The distribution of TmDOTP5− in rat tissue was compared with CoEDTA−, an anionic complex previously used as a marker of extracellular space. Heart, liver, muscle, blood, and urine were collected from rats after infusion of either complex and were quantitatively analyzed by atomic absorption spectroscopy. Although total TmDOTP5− in blood and tissue was consistently lower (0.88 ± 0.04;n = 6) than CoEDTA− after an identical infusion protocol (presumably because of some association of the phosphonate complex with bone), a comparison of blood and tissue contents indicated that the two anionic complexes distributed into identical extracellular spaces. Relative extracellular space in the in vivo liver, as determined by TmDOTP5− and CoEDTA−, was 0.18 ± 0.02 and 0.15 ± 0.01, respectively. The corresponding relative extracellular space values for the in vivo heart reported by the two agents were identical (0.11 ± 0.02). Experiments were also performed to evaluate the washout kinetics of TmDOTP5− from anesthesized rats. In rats given a total dose of 0.16 mmol TmDOTP5−, 81% appeared in urine by 180 min, <2% was found in all remaining soft tissue, leaving ∼18% undetected. The rate of Tm appearance in urine was fit to a standard pharmacokinetic model that included four tissue compartments: plasma, one fast equilbrating space, one slow equilibrating space, and one very slow equilibrating space (presumably bone). The best fit result suggests that the highly charged TmDOTP5− complex is cleared from plasma more rapidly than is the typical lower charged Gd-based contrast agents and that release from bone is slow compared with renal clearance.
- extracellular space marker
- intracellular sodium
- tissue distribution
there is considerable interest in monitoring and measuring the concentration of intracellular Na in intact cells, perfused tissue, and in vivo (12, 21, 27, 30). The most common analytical methods available for measuring Na in biological tissue include ion-selective microelectrodes, fluorescent indicator dyes, atomic absorption spectroscopy (AAS), and nuclear magnetic resonance (NMR) spectroscopy. The first method measures Na activity directly and does not require knowledge of intracellular or extracellular volumes. Although extracellular ions are most commonly detected by using microelectrodes (16), intracellular measurements are possible in single cells. The second method relies on introducing an Na-sensitive ligand into the intracellular space (5). Again, this method does not require knowledge of the intracellular volume and is usually limited to studies of isolated cells. The last two analytical methods do require an independent estimate of the volume of the intracellular space and extracellular space (ECS).
Traditionally, the volume of the ECS is measured by using a chemical marker that is known to be confined to that space. An ideal marker must be noninvasive, chemically stable and not degraded by the tissue, have high analytical sensitivity, and distribute rapidly and uniformly into all ECS. Common ECS markers include various inert, negatively charged transition metal complexes (6, 28) and various inert sugars such as inulin, sucrose, and sorbital (17, 18). These markers can be quantitated by enzymatic assay, using radiolabeled materials, or, in the case of metal complexes, by either AAS or emission spectroscopy.58CoEDTA−, a gamma-emitting isotope of cobalt, has been used as an ECS marker and found to measure the same space as [14C]sucrose (7). Nonradioactive CoEDTA− has also been successfully used to measure ECS, with Co being determined by AAS (25).
Each of the above methods is, by necessity, destructive. NMR offers the potential to monitor intracellular space and ECS completely noninvasively. Although these techniques have been demonstrated most often in studies of isolated cells and perfused organs, NMR methods can in principle be applied in vivo. A combination of1H- and59Co-NMR using the water signal as a measure of total volume and Co(CN) as an ECS marker has been used both in isolated cell studies (28) and in perfused tissues (1,13). 31P-NMR methods, first demonstrated by Kirk and Kuchel (15), using a combination of neutral phosphorus-containing markers that distribute into all water space and of charged phosphonates markers that are restricted to the ECS, have also become popular (9).
