Fluorinated Proteins as Potential 19F Magnetic Resonance Imaging

Dec 17, 1993 - imaging (MRI) and spectroscopy (MRS) agents. Proteins labeled ... the potential of these new fluorinated proteins as in vivo MRI/MRS pr...
0 downloads 0 Views 1MB Size
Bloconjugate Chem. 1994, 5, 257-261

257

Fluorinated Proteins as Potential 19F Magnetic Resonance Imaging and Spectroscopy Agents? Vimal D. Mehta,' Padmakar V. Kulkarni, Ralph P. Mason, Anca Constantinescu, and Peter P. Antich Department of Radiology, University of Texas Southwestern Medical Center, Dallas, Texas 75235. Received December 17, 1993"

Fluorinated proteins have been synthesized and characterized as potential in vivo l9Fmagnetic resonance imaging (MRI) and spectroscopy (MRS) agents. Proteins labeled with fluorine include bovine serum albumin, y-globulin, and purified immunoglobulin (IgG). The amino groups in proteins were selectively trifluoroacetylated using S-ethyl trifluorothioacetate to synthesize fluorinated proteins (TFA-protein; 1-3). In another approach, trifluoroacetamidosuccinic anhydride has been used to prepare corresponding fluorinated derivatives of proteins (TFASA-protein; 4-6). The fluorinated proteins have been purified by exhaustive dialysis and isolated in good yields (55-76 7%). The fluorinated proteins exhibit useful NMR characteristics and the biocompatibility for in vivo studies. The initial investigations demonstrate the potential of these new fluorinated proteins as in vivo MRI/MRS probes.

INTRODUCTION

Macromolecular carrier systems have been developed in an attempt to optimize the delivery of antineoplastic agents to tumors (1).Indeed, conjugation of antineoplastic drugs to carrier polymers and proteins may improve the biodistribution and therapeutic efficacy of drugs (2). Macromolecules such as albumin, globulins, and synthetic polymers accumulate in tumor tissue because the vascular network shows both enhanced permeability of the neovasculature and lack of a lymphatic recovery system (3,4). Radionuclides have been attached to serum albumin in an effort to detect micrometastases using scintigraphy (5). In addition, lH NMR paramagnetic tissue contrast agents have been conjugated to albumin with considerably enhanced vascular retention and improved contrast (6). Polyclonal human nonspecific IgG has been used as a carrier molecule (lllIn-labeled) in clinical trials to detect areas of inflammation (7). Additionally, human polyclonal IgG attached to monodisperse iron oxides (MION) has been used to enhance MRI of inflammation (8). Kilbourn et. al. have labeled proteins (HSA and IgA) with 18Fand suggested that these approaches could be used for 19F labeling of proteins (9). l9F NMR offers a unique quantitative means of imaging infused agents, as there is essentially no l9F background signal in tissue. We have been developing both novel fluorine-labeled molecular probes (IO, 11)and enhanced MR techniques to exploit the exceptional characteristics of fluorine (12, 13). Small fluorinated molecules, e.g., trifluoroacetylglucosamine, are subject to very rapid renal clearance (10). A t the other extreme of molecular size, perfluorocarbon emulsions (nanoparticles) are subject to extensive uptake by the liver and spleen and exhibit excessive tissue retention (14). The development of fluorine-labeled probes with prolonged retention in the vascular compartment should facilitate 19FNMR investigations of lesions like tumor, ischemia, or inflammation. In addition, the fluorine-labeled proteins may permit quantification and characterization of the catabolic processes of proteins (15)in tissues in vivo that could not be achieved with conventional radiolabels. ~~

* To whom correspendence should be addressed. + Presented at the 204th ACS National Meeting (Division of Medicinal Chemistry), Washington, D.C., Aug 23-28, 1992. Abstract published in Advance ACS Abstracts, April 1,1994. 1043-1802/94/2905-025?$04.50/0

The purpose of this study was to develop molecular probes with appropriate NMR characteristics and the biocompatibility for in vivo applications using magnetic resonance imaging (MRI) and spectroscopy (MRS). Albumin and the globulins have been used as model proteins for labeling with fluorine ('9F). The fluorine-labeling procedures and 19F NMR characteristics of fluorinated conjugates are reported together with preliminary examinations of their biological behavior, such as blood clearance and i n vivo spectroscopy. EXPERIMENTAL PROCEDURES

