. (1999) Polymeric gadolinium chelate magnetic resonance imaging contrast agents: design, synthesis, and properties. Bioconjugate Chem. 10, 361–370. (11) Duarte, M. G., Gil, M. H., Peters, J. A., Colet, J. M., Vander Elst, L., Muller, R. N., and Geraldes, C. F. G. C. (2001) Synthesis, characterization, and relaxivity of two linear Gd(DTPA)-polymer conjugates. Bioconjugate Chem. 12, 170–177. (12) Allen, M. J., Raines, R. T., and Kiessling, L. L. (2006) Contrast agents for magnetic resonance imaging synthesized with ring-opening metathesis polymerization. J. Am. Chem. Soc. 128, 6534–6535. (13) Wiener, E. C., Brechbiel, M. W., Brothers, H., Magin, R. L., Gansow, O. A., Tomalia, D. A., and Lauterbaur, P. A. (1994) Dendrimer-based metal chelates: a new class of magnetic resonance imaging contrast agents. Magn. Reson. Med. 31, 1–8. (14) Bryant, L. H., Jr, Brechbiel, M. W., Wu, C., Bulte, J. W. M., Herynek, V., and Frank, J. A. (1999) Synthesis and relaxometry of high-generation (G ) 5, 7, 9, and 10) PAMAM dendrimerDOTA-gadolinium chelates. J. Magn. Reson. Imaging 9, 348– 352. (15) Fulton, D. A., Elemento, E. M., Aime, S., Chaabane, L., Botta, M., and Parker, D. (2006) Glycoconjugates of gadolinium complexes for MRI applications. Chem. Commun. 1064–1066.

Bioconjugate Chem., Vol. 19, No. 1, 2008 27 (16) Jászberényi, Z., Moriggi, L., Schmidt, P., Weidensteiner, C., Kneuer, R., Merbach, A. E., Helm, L., and Tóth, É. (2007) Physicochemical and MRI characterization of Gd3+-loaded polyamidoamine and hyperbranched dendrimers. J. Biol. Inorg. Chem. 12, 406–420. (17) Sitharaman, B., Kissell, K. R., Hartman, K. B., Tran, L. A., Baikalov, A., Rusakova, I., Sun, Y., Khant, H. A., Ludtke, S. J., Chiu, W., Laus, S., Tóth, É., Helm, L., Merbach, A. E., and Wilson, L. J. (2005) Superparamagnetic gadonanotubes are highperformance MRI contrast agents. Chem. Commun. 3915–3917. (18) Livramento, J. B., Tóth, É., Sour, A., Borel, A., Merbach, A. E., and Ruloff, R. (2005) High relaxivity confined to a small molecular space: a metallostar-based, potential MRI contrast agent. Angew. Chem., Int. Ed. 44, 1480–1484. (19) Pierre, V. C., Botta, M., Aime, S., and Raymond, K. N. (2006) Fe(III)-templated Gd(III) self-assemblies—a new route toward macromolecular MRI contrast agents. J. Am. Chem. Soc. 128, 9272–9273. (20) Raymond, K. N., and Pierre, V. C. (2005) Next generation, high relaxivity gadolinium MRI contrast agents. Bioconjugate Chem. 16, 3–8. (21) Bottrill, M., Kwok, L., and Long, N. (2006) Lanthanides in magnetic resonance imaging. Chem. Soc. ReV. 35, 557–571. (22) Liu, Y., and Reineke, T. M. (2005) Hydroxyl stereochemistry and amine number within poly(glycoamidoamine)s affect intracellular DNA delivery. J. Am. Chem. Soc. 127, 3004–3015. (23) Dobashi, Y., and Hara, S. (1987) A chiral stationary phase derived from (R,R)-tartramide with broadened scope of application to the liquid chromatographic resolution of enantiomers. J. Org. Chem. 52, 2490–2496. (24) Goodacre, J., Ponsford, R. J., and Stirling, I. (1975) Selective removal of the t-butyloxycarbonyl protecting group in the presence of t-butyl and p-methoxybenzyl esters. Tetrahedron Lett. 3609–3612. (25) González, A. G., Brouard, I., León, F., Padrón, J. I., and Bermejo, J. (2001) A facile chemoselective deacetylation in the presence of benzoyl and p-bromobenzoyl groups using ptoluenesulfonic acid. Tetrahedron Lett. 42, 3187–3188. (26) Horrocks, W. D., Jr., and Sudnick, D. R. (1979) Lanthanide ion probes of structure in biology. Laser-induced luminescence decay constants provide a direct measure of the number of metalcoordinated water molecules. J. Am. Chem. Soc. 101, 334–340. (27) Supkowski, R. M., and Horrocks, W. D., Jr. (2002) On the determination of the number of water molecules, q, coordinated to europium(III) ions in solution from luminescence decay lifetimes. Inorg. Chim. Acta 340, 44–48. (28) Tóth, É., Helm, L., Merbach, A. E., Hedinger, R., Hegetschweiler, K., and Jánossy, A. (1998) Structure and dynamics of a trinuclear gadolinium(III) complex: the effect of intramolecular electron spin relaxation on its proton relaxivity. Inorg. Chem. 37, 4104–4113. (29) Aime, S., Botta, M., Fasano, M., Paoletti, S., Anelli, P. L., Uggeri, F., and Virtuani, M. (1994) NMR evidence of a long exchange lifetime for the coordinated water in Ln(III)-bis(methyl amide)-DTPA complexes (Ln ) Gd, Dy). Inorg. Chem. 33, 4707–4711. (30) Jacques, V., and Desreux, J. F. (2002) New classes of MRI contrast agents. Top. Curr. Chem. 221, 123–164. BC700375M