In the present study, thulium-1,4,7,10-tetraazacyclododecane-1,4,7,10-tetrakis(methylene phosphonate) (TmDOTP5−) was compared with CoEDTA− as a marker of ECS in the in vivo heart and liver. The concentrations (marked by brackets) of Tm, Co, and Na in tissue and blood were determined by AAS, and the [Tm]tissue/[Tm]plasmaratio (or [Co]tissue/[Co]plasmaratio) was used to determine the volume of the ECS. Given the ECS and relative dry weight (rDW) of the tissue, intracellular Na concentration ([Na+]i) was evaluated by measuring the difference between total tissue and plasma Na, as originally described by Scheufler and Peters for perfused hearts (25). Additional experiments were performed to compare the biodistribution of TmDOTP5−and CoEDTA− and to measure the rate of clearance of TmDOTP5− from live animals. The later kinetic data agreed well with an established pharmacokinetic model for clearance of negatively charged contrast-agent complexes from three tissue compartments.
MATERIALS AND METHODS
All protocols were approved by the University’s Institutional Animal Care and Research Advisory Committee. Male Sprague-Dawley rats (260–390 g) were initially anesthetized with a 0.5-ml intramuscular injection of ketamine (87 mg/ml) and xylazine (13 mg/ml). A tracheotomy was performed to ensure a clear breathing passageway throughout the experiment. The animals were not ventilated. The carotid artery and one jugular vein were cannulated through a midline neck incision. The line from the carotid artery was used to collect blood samples while the line from the jugular vein was used to administer 80 mM TmDOTP5− or 80 mM CoEDTA−. Stock solutions of TmDOTP5− and CoEDTA− were prepared from Na4HTmDOTP (Magnetic Resonance Solutions, Dallas, TX) and KCoEDTA (10), respectively.
Three different infusion studies were performed. In the first, 80 mM TmDOTP5− were infused at 2 ml/h (6 min), 4 ml/h (6 min), 6 ml/h (6 min), 8 ml/h (6 min), and 6 ml/h (81 min) while the blood samples were periodically withdrawn for determination of total Tm, Na, Mg, and Ca by AAS. In the second, either 80 mM TmDOTP5− or 80 mM CoEDTA− was infused at 2 ml/h (6 min), 4 ml/h (6 min), 6 ml/h (6 min), and 8 ml/h (6 min). Animals were then killed, and the heart, liver, one kidney, and one thigh muscle were freeze-clamped and weighed. All urine was collected from the bladder, a sample of blood was taken, and one femur removed and cleaned free of tissue. The tissues were dissolved in a 5:1:1 mixture of nitric, sulfuric, and perchloric acids, respectively, and Tm or Co in each tissue type (expressed as a percentage of total metal ion infused) was determined by AAS. The femur appeared to dissolve completely in hot acid, but an insoluble precipitate that later formed precluded quantitation of Tm and Co in this tissue.
In the third type of infusion experiment, 80 mM TmDOTP5− were infused at 2 ml/h (6 min), 4 ml/h (6 min), 6 ml/h (6 min), and 8 ml/h (6 min), followed by infusion of normal saline at 6 ml/h for an additional period of 156 min. This experiment was performed to evaluate washout of TmDOTP5− from tissues and appearance of TmDOTP5− in urine. During prolonged infusions, 100 μl of anesthetic were injected into a muscle every 30 min to maintain proper anesthesia. All animals were positioned supine on a small heating pad to maintain normal body temperature. The bladder was catheterized to collect urine samples. Blood samples were withdrawn from the arterial line before, during, and after the infusion period. All tissue samples were blotted to remove excess blood, weighed wet, dried at 60°C overnight, then reweighed to determine a dry-to-wet ratio.
Sample preparation and atomic absorption methods. Urine and blood samples were diluted 1:50 (or as needed) for Tm, Co, Na, Mg, and Ca determinations by AAS. Each tissue sample was digested in 2 ml of an acid mixture containing 5:1:1 concentrated nitric, sulfuric, and perchloric acids, respectively. The tissue digests were heated at 60°C overnight in a water bath. After cooling, the digests were diluted to 10 ml, then diluted again as needed. KCl at a final concentration of 2,000 parts/million was added to all samples to prevent overionization of Na (24). A Varian SpectrAA-5 atomic absorption spectrometer was used to analyze samples for Tm, Na, Co, Mg, and Ca at 372.2, 330.2, 241.0, 285.2, and 239.9 nm, respectively. Tm and Co were measured in a nitrous oxide-acetylene flame, whereas Na, Mg, and Ca were measured in an air-acetylene flame.