Reagents and Materials. Bovine serum albumin (BSA), y-globulin, and purified immunoglobulin (IgG) were purchased from Sigma Chemical Co. S-Ethyl trifluorothioacetate, sodium trifluoroacetate (CF&OONa), sodium fluoride (NaF),and sodium bicarbonate (NaHC03) were purchased from Aldrich Chemical Co. Trifluoroacetamidosuccinic anhydride was synthesized by a literature-reported procedure (16) and stored under dry conditions. Spectra Por dialysis tubings (Fisher Scientific) with molecular weight cutoff (MWCO) 6-8 K and 12-14 K were used. All other reagents and solvents were reagent grade unless otherwise specified. Magnevist (Berlex Laboratories Inc., Wayne, NJ), a 0.5 M solution of N-methylglucamine salt of the gadolinium complex of diethylenetriamine pentaacetic acid (Gd-DTPA), was used as the paramagnetic relaxation agent. Equipment and Physical Measurements. 19FNMR spectroscopy was performed using a 7 T (300 MHz, lH) Oxford magnet under control of a Techmag console. Imaging ('9F and lH) experiments were performed using an Omega CSI 4.7 T system with actively shielded gradients (Acustar, Bruker Instruments, Inc.). Infrared spectra (IR) were recorded in KBr on a Mattson Galaxy (2000) FT-IR spectrometer. Gel permeation chromatography (GPC) was carried out using a Waters HPLC system. Fluorine LabelingProcedures. Reaction of S-Ethyl Trifluorothioacetate with Proteins. The appropriate protein (BSA, y-globulin, or IgG) (50-100 mg) was dissolved in 0.1 M NaHC03 (40-50 mL). S-Ethyl trifluorothioacetate (SETFA) (I7) was added (50 molar excess) dropwise to the protein solution at 4 "C while a pH -8.0 was maintained (by adding NaHC03 solution). 0 1994 American Chemical Society

258

Mehta et al.

Bioconjugate Chem., Vol. 5, No. 3, 1994

The reaction was allowed to proceed overnight at room temperature. The reaction mixture was extensively dialyzed using MWCO 6-8-K and 12-14-K dialysis tubing (sequentially) against water to remove low molecular weight materials (excess of reagent) for not less than 72 h. The solution was lyophilizedto yield trifluoroacetylated protein derivatives (TFA-protein; 1 and 3). Reaction of y-globulin with SETFA resulted in water-insolubleproduct 2, which was isolated by centrifugation. During reaction and purification the temperature was maintained at 4 "C for IgG. Reaction of Trifluoroacetamidosuccinic Anhydride with Proteins. The appropriate protein (BSA,y-globulin, or IgG) (50-100 mg) was dissolved in 0.1 M NaHC03 (4050 mL). Trifluororoacetamidosuccinic anhydride (16) was added (50 molar excess) in small lots to the protein solution at 4 "C while a pH -8.0 was maintained (by adding NaHC03 solution). The reaction was allowed to proceed overnight at room temperature. The conjugate was extensively dialyzed using MWCO 6-8-K and 12-14-K dialysis tubing (sequentially) against water to remove low molecular weight materials (excess of reagent) for not less than 96 h. The solution was lyophilized to yield trifluorosuccinylated protein derivatives (TFASA-protein; 4-6). During reaction and purification the temperature was maintained at 4 "C for IgG. Gel Permeation Chromatography. Gel permeation chromatography (GPC) was performed using Waters ultrahydrogel columns at 25 OC. Phosphate buffer (2M) was used for elution at a flow rate of 0.8 mL/min with a Waters 484 (UV/vis) and 410 differential refractometer. 19FNMR Spectroscopy and Sensitivity Measurements. Typically, fluorinated protein (viz. 1; -10 mg; 0.28 M) was dissolved in water or fresh heparinized blood (-500 pL). The pH of the protein solution in water was adjusted to neutral (with NaOH solution) for NMR experiments. Sodium trifluoroacetate (CF3COONa;0.46 M) and sodium fluoride (NaF; 0.1 M) in water were used as chemical shift reference standards. 19FNMR detection sensitivity was estimated using NaF solution as an internal standard with relative sensitivity per mg material. 2'1 and Tz were estimated using standard pulse burst saturation recovery and Hahn spin-echo methods. In Vivo Evaluations. Blood Clearance Kinetics. TFA-albumin (1, 11.3 mg) was injected iv in a group of six Balb/C mice. Blood samples were drawn from pairs of mice at 5 min, 2 h, 5 h, 48 h or 1h, 4 h, 72 h o r 2 hand 24 h to minimize blood volume loss for any given mouse. The concentration of fluorine in blood samples was estimated with 19FNMR spectroscopy by comparison with an external standard of NaF. In Vivo Spectroscopy. TFA-albumin (1) was administered (iv) in two consecutive doses of 50 mg 2 h apart to a tumor-bearing mouse (Meth-A). The mouse was anesthetized [ketamine (100mg/kg) xylazine (5 mg/kg) (IM)], and a surface coil (1.8-cm diameter) was placed against the tumor. Spectra were acquired at 282.3 MHz using an 80-ps pulse with 400 scans in 10 min and spectral width of 12 000 Hz. A 20-Hz exponential line broadening was applied prior to Fourier transformation. Phantom Imaging. The phantom comprised three glass vials containing 40 mg of TFA-albumin (1) in 1mL of water each. In addition, two of the vials had, respectively, -5 and 10 mM Gd-DTPA. The phantom was placed in a 3-cm single turn solenoid lH/'9F tunable coil. lH images were obtained at 200 MHz using TR = 250 ms and T E = 8 ms with two acquisitions at each of 64 increments for a total imaging time of 32 s. 19F images

