Novel Paramagnetic Macromolecular Complexes Derived from the

Macromolecular Gd(III) complexes may find useful application as contrast agents for magnetic resonance angiography (MRA). Herein two novel systems are...
0 downloads 0 Views 178KB Size
Download PDF
192

Bioconjugate Chem. 1999, 10, 192−199

Novel Paramagnetic Macromolecular Complexes Derived from the Linkage of a Macrocyclic Gd(III) Complex to Polyamino Acids through a Squaric Acid Moiety Silvio Aime,*,† Mauro Botta,† Simonetta Geninatti Crich,† Giovanni Giovenzana,‡ Giovanni Palmisano,§ and Massimo Sisti‡ Dipartimento di Chimica IFM, Universita` di Torino, via P. Giuria 7, I-10125 Torino, Italy, Dipartimento di Chimica Organica ed Industriale, Universita` di Milano, Viale Venezian 21, I-20133 Milano, Italy, and Dipartimento di Scienza e Tecnologia del Farmaco, Universita` di Torino, via P. Giuria 9, I-10125 Torino, Italy. Received March 16, 1998; Revised Manuscript Received October 15, 1998

Macromolecular Gd(III) complexes may find useful application as contrast agents for magnetic resonance angiography (MRA). Herein two novel systems are reported, namely Gd(DO3ASQ)3-lys16 and Gd(DO3ASQ)30-orn114. Their syntheses are based on the ability of the squaric acid moiety to act as a linker between the DO3A (1,4,7,10-tetraazacyclododecane-1,4,7-triacetic acid) chelate moiety and the polyamino acidic chain. Moreover, the squaric acid participates in the coordination cage of the Gd(III) ion. The investigation of 1H and 17O NMR relaxation processes of solvent water nuclei allowed a detailed characterization of the systems under study. Gd(DO3ASQ)30-orn114 displays a remarkable ability to enhance the water proton relaxation rate of its solutions, and it may be considered as potential contrast agent for MRA applications.

INTRODUCTION

The search for macromolecular Gd(III) complexes (1, 2) has been an important issue in the past decade of development of contrast agents (CA) for magnetic resonance imaging (MRI) (3-5). This interest is even stronger now that magnetic resonance angiography (MRA) is becoming a procedure of widespread use. These macromolecular systems provide both an enhanced ability to catalyze solvent proton relaxation rates [through the occurrence of a long molecular reorientational time τR (6, 7)] and an increased lifetime of the CA in the circulating blood by avoiding the extravasation typical of the smallsized Gd(III) complexes commonly employed in MRI investigations. It was earlier reported that an easy way to attach a chelating moiety to a macromolecule consists of forming an amide linkage with the reaction of an amino group on the macromolecule and the bis(anhydride) derivative of DTPA. This procedure has been used to prepare systems such as (GdDTPA)n-albumin and (GdDTPA)n-polylysine (8, 9). The main drawback in the use of the bifunctional dianhydride DTPA stems from the occurrence of undesiderable intra- and intermolecular cross-linking reactions. Intermolecular cross-linking may be reduced by the use of low protein concentrations, but this has no effect on intramolecular cross-linking, which represents a major risk in the case of highly flexible macromolecules such as polylysine (10). The diconjugation of DTPA is expected to cause a large reduction of the thermodynamic stability of the metal complexes as proved in a detailed study on model systems (11). The search has then been addressed to methods that preclude diconjugation of the chelate, * Author to whom correspondence should be addressed [fax (+39)-11-670-7524; e-mail [email protected]]. † Dipartimento di Chimica IFM. ‡ Dipartimento di Chimica Organica ed Industriale. § Dipartimento di Scienza e Tecnologia del Farmaco.

for instance, through the use of the monoanhydride derivative of DTPA (12). However, as outlined also by other authors (10), the synthesis of the latter ligand appears to be difficult to reproduce and the yields are invarianty low. For the same purpose other authors have dealt with functionalized DOTA chelates (13). Again, the employed procedures are rather laborious and troublesome. On this basis, the need for a monoreactive chelate derivative that would enable extensive and specific activation of amino groups in macromolecules is still vivid. The effectiness of paramagnetic CAs is represented by their relaxivity (r1p); that is, the water proton relaxation rate of a 1 mM solution of the Gd(III) chelate, that, at the magnetic fields usually employed in MRI (0.5-1.5 T), is roughly proportional to the molecular reorientation time (τR) of the complex. Unfortunately, the relaxivity of such systems is not as high as one would expect on the basis of their molecular weights. Possible causes for such behavior may be either the occurrence of internal motions that reduce the effective molecular reorientation time or the presence of a relatively long exchange lifetime (τM) of the coordinated water molecule or both. As far as the exchange of the water molecule coordinated to the Gd(III) ion is concerned, the enhancement of the solvent relaxation rate (Ris 1p) is “quenched” when τM is longer than the relaxation time of these protons (T1M) (14). Indeed, a number of bis(amide) DTPA Gd(III) complexes have been reported to have τM values of ∼1 µs at room temperature (15), whereas other DTPA- and DOTAlike (1,4,7,10-tetraazacyclododecane-N,N′,N′′,N′′′-tetraacetic acid) Gd(III) complexes display shorter τM values (14). On this basis, it was deemed of interest to explore alternative routes to obtain conjugates between macromolecular substrates and Gd(III) chelate moieties different from DTPA-amide. Since it has been recently reported that squaric esters readily react with amino groups (16), under mild conditions, we decided to explore