Pharmacokinetic model. A kinetic model was generated by using ModelMaker 3.0 (Cherwell Scientific, Oxford, UK) and used to predict the appearance of Tm (as detected by AAS) in urine with time after an initial intravenous infusion of 0.15 mmol of TmDOTP5− at a known rate. The model was equivalent to that of Wedeking et al. (31), who described the pharmacokinetics of GdDTPA2− and99MTcDTPA in rats and included three exchanging compartmental spaces: plasma, a fast exchanging space (FES), and a slow exchanging space (SES) plus a urine compartment. The kinetic rate constants for movement of TmDOTP5− between these compartments and into urine were initially set equal to those given by Wedeking et al. (31) and later altered to optimize the fit to our experimental data.
Statistical data analysis. All data are means ± SD. Statistical significance was determined by using ANOVA or the Student t-test, as appropriate. A P value < 0.05 was considered significant.
TmDOTP5− infusions over extended periods.
As reported previously (2, 3, 26), infusion of 80 mM TmDOTP5− into rats by using the infusion protocol outlined in materials and methods produces a modest decrease in mean arterial blood pressure. Blood samples removed from rats during such a protocol have been analyzed for total Tm, Ca, and Mg by AAS; those results are shown in Fig. 1. Although total concentations of Ca and Mg ([Ca2+]totand [Mg2+]tot, repectively) in blood tended to drop during early periods of TmDOTP5− infusion (compare 0- to 10- vs. 13- to 18-min infusion periods), these differences were not significant. Thereafter, [Mg2+]totremained constant throughout the remaining infusion periods, whereas [Ca2+]totincreased in proportion to total concentration of TmDOTP5− (Fig.2). A linear regression of the [Ca2+]totdata shown in Fig. 2 (n = 12 animals) indicated that [Ca2+]tot= 0.30 ± 0.05 TmDOTP5−concentration + 1.81 ± 0.13.
Biodistribution of two different ECS markers. Either 80 mM TmDOTP5− or 80 mM CoEDTA− was infused into live animals for 24 min by using the protocol described inmaterials and methods. The infusions were then discontinued, and the heart, liver, one thigh muscle, and one kidney were removed, a blood sample was collected, and urine was collected from the bladder. All tissues were analyzed for total Co and Tm by AAS. Total Tm and Co levels in all skeletal muscle were determined from the thigh muscle analyses by assuming that skeletal muscle was 45% of total animal body weight (8). Similarly, total Tm and Co concentrations in blood were determined by assuming that total blood volume was 6.5% of total body weight (11). Those results, expressed as percentage of total TmDOTP5− or CoEDTA− infused at 24 min, are shown in Fig. 3. These data show that the tissue distribution of the two ECS markers paralleled one another, but percent Tm was typically lower than percent Co (this was especially evident in those tissues containing greater amounts of ECS marker, such as kidney, muscle, blood, and urine). The average Tm/Co in kidney, muscle, blood, and urine was 0.88 ± 0.04. Because the amount detected in heart and liver was relatively small, the Tm/Co ratio was not significantly different from unity in those tissues.
The results described above suggest there may be another tissue compartment that sequesters a significant amount of TmDOTP5−, but not of CoEDTA−. To examine this question further, we measured the washout kinetics of TmDOTP5− from tissue after an initial infusion of the ECS marker for 24 min, followed by continued infusion of normal saline for an additional 156 min. The experimental data for appearance of Tm in urine during infusion of saline are shown in Fig. 4 (experimental points on top curve). At 180 min, the infusion was discontinued, and tissues were removed and analyzed as described above. An analysis of those data indicated that ∼81% of the total TmDOTP5− infused after 24 min (0.15 mmol) was found in urine, but <2% was found in liver, thigh muscle, kidney, heart, and blood at 180 min. Thus ∼18% of the TmDOTP5− could not be accounted for in those tissues examined at 180 min. This is shown as the single data point on the curve labeled “bone” in Fig. 4 (see below).
[Na+]iand relative ECS in tissue.