-

+

Scheme 1 PROTEIN

a

-NH,

*

PROTEIN TFA

-NH -RI

-PROTEIN

ALBUMIN (1)

=

PROTEIN

0

GLOBULIN (2) R, =

IgG (3) a

II

-C-CF,

Key: (a) 0.1 M NaHC03, EtSCOCF3.

Scheme 2 PROTEIN

b

-NH,

t

PROTEIN TFASA

-NH -R,

- PROTEIN 0

II

ALBUMIN (4) PROTEIN

=

GLOBULIN (5) IgG ( 6 )

R,=

-C

diC -cFJ

II 0

II

0

a Key: (b) 0.1 M NaHC03, trifluoroacetamidosuccinic anhydride.

Table 1. Characteristics of Fluorinated Proteins 1-6 fluorinated proteins 19F NMR yield: svnthesizeda sensitivityb IR (KBr).cm-1 % TFA-albumin (1) 10 mg/mg NaF 1658 (M), 1535 (NH) 76 TFA-IgG (3) 15 mg/mg NaF 1641 (C=O), 1529 (NH) 55 TFASA-albumin (4) 20 mg/mg NaF 1656 (C=O), 1539 (NH) 73 TFASA-globulin (5) 28 mg/mg NaF 1647 (C==O), 1535 (NH) 64 TFASA-IgG (6) 23 mg/mg NaF 1643 (C=O), 1531 (NH) 60 All fluorinated proteins except 2 were water soluble and had a chemical shift coincident with CFsCOzNa and -44.10 ppm downfield from NaF at neutral pH (-7.0). lBFMR sensitivities obtained under fully relaxed conditions with respect to NaF (internal standard) at neutral pH (&l5-20% error). Based on amount (mg) of protein recovered. a

l9F

*

were obtained with TR = 150 ms and T E = 8 ms using a driven equilibrium sequence to enhance data collection efficiency (18)with 256 acquisitions at each of 32 gradient increments for a total imaging time of 20 min. Images were processed with sin2 apodization prior to Fourier transformation. RESULTS AND DISCUSSION