10.1021/bc980030f CCC: $18.00 © 1999 American Chemical Society Published on Web 02/05/1999

Novel Paramagnetic Macromolecular Complexes Scheme 1

their use in the formation of stable linkages between Gd(III) complexes and NH2 moieties on macromolecular substrates. Furthermore, once the squaric ester derivatives are bonded to an amino nitrogen of a tetraazacyclododecane macrocycle as in DO3ASQ and DO3ASQ-est ligands (Scheme 1), they may show coordination properties analogous to those displayed by an acetate arm in the parent DOTA ligand. EXPERIMENTAL PROCEDURES

Syntheses. Synthesis of 10-(2-Ethoxy-3,4-dioxo-1-cyclobutenyl)-1,4,7,10-tetraazacyclododecane-1,4,7-triacetic Acid, (Tri)-tert-Butyl Ester (3). To a stirred solution of the triester (1) [1,4,7,10-tetraazacyclododecane-1,4,7triacetic acid, (tri)-tert-butyl ester] (17), kindly provided by Bracco (Bracco S.p.A., Milano, Italy) (3.3 g, 6.1 mmol) in ethanol (50 mL) is added a solution of diethyl squarate (1.24 g, 7.73 mmol) in one portion. The resulting mixture is stirred overnight at room temperature and then evaporated in vacuo. The product is purified by column chromatography (SiO2, CHCl3/acetone 50:50 f 10:90), yielding 3.60 g (92.7%) of 3 as a light yellow viscous oil: 1 H NMR (300 MHz, CDCl3) δ 4.72 q[2H] (J ) 7.1 Hz), 3.93 t[2H] (J ) 6.0 Hz), 3.78 t[2H] (J ) 6.0 Hz), 3.26 s[2H], 3.25 s[4H], 2.98 m[4H], 2.72 m[8H], 1.42 s[27H], 1.30 t[3H] (J ) 7.1 Hz); 13C NMR (75.5 MHz, CDCl3) δ 189.0, 182.1, 176.1, 172.0, 170.6, 81.0, 69.2, 58.0, 57.6, 56.7, 55.6, 54.8, 54.6, 53.2, 53.0, 52.1, 52.0, 49.6, 48.7, 28.1, 15.8; IR (neat) 2979, 2932, 2848, 1797, 1731, 1714, 1613 cm-1; MS(CI), m/z 640 (MH+), 611 (-CO); Rf (TLC) ) 0.78 (SiO2, acetone). Synthesis of 10-(2-Hydroxy-3,4-dioxo-1-cyclobutenyl)1,4,7,10-tetraazacyclododecane-1,4,7-triacetic Acid (Trifluoroacetate) (DO3ASQ). To a stirred solution of the tetraester 3 (3.27 g) in THF (20 mL) is added concentrated HCl (37%, 2.5 mL), and the mixture is stirred for 24 h at room temperature. Diethyl ether (20 mL) is slowly added to form a two-phase system, which is allowed to stand at room temperature overnight. The crystalline product DO3ASQ is isolated by filtration, washed with diethyl ether, and dried in vacuo. The product is recrystallized from water, obtaining 2.88 g: mp 223-225 °C (dec); 1H NMR (D2O, 200 MHz) δ 4.25 m[4H], 3.60 m[4H], 3.48 m[4H], 3.46 s[6H], 3.08 m[4H]; 13C NMR (D2O, 50.3 MHz) δ 198.6, 190.0, 184.2, 179.4, 170.5, 56.4, 55.6, 55.2,

Bioconjugate Chem., Vol. 10, No. 2, 1999 193

54.1, 52.2, 49.3; IR (Nujol mull) 3500-2500 (bb), 1813, 1751, 1694, 1665 cm-1. Synthesis of 10-(2-Ethoxy-3,4-dioxo-1-cyclobutenyl)1,4,7,10-tetraazacyclododecane-1,4,7-triacetic Acid (Trifluoroacetate) (DO3ASQ-est). A solution of compound 3 (204.0 mg, 0.3194 mmol) in trifluoroacetic acid (1 mL) was stirred overnight at room temperature. After evaporation to dryness, the residue was dissolved in methanol (1 mL); slow addition of diethyl ether (10 mL) caused the precipitation of DO3ASQ-est as a white powder, which was collected by centrifugation, washed with diethyl ether, and dried in vacuo (140.7 mg): mp 170-175 °C (dec); 1H NMR (D2O, 200 MHz) δ 4.65 q[2H] (J ) 7.0 Hz), 3.82-3.51 m[10H], 3.16-3.00 m[12H], 1.36 t[3H] (J ) 7.1 Hz). Syntheis of Polylysine Conjugate. Polylysine hydrobromide (5, 49.1 mg; Sigma, St. Louis, MO) is dissolved in 250 µL of H2O; triethylamine (500 µL) is added to the clear solution and, after 5 min, a solution of the tetraester 4 (33.6 mg) in 750 µL of ethanol is added in one portion. The mixture is stirred at room temperature for 72 h, with the disappearance of 4 (SiO2, CHCl3/MeOH 9:1) monitored by TLC. The mixture is evaporated in vacuo and the residue dissolved in neat trifluoroacetic acid (3 mL) and stirred overnight at room temperature; the solution is evaporated, and the residue is dissolved in 1 mL of methanol. By slow addition of a 4-fold volume of diethyl ether, the conjugate (7) precipitated as a white amorphous powder, which was collected by centrifugation, repeatedly washed with diethyl ether, and dried in vacuo (yield ) 69.7 mg). The presence of unconjugated DO3ASQ monomer has been ruled out on the basis of HPLC analysis. The experimental setup of HPLC analysis is as follows: Hamilton PRP-1 column (5 µm, 250 × 4 mm), UV detector at 260 nm, isocratic elution with a 10:90 mixture of acetonitrile/TBOH, 0.02 M, pH 6.5 by H2SO4; elution flux ) 1 mL/min. The samples were derivatized by treatment with a 0.01 M CuSO4 solution. Synthesis of Polyornithine Conjugate. Polyornithine hydrochloride (5, 51.6 mg; Sigma) is dissolved in 500 µL of H2O; triethylamine (500 µL) is added to the clear solution and, after 5 min, a solution of the tetraester 4 (60.9 mg) in 500 µL of ethanol is added in one portion. The mixture is stirred at room temperature for 72 h, with the disappearance of 4 (SiO2, CHCl3/MeOH 9:1) monitored by TLC. The mixture is evaporated in vacuo and the residue dissolved in neat trifluoroacetic acid (3 mL) and stirred overnight at room temperature; the solution is evaporated, and the residue is dissolved in 1 mL of methanol. By slow addition of a 4-fold volume of diethyl ether, the conjugate (7) precipitated as a white amorphous powder, which was collected by centrifugation, repeatedly washed with diethyl ether, and dried in vacuo (yield ) 72.7 mg). The absence of unconjugated DO3ASQ monomer has been checked by HPLC as described above for the polylysine conjugate. Preparation of the Gd(III) Complexes. The complexation has been carried out by adding stoichiometric amounts of GdCl3 to the aqueous solutions of the ligands at neutral pH and at room temperature. The formation of the complex has been followed by measuring the solvent proton relaxation rate (1/T1). The presence of an excess of free Gd(III) ions, which yields a noticeable increase of the observed relaxation rate, may be quickly removed by centrifugation of the solution brought to basic pH. NMR Measurements. The longitudinal water proton relaxation rate was measured by using a Stelar Spin-