Although the absolute amount of TmDOTP5− and CoEDTA− in the various tissues does appear to differ after an identical infusion period, the ratio of marker in tissue to blood plasma was constant. Assuming that [Tm]plasma = [Tm]ECS, [Co]plasma = [Co]ECS, and [Na]plasma = [Na]ECS after a suitable equilibration period, the relative ECS (rECS), and hence the [Na+]i, could be determined by using the method outlined in Scheufler and Peters (25). The relevant equations are as follows Equation 1 Equation 2rECS, [Na+]i, and rDW values in hearts and livers removed from live rats after infusion of either TmDOTP5−or CoEDTA− are summarized in Table 1. These parameters were not significantly different (P < 0.05) for the two ECS markers in either heart or liver (although values for the two tissues did differ). Both markers indicated that rECS was larger and rDW was smaller in heart than in liver. Despite these differences, [Na+]idid not differ between the two tissues.
TmDOTP5− has been shown to be effective as an in vivo shift reagent for resolving23Na-NMR signals of intra- and extracellular Na+ (2, 3, 19, 22,26). Prior NMR experimental results suggest that this compound distributes into all ECS and does not cross the blood-brain barrier (2). The highly charged phosphonate surface of this molecule does, however, clearly alter Ca2+homeostasis. Although free Ca2+concentration (as measured by a Ca-specific electrode; see Ref.3) has been shown to decrease by ∼10–20% during infusion of TmDOTP5− into rats, the data shown here indicated that [Ca2+]totdoes increase in parallel with total shift reagent concentration in blood (see Fig. 2). The shift reagent is also known to bind divalent Mg2+ (with a lower affinity than Ca2+; see Ref. 23), but analysis of total blood Mg2+indicated that the concentration of this ion was not substantially altered during infusion of TmDOTP5− into live rats. This apparent difference in homeostasis between Ca2+ and Mg2+ parallels the affinity of TmDOTP5− for these two ions (23).
The tissue distribution of TmDOTP5− in live anesthetized rats parallels the distribution of CoEDTA− in heart, liver, kidney, muscle, blood, and urine after infusion of identical amounts of each ECS marker (Fig. 3). Although the absolute amount in each tissue was lower for TmDOTP5− than for CoEDTA−, these data show that both compounds act as typical low- molecular-weight extracellular agents. The AAS results (Table 1) indicate that both compounds were equally effective as markers of ECS in heart and liver. rECS was smaller in the in vivo hearts (average 0.16) than in isolated perfused hearts (1), reflecting the known tendency of isolated, perfused hearts to become edematous. rECS, as measured by either marker in the in vivo rat liver (0.11 ± 0.02), was significantly lower than rECS for the in vivo heart (0.16 ± 0.03), but [Na+]iin the two tissues was identical. Mavroudis et al. (20) have reported values of rECS = 0.15 and [Na+]i= 20.2 mM for in vivo rat heart using radioactive inulin as an ECS marker. Both agree well with values found here with the use of either TmDOTP5− or CoEDTA− as an ECS marker. Blum et al. (4) and Bansal et al. (2) have measured [Na+]iin the in vivo rat liver using NMR techniques, finding values of 26 ± 7 and 21 ± 6, respectively. Again, both agree well with the values determined here by AAS.
The in vivo clearance rate of TmDOTP5− has been measured previously by NMR in rat kidney (26), liver (3), and skeletal muscle (2). The shift reagent was found to clear with similar time constants from kidney and liver (12–13 min) but with a longer time constant from muscle (53 min). These results were consistent with elimination of TmDOTP5− from at least two tissue compartments with different time constants, similar to that observed for anionic MRI contrast agents (31). Data presented here indicate that ∼18% of the total TmDOTP5− infused into animals was not detected in the tissues after 180 min. Because highly charged phosphonates of this type are known to localize on bone surfaces (29), a reasonable assumption is that much of the missing TmDOTP5− remained bound to bone after the 180-min washout period (we were unable to fully dissolve bone for quantitative analysis by AAS). G. Kiefer (unpublished observations) has examined the biodistribution of159GdDOTP5−in rats and found that 55% of the total injected radioactive dose was localized on bone surfaces after 30 min. When this experiment was repeated by using159GdDOTP5−doped with millimolar quantities of unlabeled GdDOTP5−, ∼25% was found tightly bound to bone after 2 h (G. Kiefer, unpublished observations). This suggests that the binding sites for these highly charged complexes on bone surfaces may be saturable.