We have used proteins which are readily available and occur naturally at high concentration in the body. In addition, bovine serum albumin and immunoglobulins were cost-effective model proteins for the development and standardization of fluorine-labeling procedures. We adopted two approaches to the fluorination of these proteins. In the first approach, the amino groups in the proteins were selectively trifluoroacetylated using 5'-ethyl trifluorothioacetate (SETFA) (TFA-protein; 1-3) (Scheme 1). 7-Globulin on reaction with SETFA produced a waterinsoluble product (2). In an alternative approach we used trifluoroacetamidosuccinic anhydride (TFASAN) to prepare corresponding fluorinated derivatives of proteins (TFASA-protein; 4-6) (Scheme 2). The first approach, trifluoroacetylation of proteins, yielded products with higher fluorine labeling (i.e., proteins with higher 19F detection sensitivity) as compared to labeling with trifluoroacetamidosuccinic anhydride (Table 1). This may be due to the competing hydrolysis reaction of anhydride with trifluorosuccinylation of proteins in aqueous solution. The additional free carboxylic acid groups introduced in

Fluorinated Proteins as MRI and MRS Agents

fluorinated proteins 4-6 using TFASAN may alter the charge on proteins and favorably enhance the water solubility, as is evident from y-globulin derivative 5, which was obtained as a water-soluble product. For in vivo applications, it is desirable to minimize the injection volume. Usually, labeling of proteins with fluorinating agents involves the modification of amino acid side chain polar functional groups (e.g., NH2 groups), which in turn results in reducing hydrophilicity of derivatized products. The introduction of polar groups simultaneously with fluorine labeling will maintain or even enhance the solubility of derivatized products. The fluorinated proteins 1-6 were purified by exhaustive dialysis and isolated in good yields (55-76%; Table 1).A control experiment showed that dialysis successfully removed the reaction byproduct, trifluoroacetamidosuccinic acid (TFASA),from proteins. A sample of TFASAN was first hydrolyzed to TFASA by adding to 0.1 M NaHC03. The TFASA solution was dialyzed independently and also after adding to albumin solution. Lyophilization of TFASA dialysate yielded no visible product. The dialysis solution containing albumin showed no 19F signal in the product obtained after lyophilization. To evaluate the product purity and integrity of the proteins following the reaction, gel permeation chromatography (GPC) was performed. Analyses of fluorinated proteins showed a smooth GPC profile without formation of additional low or high molecular weight products. The IR spectra of TFA-protein 1-3 and TFASA-protein 4-6 showed two enhanced bands at 1641-1658 and 1529-1535 cm-' and 1643-1656 and 1531-1539 cm-l, respectively, compared to native proteins besides a broad band (29003500 cm-l). The fluorinated proteins 1to 6 exhibit asingle sharp 19F NMR signal as demonstrated in Figure 1 for TFA-albumin (1) and TFA-IgG (3) with a typical line width of -80-100 Hz. The chemical shifts of proteins 1-6 are essentially coincident with CF3COONa and -44.10 ppm downfield from NaF. l9F spin-lattice relaxation time Ti= 0.75-1.0 s and spin-spin relaxation time T2 i= 10-25 ms were obtained at 7 T for derivatized proteins 1-6 (19). The line width as well as relaxation parameters of derivatized proteins were significantly different from small molecular weight fluorinated compounds (e.g., CF3COONa, line width -20 Hz, TI -2.9 s, TZ-760 ms) (19). The conjugation of the fluorinated moiety to macromolecules (e.g., proteins) results in a reduction of the relaxation times TI and T2. Reduced 19FTI is favorable for imaging experiments. It permits rapid data acquisition and thus shortens imaging time without sacrificing resolution. The NMR characteristics exhibited by derivatized proteins together with the other methods of characterization described above confirm the purity and identity of fluorinated proteins. The observation of a single sharp l9F resonance for fluorine-labeled proteins 1-6 contrasts previous reports of protein labeling with fluorine (15), and facilitates imaging without chemical shift artifacts. We have determined the relative 19F NMR sensitivity of the fluorinated proteins compared to sodium fluoride (NaF). In other words, the relative sensitivity is defined as the number of mg of fluorinated protein needed to give the same intensity of 19F NMR signal as that obtained with 1mg of NaF (Table 1). The NMR sensitivity provides an estimate for the relative efficiency of the labeling methods and in vivo NMR detectability of derivatized proteins. The 19Fsignals from fluorinated proteins were readily observed when added to whole blood as shown for TFAalbumin (1) andTFA-IgG (3) inFigure 1B,D. Theyexhibit