194 Bioconjugate Chem., Vol. 10, No. 2, 1999

Aime et al.

master spectrometer [Stelar, Mede (PV) Italy] operating at 20 MHz, by means of the standard inversion-recovery technique (16 experiments, 2 scans). A typical 90° pulse width was 3.5 µs, and the reproducibility of the T1 data was (0.5%. The temperature was controlled with a Stelar VTC-91 air-flow heater equipped with a copper-constantan thermocouple (uncertainty of (0.1 °C). The proton 1/T1 NMRD profiles were measured over a continuum of magnetic field strength from 0.00024 to 1.2 T (corresponding to 0.01-50 MHz proton Larmor frequency) on the Koenig-Brown field-cycling relaxometer installed at the NMR relaxometry laboratory of the University of Torino (Italy). The relaxometer works under complete computer control with an absolute uncertainty in 1/T1 of (1%. Details of the instrument and of the data acquisition procedure may be found elsewhere (18). Variable-temperature 17O NMR measurements were recorded on a JEOL EX-90 (2.1 T) spectrometer, equipped with a 5 mm probe, by using a D2O external lock. Experimental settings were as follows: spectral width ) 10000 Hz, pulse width ) 7 µs, acquisition time ) 10 ms, 1000 scans, and no sample spinning. Solutions containing 2.6% of 17O isotope (Yeda, Israel) were used. The observed transverse relaxation rates (RO 2obs) were calculated from the signal width at half-height. The 1H and 13C NMR spectra of the ligands were performed on a Bruker AC-200 spectrometer (7.1 T). tert-Butyl alcohol (1%) was used as internal reference (δ1H ) 1.29 ppm, δ13C ) 31.3 ppm, and δ31P ) 0 ppm) for 1H/13C NMR spectra recorded using as solvent D2O (99.8%, Merck, Darmstadt, Germany). THEORY 1H

Water Relaxation Rate. The longitudinal water proton relaxivity rH 1p is defined as the paramagnetic contribution to the observed water proton relaxation rate ( RH 1obs) of a 1 mM solution of the paramagnetic metal complex: H H rH 1p ) R1obs - R1d

(1)

In eq 1 RH 1d is the water proton relaxation rate in the presence of an equimolar amount of a diamagnetic analogue of the paramagnetic compound which, for diluted solutions, may be made equal to the value of 0.38 s-1, that is, the relaxation rate of pure water at 298 K. The observed relaxivity results from contributions arising from water molecules in the inner- and outer-coordination spheres of the metal ion: His Hos rH 1p ) r1p + R1p

(2)

rHis 1p deals with the contribution arising from the exchange of the water directly coordinated to the paramagnetic metal ion and is given by H H RHis 1p ) (Cq/55.6)(T1M + τM)

(3)

where C is the molar concentration of the complex, q the H the longitudinal relaxation time hydration number, T1M of the inner-sphere water protons, and τH M their residence lifetime. The Solomon-Bloembergen theory (19) describes the magnetic field dependence of TH 1M, which, for a Gd(III) chelate, is given by

[

]

2 2 2 7τc2 3τc1 1 2 2γH ge µBS(S + 1) ) + H 6 2 2 15 T1M rH 1 + ωH τc1 1 + ω2S τ2c2 (4)

where S is the electron spin quantum number [7/2 for Gd(III)], γH is the proton nuclear magnetogyric ratio, µB is the Bohr magneton, ge is the Lande` factor for the free electron, rH is the distance between the metal ion and the inner-sphere water protons, ωH and ωS are the proton and electron Larmor frequencies (ωS ) 658ωH), respectively, and τci (i ) 1, 2) are the correlation times related to the modulation of the electron-proton dipolar coupling. Such an interaction may be modulated by the reorientation of the paramagnetic species, τR, by the residence lifetime of the inner-sphere water molecule (τM) and by the electronic relaxation times, TiE. -1 -1 -1 ) τ-1 τci R + τM + TiE

(5)

In analogy with the nuclear relaxation time also the electronic relaxation processes depend on the magnetic field strength. For Gd(III) complexes TiE are related to the modulation of the zero field splitting (ZFS) of the electronic spin states due to the dynamic distortions of the ligand field interaction and, according to the Blombergen-Morgan theory (20), their magnetic field dependence is given by the following equations:

(

)

T-1 1E )

1 2 4 1 ∆ τv[4S(S + 1) - 3] + 2 2 25 1 + ωS τv 1 + 4ω2S τ2v (6)

T-1 2E )

1 2 ∆ τv[4S(S + 1) - 3] × 50 5 2 3+ + (7) 1 + ω2S τ2v 1 + 4ω2S τ2v

(

)

where ∆2 is the trace of the square of the transient ZFS tensor and τv is the correlation time related to its modulation. The outer-sphere term, RHos 1p , describes the contribution arising from the water molecules diffusing near the paramagnetic chelate and, according to the model of Hwang and Freed (21), may be related to the minimum distance between the metal and the diffusing water molecules, a, the relative solute-solvent diffusion coefficient, D, and the electronic relaxation times, TiE

(aD1 )[7J(ω ) + 3J(ω )]

os RHos 1p ) C

S

H

(8)

where Cos is a constant (5.8 × 10-13 s-2 M-1) and the dependence on the electronic relaxation times is expressed in the non-Lorentzian spectral density functions J(ωi). 17 O Water Relaxation Rate. The residence lifetime of a water molecule directly coordinated to a paramagnetic metal ion (τO M) may be evaluated by measuring the temperature dependence of the paramagnetic contribu17 O water solvent transverse tion (RO 2p) to the observed relaxation rate O O RO 2p ) R2obs - R2d

(9)

where, in analogy with 1H measurements, the diamagnetic term RO 2d is measured on a solution containing a diamagnetic analogue of the chelate of interest. (We used 17O NMR data of LaDOTA.) RO is related to τO through 2p M

Novel Paramagnetic Macromolecular Complexes

Bioconjugate Chem., Vol. 10, No. 2, 1999 195

Scheme 2

Scheme 3

Scheme 4

17O chemical shift the values of ∆ωO M (which is the difference between coordinated and bulk water molecule) and RO 2M (which is the transverse relaxation rate of the coordinated water oxygen) (22):

RO 2p )

O 2 O-1 O O2 qC O-1 R2M + τM R2M + ∆ωM τM 55.6 (RO + τO-1 )2 + ∆ωO 2 2M

M

(10)

M

The temperature dependence of ∆ωO M is described by the equation

∆ωO M )

geµBS(S + 1)B0 A 3kBT p

(11)

where B0 is the magnetic field strength (2.11 T in this work) and A/p is the Gd-17O scalar coupling constant [the value of which for polyaminocarboxylate Gd(III) complexes may be reasonably fixed to -3.8 × 106 rad s-1]. For relatively small-sized Gd(III) chelate and at the field of 2.11 T used in this work, RO 2M is dominated by the electron-nucleus scalar interaction:

RO 2M )

()

1 A 3 p

2

(

S(S + 1) τE1 +

τE2 1 + ω2S τ2E2

)

-1 O -1 τ-1 E2 ) TiE + (τM)

(12) (13)

Finally, the temperature dependence of RO 2P is expressed in terms of the Eyring relationship for τO M and τv

(τj) -1 T )

[ (

298.15 (τ-1 T ∆Hj 1 j ) 1 exp 298.15 R 298.15 T

)]

(14)

where j refers to the two different dynamic processes involved (j ) v, M) and ∆Hj is the corresponding activation enthalpy. RESULTS AND DISCUSSION

Synthesis of the Ligands DO3ASQ and DO3ASQest. The ligands reported herein were prepared starting from diethyl squarate 2 (Scheme 2). The reactivity of this compound toward amines has been already investigated (23-25). Diethyl squarate may be selectively converted into the corresponding mono- or bis(amide)s (26, 27) simply by varying the experimental conditions. Reaction of a slight excess of diethyl squarate 2 with DO3A-tertbutyl ester (1) in ethanol at room temperature gave, after