Bioconjugate Chem., Vol. 5, No. 3, 1994 259 NaF TFA-aibumln (1)

PHANTOM (A)

WHOLE BLOOD (B)

I

TFA-IgC (3)

WHOLE BLOOD @)

50.0

b.0

!,o.o

$0.0

b.0

'$0.0

!S.O

19F CHEMICAL SHIFT (PPM)

Figure 1. 282.3-MHz l9F NMR spectra of TFA-albumin (1) or TFA-IgG (3) and NaF. Fluorinated proteins 1 and 3 are shown, respectively, in water (A and C) and in whole rabbit blood (Band D). 1 and 3 both have a single sharp 19Fsignal, and there is no line broadening or change in chemical shift in blood.

a single sharp resonance without any substantial line broadening, change in signal intensity, or chemical shift in blood. TFA-albumin (1) was also detectable as a single resonance in blood drawn from an animal following iv infusion and, indeed, was detectable in excised tissues (liver and tumor) 72 h after administration (48 mg) to a tumorbearing mouse. The blood clearance kinetics of TFA-albumin (1) in mice were investigated for a period of 6 h following iv injection (11.3 mg). Clearance of the agent from the blood as measured by the decrease in the intensity of 19Fsignal is shown in Figure 2. The signal to noise ratio of 19Fsignal ranged from -11 to 6 for blood samples at 5 min postinjection to 5 h. TFA-albumin (1) showed prolonged retention in the blood pool compared to monomeric molecules (10) and the half-life (4-5 h) is similar to GdDTPA-labeled albumin (20)and lsF-labeled human serum albumin (9). Clearance of TFA-albumin (1) appeared linear contrasting perfluorocarbon emulsions,which follow a nonlinear kinetic (21). Prolonged retention of TFAalbumin (1) in the intravascular space suggests enhanced probability of leakage of these agents into the tumor interstitium. Although no formal toxicity studies have been performed, we have injected 40-50 mg of TFA-albumin (1) iv in tumor-bearing mice (Meth-A;n = 3) without apparent acute toxicity. Similarly, in a rat there was no toxicity for 72 h following administration of 300 mg by combination of iv and ip. The TFA-albumin (1) has significantly lower

260

Mehta et al.

Bioconjugafe Chem., Vol. 5, No. 3, 1994

BLOOD CLEARANCE KINETICS OF TFA-ALBUMIN

T

0

1

2

3

4

5

6

Time (hour)

Figure 2. Blood clearance profile of TFA-albumin (1) in Balb/C mice from 5 min to 5 h following iv injection (11.3 mg). TFA-Albumin (1)

19F CHEMICAL SHIFT (p.p.m)

Figure 3. 19FMR spectrum of a Meth-A tumor obtained using a surface coil following administration of two consecutive doses of 50 mg each of TFA-albumin (1) iv 2 h apart. Spectra were acquired a t 282.3 MHz using an 80-ps pulse with 400 scans in 10 min and spectral width of 12 000 Hz. A 20-Hz exponential line broadening was applied prior to Fourier transformation.

toxicity than observed for fluorinated polylysines (10)and could be used at higher doses to provide the 19FNMR detection sensitivity for i n vivo studies. In Vivo Spectroscopy. Preliminary experimentsusing localized surface-coil spectroscopyshowed 19FNMR signal from TFA-albumin (1) in the Meth-A tumor in a mouse following successive iv infusions (Figure 3). The concentration of 1in the tumor remained constant over a period of 2 h as studied at 10-min intervals. This indicates the feasibility of in vivo investigations with fluorine-labeled proteins. PhantomImaging. We have determined the 19FNMR spin-lattice relaxation time 2'1 -620 ms and spin-spin relaxation time 2'2 -70 ms for TFA-albumin (1)at 4.7 T (22 "C) to assess their imaging potential. Here, we show 19Fand corresponding lH MR images of a phantom of TFA-albumin (1)(Figure 4 (top and bottom)). 19Fimages were obtained with a S/N 10-20 using driven equilibrium. We have previously noted that addition of Gd-DTPA substantially reduces both 2'1 and 2'2 of the TFA-albumin (19). Use of driven equilibrium substantially enhances signal from the vial with longer 2'1 (no Gd-DTPA) but had relatively little effect on the two other vials (with Gd-DTPA). We note that reducing the 21' of TFAalbumin (1)with a paramagnetic contrast agent can have more significant impact on S/N than using driven equilibrium. However, excessive concentration of relaxation agent significantly shortens 2'2 (C) in Figure 4 (top and bottom), thereby reducing the benefits of the increased signal strength (22) caused by shortened 2'1 (B). TFA-albumin (1) has been evaluated as a prototype macromolecular agent, and although it may not prove to be the optimal agent, it does provide an opportunity to investigate the applications of an intravascular 19Fagent