silica gel chromatography, nearly quantitative yields of tetraester 3. The ligand 4 was obtained in good yield by reacting 3 with concentrated HCl/THF at room temperature, which promotes the simultaneous hydrolysis of the tert-butyl and ethyl ester groups. 13C NMR Characterization of DO3ASQ and DO3ASQ-est Ligands. The 13C NMR spectra of DO3ASQ and DO3ASQ-est provide useful insights to predict the coordination properties of these ligands. In fact, whereas the chemical shifts of the 13C resonances of the macrocyclic ring and of the acetate arms (see Experimental Procedures) are very similar in the two ligands, a noticeable difference exists as far as the cyclobutenedione (squaric) group resonances are concerned. In DO3ASQ the squaric acid moiety gives raise to three resonances with the relative intensity ratio of 1:2:1 at 197.1 (C-3), 189.7 (C-2 and C-4), and 183.2 (C-1) ppm, respectively, owing to the delocalization of the negative charge (Scheme 3) as outlined by the remarkable downfield shift of C-3 with respect to C-2/C-4. On the other hand, in the case of DO3ASQ-est, there are four, equally intense, squaric ester derivative resonances at 190.3 (C-2), 184.8 (C-1), 175.4 (C-4), and 174.5 (C-3) ppm, respectively. Because it is well established that a heteroatom substituent (alkoxy or amino) deactivates the nonadjacent carbonyl group through resonance delocalization (28, 29), the contribution of this resonance form A-2, reported in Scheme 4, is also evident from the large upfield shift of C-3. Thus, the strong upfield shift of C-3 suggests a large decrease of its carbonylic character, which is consistent with a shift in the equilibrium represented in Scheme 4. Besides the squaric acid resonances, the overall 13C NMR spectrum of the DO3ASQ-est, recorded at 25 °C, supports this hypothesis as it shows 18 different signals resulting from the hindered rotation around the bond between carbon (C-1) of the squaric ester moiety and the macrocyclic nitrogen (i.e., vinylogous amide functionality). Thus, we expect that in DO3ASQ the squaric moiety may participate in the chelation of a Ln3+ ion through the negatively charged oxygen at one of the equivalent C-2 or C-4 sites, whereas in DO3ASQ-est the coordination might involve either the oxygen of the carbonyl C-2 functionality or the estereal oxygen. Chelation of the Gd(III) ions from both DO3ASQ and DO3ASQ-est occurs instantaneously, at room temperature, by mixing equimolar solutions of the reagents. 1 H and 17O Relaxometric Investigations of DO3ASQ and DO3ASQ-est Gd(III) Complexes. The

196 Bioconjugate Chem., Vol. 10, No. 2, 1999

Aime et al.

Table 1. Best Fitting Parameters Obtained from the Analysis of the NMRD Profiles (298 K, pH 7.0, 1 mM) and of the Temperature Dependence of 17O-R2p (2.1 T, pH 7.0, 50 mM) for Aqueous Solutions of DO3ASQ and DO3ASQ-est-Gd(III) Chelates Gd-DO3ASQ Gd-DO3ASQ-est

q

∆2 (s-2)/1019

τv (ps)

τR (ps)

τM (ns)

r (Å)

∆Hv (kJ mol-1)

∆HM (kJ mol-1)

1 1

6.3 ( 0.4 5.5 ( 0.4

19 ( 3 18 ( 3

78 ( 2 83 ( 2

134 200

2.98 2.98

1.3 2.0

41 42

Figure 1. 1/T1 NMRD profile of a 1 mM aqueous solution of GdDO3ASQ (9) and GdDO3ASQ-est (0) at 25 °C and pH 7. The solid curves through the data points were calculated with the parameters of Table 1. The lower curves represent the outer sphere contributions to the overall relaxivity.

Figure 2. Temperature dependence of the transverse water 17O relaxation rate at 2.1 T and pH 7 for 50 mM solution of GdDO3ASQ (9) and GdDO3ASQ-est (0). The solid curves through the data points were calculated with the parameters of Table 1.

relaxivities at 20 MHz and 25 °C of GdDO3ASQ and GdDO3ASQ-est are 5.3 and 5.6 mM-1 s-1, respectively, that is, in the upper range of values for related polyamino polycarboxylate Gd(III) complexes containing one coordinated water molecule (q ) 1) (5, 14). The NMRD

profiles of GdDO3ASQ and GdDO3ASQ-est (Figure 1) closely resemble those obtained for a number of DOTAlike Gd(III) complexes, and Table 1 shows the parameters obtained by fitting the experimental data to the Solomon-Bloembergen-Morgan and Freed equations. Thus,

Scheme 5

Novel Paramagnetic Macromolecular Complexes

the occurrence of q ) 1 implies that the squaric acid moiety is involved in the coordination scheme of these complexes. A parameter particularly important for the forthcoming application of these complexes is represented by the residence lifetime of the coordinated water molecule at the metal site (τM). As has been demonstrated in several recent publications (15, 30, 31), the exchange rate of the coordinated water molecule (τO-1 M ) may be accurately determined by the measurement of the solvent water 17O-T at variable temperature. In fact, the temperature 2 dependence of RO 2M may be fitted in terms of the temO perature dependence of τO M, τv, and ∆ωM. In Figure 2 the O R2p data for GdDO3ASQ and GdDO3ASQ-est measured over an extended temperature range are reported, and the calculated values of the relevant parameters obtained from the fitting procedure to the above-reported theory are shown in Table 1. From this procedure we obtained, at 298 K, τM values of 134 ns for GdDO3ASQ and 200 ns for GdDO3ASQest. These values are similar to those reported for other DTPA- and DOTA-like Gd(III) complexes and are almost 1 order of magnitude shorter than those obtained at the same temperature for the corresponding complexes with bis(amide)-DTPA ligands. Thus, although still far from optimum τM values of 10-20 ns, which ensure maximum relaxation enhancement (14) for systems with long τR and T1E, the exchange lifetime of the coordinated water in these Gd(III) chelates of squaric acid containing ligands should not represent a limiting factor in attaining a relaxation enhancement for conjugates in which the paramagnetic moiety is attached to macromolecular substrates of molecular mass up to 30-40 kDa. Synthesis of the Conjugates with Polylysine and Polyornithine. The tetraester 3 (Scheme 5) was employed for the synthesis of polylysine-DO3ASQ and polyornithine-DO3ASQ conjugates. For this purpose we chose two polyamino acids endowed with a large difference in the number of monomeric units in their chain, that is, polylysine (n ) 16) and polyornithine (n ) 114). The “second amidation step” implied the reaction with an amino group in the presence of a base (generally a tertiary amine). Due to the low solubility of polylysine and polyornithine in solvents other than water, the coupling reaction was run in a water/ethanol mixture, in the presence of triethylamine. The reaction of the two polymers with 3 required 72 h to reach completion; the two conjugates were isolated by adding diethyl ether to a methanolic solution of the crude hydrolysis product. All of the added chelate bound almost quantitatively to yield (DO3ASQ)n-lys16 and (DO3ASQ)m-orn114 respectively. n and m numbers were estimated to be ∼3 and ∼30 on the basis of the observation that all of the added compound 3 was consumed. These numbers have been confirmed by the results of the titration with Gd3+ ions (vide infra). The number of chelate moieties on each polyamino acidic chain is an average value as we have not attempted any separation of the reaction products. The conjugates so obtained were directly used for the complexation of the paramagnetic ion. Synthesis and Relaxometric Characterization of DO3ASQ-lys16 and DO3ASQ-orn114 Gd(III) Complexes. The dropwise addition of a GdCl3 solution to 1 mM aqueous solutions of (DO3ASQ)n-lys16 and (DO3ASQ)m-orn114 at room temperature led to the prompt chelation of the metallic ions by the chelate moieties on the polyamino acidic substrate. The formation of the macromolecular complex has been monitored

Bioconjugate Chem., Vol. 10, No. 2, 1999 197

Figure 3. Longitudinal water proton relaxation rate of 0.078 mM (DO3ASQ)30-orn114 ([) and 1 mM (DO3ASQ)3-lys16 (O) ligand solutions as a function of the Gd(III) ion concentration added to solutions.

Figure 4. Temperature dependence of the longitudinal proton relaxivity at 20 MHz and pH 7 of Gd(DO3ASQ)30-orn114 (9) and Gd(DO3ASQ)3-lys16(0) water solutions.

by measuring the relaxation time of the water protons. Each addition of Gd(III) ions corresponds to an increase of the observed relaxation rate (not a linear increase as the successive coordination of metal ions causes a progressive increase of the molecular weight and an eventual change in the overall polymer structure). The titration end point (Figure 3) is clearly detected and allows us to establish the relative metal chelate/polyamino acid ratio, thus yielding the following stoichiometries: Gd(DO3ASQ)3-lys16 (MW = 5000) and Gd(DO3ASQ)30orn114 (MW = 35000); their relaxivities are 15.6 and 31.0 mM-1 s-1, respectively. To get more insight into the understanding of the various parameters determining the observed relaxivities of these paramagnetic macromolecular complexes, we measured both the temperature dependence of proton r1p at 0.47 T and the temperature dependence of 17O-R2 at 2.1 T. From the data reported in Figure 4, it is clearly shown that from room to high temperature, the relaxivities of Gd(DO3A)3-lys16 and Gd(DO3A)30-orn114 are not limited by τM; that is, the condition T1M > τM holds (eq 1). At lower temperatures the decreased exchange rate and the concomitant shortening of T1M (as a consequence of the elongation of τR) yield the typical flattening of the profile of r1p versus T, expected for systems whose τM > T1M. This is confirmed by analyzing the water 17O-R2p variable temperature profiles (Figure 5) from which the τM values obtained for both macromolecular complexes are relatively short and similar to those determined for the free Gd complexes. Table 2 shows that the other parameters and the reorientational correlation time (τR) display values comparable to those of the free complexes. This finding supports the view that each Gd ion owns one water molecule, pointing

198 Bioconjugate Chem., Vol. 10, No. 2, 1999

Aime et al.

Table 2. 17O NMR Best Fitting Parameters Obtained from the Analysis of the Temperature Dependence of RO 2p for 11 mM Aqueous Solution of the Gd(III) Complexes Gd(DO3ASQ)30-orn114 Gd(DO3ASQ)3-lys16

q

∆2 (s-2)/1019

τv (ps)

τM (ns)

∆Hv (kJ mol-1)

∆HM (kJ mol-1)

1 1

4.4 4.5

11 29

153 197

4.3 1.5

31 45

the 0.01-3 MHz region are indicative of additional contributions arising mainly from water molecules bound to the surface of the macromolecular substrate in the proximity of the paramagnetic center (32, 33). CONCLUDING REMARKS

Figure 5. Temperature dependence of the transverse water 17O relaxation rate at 2.1 T and pH 7 for 11.0 mM solutions of Gd(DO3ASQ)30-orn114 (9) and Gd(DO3ASQ)3-lys16 (0). The solid curves through the data points were calculated with the parameters of Table 2.

In summary, the results herein reported show that, for the formation of conjugates with macromolecular substrates, an easy alternative to the use of DTPA anhydride is now available leading to Gd(III) chelates that display quite good relaxivities and do not suffer the limitation of the long exchange lifetime. Particularly interesting is the (GdDO3ASQ)30-orn114 conjugate, which has an r1p value, at 39 °C, of 26.5 mM-1 s-1, significantly higher than the r1p values reported by Brasch et al. for other macromolecular CAs such as (GdDTPA)n-polylysine (r1p ) 13.1 mM-1 s-1) and (GdDTPA)n-albumin (r1p ) 14.4 mM-1 s-1), respectively (1). Furthermore, with respect to the use of DTPA-bis(anhydride), there is the advantage that, by applying the method herein described, no cross-linking between the macromolecular chain occurs. The easiness found in the anchoring procedure of the chelate moiety to the macromolecular substrates suggests that this route based on squaric acid conjugates may find a wide generalization. As far as the specific substrates considered in this work are concerned, polyamino acidic chains containing a high number of Gd(III) chelates may be used as high relaxivity markers for systems able to target selected tissues or organs. ACKNOWLEDGMENT