-

Figure 4. (Top) driven equilibrium fluorine (l9F)and (bottom) corresponding proton (lH) images of a phantom containing 40 mg of TFA-albumin (1) in 1mL of water (A). Magnevist (GdDTPA) 20 p L (- 5 mM) and 40 p1 (- 10 mM) has been added to B and C. 19Fimages were obtained with TR = 150 ms and T E = 8 ms using a driven equilibrium sequence to enhance data collection efficiency (18) with 256 acquisitions a t each of 32 gradient increments for a total imaging time of 20 min. Corresponding lH images were obtained a t 200 MHz using TR = 250 ms, T E = 8 ms, and two acquisitions a t each of 64 increments for total imaging time of 32 s. Images were processed with sin2 apodization prior to Fourier transformation.

for MRI. Recently, we successfully conjugated 19FpH sensor to carrier polymers (23)without loss in pH sensing capability, providing the opportunity for tissue targeting of 19Fphysiological markers. The optimal macromolecular 19Fagent must have appropriate NMR characteristics and prove to be safe and effective. The fluorinated macromolecular agents will be useful in the evaluation of tissue perfusion and for the diagnosis of ischemic and neoplastic disorders. In summary, proteins have been successfully derivatized providing fluorine labeled biocompatiblemolecular probes with use€ul NMR characteristics. Our initial results demonstrate the potential of these new fluorinated proteins as MRI/MRS probes. However, more detailed studies with different proteins and fluorine labels involving biodistribution (plasma clearance and organ distribution),toxicity, metabolic fate, and immunologic consequences are necessary to establish their potential as i n vivo MRI/MRS probes. ACKNOWLEDGMENT

This work was supported by Dallas Biomedical Corporation/Hartford Foundation, and MR studies were conducted at the Mary Ne11 & Ralph B. Rogers NMR center. We are grateful to Dr. H. Sinn (DKFZ, Heidelberg) for valuable discussions and thank Ms. P. Lea for help in the NMR experiments.

Fluorinated Proteins as MRI and MRS Agents

LITERATURE CITED (1) Friend, D. R., and Pangburn, S. (1987) Site-specific drug delivery. Med. Res. Rev. 7, 53-106. (2) Maeda, H., Seymour, L. W., and Miyamoto, Y. (1992) Conjugates of anticancer agents and polymers: advantages of macromolecular therapeutics i n vivo. Bioconjugate Chem. 3(5), 351-362. (3) Dvorak, H. F., Nagy, J. A., Dvorak J. T., and Dvorak, A. M. (1988) Identification and characterization of the blood vessels of solid tumors that are leaky to circulating macromolecules. Am. J . Pathol. 133,95-109. (4) Jain, R. K. (1987) Transport of molecules across tumor vasculature. Cancer Metastasis Rev. 6, 559-593. (5) Sinn, H., Schrenk, H. H., Friedrich, E. A., Schilling, U., and Maier-Borst, W. (1990) Design compounds having an enhanced tumour uptake, using serum albumin as a carrier. Part I. Nucl. Med. Biol. 17(8), 819-827. (6) Ogan, M. D., Schmiedl, U., Moseley, M. E., Grodd, W., Paajanen, H., and Brasch , R. C. (1987) Albumin labeled with Gd-DTPA an intravascular contrast-enhancing agent for magnetic resonance blood imaging: preparation and characterization. Invest. Radiol. 22, 665-671. ( 7 ) Rubin, R. H., Fishman, A. J., and Callahan, R. J. (1989) ll1In-labelled nonspecific immunoglobulin scanning in the detection of focal infection. N . Engl. J. Med. 321, 935-940. (8) Weissleder, R., Lee, A. S., and Fischman, A. J. (1991) MR antibody imaging: polyclonal human IgG labelled with polymeric iron oxide. Radiol. 181, 245-249. (9) Kilbourn, M. R., Dence, C. S., Welch, M. J., and Mathias, C. J. (1987) Fluorine-18 labeling of proteins. J. Nucl. Med. 28(4), 462-470. (10) Mehta, V. D., Kulkarni, P. V., Rajur, S. B., Mason, R. P., Babcock, E. E., Constantinescu, A., and Antich P. P. (1992) Novel molecular probes for l9F magnetic resonance imaging: synthesis & characterization of fluorinated polymers. Biomed. Chem. Lett. 2(6), 527-532. (11) Mehta, V. D., Kulkarni, P. V., Mason, R. P., and Antich, P. P. (1993) Fluorinated macromolecular probes for noninvasive assessment of pH by magnetic resonance spectroscopy. Biomed. Chem. Lett. 3(2), 187-192. (12) Mason, R. P., Bansal N., Babcock, E. E., Nunnally, R. L., and Antich, P. P. (1990) A novel editing technique for 19F