Figure 6. 1/T1 NMRD profile of a 1 mM aqueous solution of Gd(DO3ASQ)30-orn114 (9) or Gd(DO3ASQ)3-lys16 (0) at 25 °C and pH 7.

out that the squaric group maintains its ability to coordinate the metal ion. The linkage to the polyamino acidic residue yields to the transformation of one of the C(2/4)-O bonds of the squaric acid moiety into a C-N bond; thus, it seems likely that the coordination to the Gd(III) center occurs through the carbonyl (C-2) functionality. On this basis it seems likely that an analogous binding scheme may occur also in the case of the monomeric GdDO3ASQ-est. The NMRD profiles of Gd(DO3A)3-lys16 and Gd(DO3A)30-orn114 (Figure 6) show a relaxivity increase at ∼20-30 MHz typical of slowly tumbling paramagnetic systems. Only in the case of Gd(DO3A)30-orn114 does the NMRD profile display the shape of the relaxivity peaks commonly found for paramagnetic protein conjugates. Furthermore, the extent of the observed relaxivity enhancements is not proportional to the molecular weights of the two systems to suggest that large internal motions overlap the reorientational processes of the whole molecules. As in the case of the noncovalent paramagnetic adducts with albumin, the data in the low-field region of the NMRD profile cannot be fitted with the r1p values calculated on the basis of the Solomon-BloenbergenMorgan theory. Likely, the high relaxivities observed in

Financial support from Bracco S.p.A. (Milano, Italy), CNR, and MURST is gratefully acknowledged. This work was carried out under the scheme of the EU-COST-D8 Program and the EU-BIOMED II-MACE Project. We gratefully acknowledge the stimulating discussion with Prof. I. Bertini, C. Luchinat, L. Casella, and their collegues. LITERATURE CITED (1) Brasch, R. C. (1991) Rationale and Applications for Macromolecular Gd-Based Contrast Agents. Magn. Reson. Med. 22, 282-287. (2) Schuhmann-Giampieri, G., Schmitt-Willich, H., Frenzel, T., Press, W., and Weinmann, H. (1991) In vivo and in vitro evaluation of Gd-DTPA-polylysine as a macromolecular contrast agent for resonance magnetic imaging. Invest. Radiol. 269, 69-78. (3) Kumar, K., and Tweedle, M. F. (1993) Macrocyclic polyaminocarboxylate complexes of lanthanides as magnetic resonance imaging contrast agents. Pure Appl. Chem. 65, 515520. (4) Rinck, P. A. (1993) Magnetic Resonance in Medicine, pp 155-174, Blackwell Scientific Publications, Oxford, U.K. (5) Peters, J. A., Huskens, J., and Raber, D. J. (1996) Lanthanide induced shifts and relaxation rate enhancements. Prog. NMR Spectrosc. 28, 283-350. (6) Lauffer, R. B. (1987) Paramagnetic Metal Complexes as Water Proton Relaxation Agents for NMR Imaging: Theory and Design. Chem. Rev. 87, 901-927. (7) Lauffer, R. B., Brady, T. J., Brown, R. D., III, Baglin, C., and Koenig, S. H. (1986) 1/T1 NMRD Profiles of solutions of

Novel Paramagnetic Macromolecular Complexes Mn(II) and Gd(III) Protein-Chelate Conjugates. Magn. Reson. Med. 3, 541-548. (8) Sieving, P. F., Watson, A. D., and Rocklage, M. R. (1990) Preparation and characterization of paramagnetic polychelates and their protein conjugates. Bioconjugate Chem. 1, 6571. (9) Spanoghe, M., Lanens, D., Dommisse, R., Van Der Linden, A., and Alderweireldt, F. (1992) Proton relaxation enhancement by means of serum albumin and poly-L-lysine labeled with DTPA-Gd3+: relaxivities as a function of molecular weight and conjugation efficiency. Magn. Reson. Imaging 109, 13-21. (10) Maisano, F., Gozzini, L., and de Haen, C. (1992) Coupling of DTPA to proteins: A critical analysis of the cyclic dianhydride method in the case of insulin modification. Bioconjugate Chem. 3, 212-217. (11) Sherry, A. D., Cacheris, W. P., and Kuan, K. T. (1988) Stability constants for Gd3+ binding to model DTPA-conjugates and DTPA-proteins: implications for their use as magnetic resonance contrast agents. Magn. Reson. Med. 8, 180-190. (12) Halpern, S., Stern, P., Hagan, P., Chen, A., Frincke, J., Bartholomew, R., David, G., and Adams, T. (1983) Labeling of monoclonal antibodies with indium(III): technique and advantages compared to radioiodine labeling. In Radioimmunoimaging and Radioimmunotherapy (S. W. Burchiel and B. A. Rhodes, Eds.), pp 197-205, Elsevier Science Publishing, New York. (13) Parker, D. (1990) Tumor targeting with radiolabeled macrocycle-antiboby congjugates. Chem. Soc. Rev. 19, 271291. (14) Aime, S., Botta, M., Fasano, M., and Terreno, E. (1998) Lanthanide(III) chelates for NMR biomedical applications. Chem. Soc. Rev. 27, 19-29. (15) Powell, D. H., Ni, Dhubhghaill, O. M., Pubanz, D., et al. (1996) Structural and Dynamic Parameters Obtained from 17O NMR, EPR, and NMRD Studies of Monomeric and Dimeric Gd3+ Complexes of Interest in Magnetic Resonance Imaging: An Integrated and Theoretically Self-Consistent Approach. J. Am. Chem. Soc. 118, 9333-9346. (16) Tietze, L. F., Schroter, C., Gabius, S., Brinck, U., GoerlachGraw, A., and Gabius, H. J. (1991) Conjugation of p-Aminophenyl Glycosides with Squaric Acid Diester to a Carrier Protein and the Use of the Neoglycoprotein in the Histochemical Detection of Lectins. Bioconjugate Chem. 2, 4853. (17) Chang, C. A., Francesconi, L. C., Malley, M. F., et al. (1993) Synthesis, Characterization, and Crystal Structure of MDO3A (M ) Fe, Gd) and Na[M(DOTA)] (M ) Fe, Y, Gd). Inorg. Chem. 32, 3501-3508. (18) Koenig, S. H., Brown, R. D., III, and Gupta, R. K., Eds. (1987) NMR Spectroscopy of Cells and Organism, CRC Press, Boca Raton, FL. (19) Banci, L., Bertini, I., and Luchinat, C. (1991) Nuclear and Electron Relaxation, VCH, Weinheim, Germany. (20) Bloembergen, N., and Morgan, L. O. (1961) Proton Relaxation Times in Paramagnetic Solutions. Effects of Electron Spin Relaxation. J. Chem. Phys. 34, 842-850. (21) Hwang, L. P., and Freed, J. H. (1975) Generalized Einstein Relations for Rotational and Translational Diffusion of Molecules Including Spin. J. Chem. Phys. 63, 118-130.

Bioconjugate Chem., Vol. 10, No. 2, 1999 199 (22) Swift, T. J., and Connick, R. E. J. (1962) NMR Relaxation Mechanisms of O17 in Aqueous Solutions of Paramagnetic Cations and the Lifetime of Water Molecules in the First Coordination Sphere. J. Chem. Phys. 37, 307. (23) Schmidt, A. H., and Ried, W. (1978) Die Praparative Chemie der Cyclobutenedione; Synthese von Cyclobutendion und dressen Alkyl-, Alkenyl- und aryl- Derivaten. Synthesis, 1-22. (24) Knorr, H., and Ried, W. (1978) Die Praparative Chemie der Cyclobutenedione; Reaktionen von Alkyl-, Alkenyl- and Arylcyclobutendionen. Synthesis, 649-666. (25) Schmidt, A. H., and Ried, W. (1978) Die Praparative Chemie der Cyclobutenedione: Synthese von Quadratsaure, Benzocyclobutendion und deren Derivaten. Synthesis 8, 6880. (26) Tietze, L. F., Arlt, M., Beller, M., Glusenkamp, K. H., Jahde, E., and Rajewsky, M. F. (1991) Bioisosteric Replacement of the a-Aminocarboxylic Functionality in 2-Amino-5phosphonopentanoic Acid Yield Unique 3,4-Diamino-3-cyclobutene-1,2-dione Containing NMDA Antagonists. Chem. Ber. 124, 1215-1221. (27) Kinney, W. A., Lee, N. E., Garrison, D. T., Podlesny, E. J., Jr., Simmonds, J. T., Bradlett, D., Notvest, R. R., Kowal, D. M., and Tasse, R. P. (1992) Squaric Acid Diethyl Ester: A New Coupling Reagent for the Formation of Drug Biopolymer Conjugates. Synthesis of Squaric Acid Ester Amides and Diamides. J. Med. Chem. 35, 4720-4738. (28) Liebeskind, L. S., Granberg, K. L., and Zhang, J. (1992) A Strategy for Generalization of the Regiospecific Synthesis of Substituted Quinones from Cyclobutenediones. J. Org. Chem. 57, 4345-4352. (29) Liebeskind, L. S., and Sun, L. J. (1995) Construction of Highly-substituted Aromatics Bearing Acyl Substituents Using Cyclobutenedione Technology: Vinylketene-Based Benzannulations using 3-Acyl-4-(substituted amino) cyclobutenediones. J. Org. Chem. 60, 8194-9201. (30) Micksei, K., Powell, D. H., Helm, L., Brucher, E., and Merbach, A. E. (1993) Water Exchange on [Gd(H2O)8]3+ and [Gd(PDTA)(H2O)2]- in Aqueous Solution: a Variable-Pressure, -Temperature and -Magnetic Field 17O NMR Study. Magn. Reson. Chem. 31, 1011-1020. (31) Micksei, K., Helm, L., Brucher, E., and Merbach, A. E. (1993) 17O NMR Study of Water Exchange on [Gd(DTPA)(H2O)]2- and [Gd(DOTA)(H2O)]- Related to NMR Imaging. Inorg. Chem. 32, 3844-3850. (32) Aime, S., Botta, M., Fasano, M., Geninatti Crich, S., and Terreno, E. (1996) Gd(III) Complexes as Contrast Agents for Magnetic Resonance Imaging: a Proton Relaxation Enhancement Study of the Interaction with Human Serum Albumin. J. Biol. Inorg. Chem. 1, 312-319. (33) Aime, S., Botta, M., Geninatti Crich, S., et al. (1997) Towards MRI contrast agents of improved efficacy. NMR relaxometric investigations of the binding interaction to HSA of a novel heptadentate macrocyclic triphosphonate Gd(III)complex. J. Biol. Inorg. Chem. 2, 470-479.

BC980030F