Bioconjugate Chem., Vol. 5, No. 3, 1994 261

MRI: molecule-specific imaging. Magn. Reson. Imaging& 729736. (13) Barker, B., Mason R. P., and Peshock, R. M. (1993) Echo planar imaging of perfluorocarbons. Magn. Reson. Imaging 11, 1165-1173. (14) Lutz, J., and Metzenauer, P. (1980) Effects of potential blood substituents on rat liver and spleen. Pflugers Arch. 387, 175-181. (15) Daugherty, A,, Becker, N. N., Schemer, L. A., Sobel, B. E., Ackerman, J. J. H., Baynes, J. W., and Thorpe, S. R. (1989) Non-invasive detection of protein metabolism i n vivo by nmr spectroscopy. Biochem. J. 264, 829-835. (16) Lapidus, M., and Sweeney, M. (1973) L-4-Cyano-342, 2, 2-trifluoroacetamido)succinalilic acid and related synthetic sweetening agents. J. Med. Chem. 16, 163-166. (17) Wolfrom, M. L., and Conigliaro, P. J. (1969) Trifluoroacetyl as an N-protective group in the synthesis of purine nucleosides of 2-amino-2-deoxy saccharides. Carbohyd. Res. 11, 63-76. (18) van Uijen, C. M. J., and Den Boef, J. H. (1984) Drivenequilibrium radiofrequency pulses in nmr imaging. Magn. Reson. Med. I, 502-507. (19) Mehta, V. D., Mason, R. P., Kulkarni, P. V., Lea, P., Constantinescu, A., and Antich, P.P. (1994) 19F MR characterization of fluorinated proteins and relaxation rate enhancement with Gd-DTPA for faster imaging. Proceedings of the International Symposium on Chemists View of Imaging Centers (A. M. Emran, Ed.) Plenum Press, New York, accepted for publication. (20) Schmeidl, U., Moseley, M. E., Ogan, M. D., Chew, W. M., and Brasch, R. C. (1987) Comparison of initial biodistribution patterns of Gd-DTPA and albumin-(Gd-DTPA) using rapid spin-echo imaging. J. Comput. Assist. Tomogr. 11, 306-13. (21) Malet-Martino, M. C., Betbeder, D., Lattes, A., Lopez, A., Martino, R., Francois, G., and Cros S. (1984) Fluosol 43 intravascular persistence in mice measured by I9F NMR. J. Pharm. Pharmacol. 36, 556-559. (22) Ratner, A. V., Quay, S., Muller, B. B., Simpson, R., Hurd, R., and Young, S. W. (1989) 19Frelaxation rate enhancement and frequency shift with Gd-DTPA. Invest. Radiol. 24,22426. (23) Mehta, V. D., Aravind, S., Kulkarni, P. V., Mason, R. P., and Antich, P. P. (1993) Fluorinated macromolecular probes as biosensors for non-invasive assessment of pH with magnetic resonance spectroscopy. J. Nucl. Med. 34(5), 124.