The clearance data of Fig. 4 appear to be minimally biphasic, consistent with a combined rapid clearance of TmDOTP5− from blood plus slower clearance from less highly perfused tissue such as muscle and, perhaps, bone. Using the kinetic constants reported by Wedeking et al. (31) for charged contrast-agent complexes, we initially modeled the data of Fig. 4 to a three-compartment model, including plasma, a FES, and a SES. That model predicted that the appearance of Tm in urine would continually increase with time (even after 180 min) and not level off as the experimental data show. This means that either the kinetic rate constants reported by Wedeking et al. (31) are too slow to model the TmDOTP5− data well or the model is too simple. Subsequently, a fourth, even slower exchanging, space was added to the model [labeled as bone in Fig.4, based on the biodistribution data of G. Kiefer (unpublished observations)], and the kinetic constants were adjusted so that total Tm on bone at 180 min equaled the missing 18% (0.028 mmol) that we failed to detect in the remaining tissues in our experiments. The solid curves of Fig. 4 show the time-dependent distribution of TmDOTP5− predicted by this model. The kinetic rate constants for distribution between the plasma, FES, SES, and urine were double those reported by Wedeking et al. (31), and the rate of TmDOTP5− removal from bone was an order of magnitude slower than the remaining kinetic constants. Because we do not have plasma levels of Tm during the 180-min washout period, we cannot say with any certainty that this kinetic model is unique, but the general shape of the appearance of TmDOTP5− in urine agrees reasonably well with that observed. The poorest agreement between the results predicted by this kinetic model and the experimental data comes at the early time points. This may be due to a short time delay between collection of TmDOTP5− in the tubules of the kidney and its actual appearance in the bladder.
Two other pieces of experimental evidence suggest that the model presented in Fig. 4 is reasonable. First, the time constants for washout of TmDOTP5− from liver (3) and kidney (26) have been measured at 12–13 min, whereas the time constant for washout from muscle was considerably longer (53 min; Ref. 2). These time constants are similar to those predicted by the kinetic model for washout of TmDOTP5− from the FES (17 min) and SES (38 min), respectively. Second, if one estimates the blood volume of a 225-g rat at 14.6 ml (6.5%; Ref. 11), then the concentration of TmDOTP5− in plasma at 24 min according to the model would be ∼1.7 mM, similar to that found experimentally (Fig. 1).
In summary, we have shown that TmDOTP5− can be used as a marker of all ECS in various rat tissues in vivo. The analytical comparisons with CoEDTA−, the rates of TmDOTP5−appearance in urine, and the pharmacokinetic modeling results all indicate that a significant fraction of TmDOTP5− is bound to bone in vivo. This, however, does not detract from its potential usefulness as an ESC marker in vivo. Although the tissue samples were destroyed for analytical purposes in the experiments reported here, the NMR hyperfine shift properties of TmDOTP5−potentially make this a unique ECS marker for in vivo applications. We have shown that 31P-NMR signal of TmDOTP5− is easily detected in tissue at levels comparable to those used here (26), so the total amount (μmol) of ECS marker in any given volume of tissue can be determined by 31P-NMR. Furthermore, the magnitude of the hyperfine shift induced in the extracellular 23Na resonance has been shown (32) to be directly proportional to the concentration (μmol/ml) of TmDOTP5− in the ECS. Thus the combination of these two measurements provides a direct measure of the ECS volume without destroying the tissue. This unique feature of TmDOTP5−has allowed us to monitor both the extracellular and intracellular concentrations of Na+ in vivo (32).
The technical help of Andrea Wiethoff and Martha Germann is gratefully acknowledged.
Address for reprint requests: A. D. Sherry, Dept. of Radiology, The Mary Nell & Ralph B. Rogers Magnetic Resonance Center, Univ. of Texas Southwestern Medical Center, 5801 Forest Park Rd., Dallas, TX 75235-9085.
This research was supported in part by a Division of Research Resources Biomedical Magnetic Resonance Facility Grant P41-RR-02584, a grant from the Robert A. Welch Foundation (AT-584), and by a Clinical Investigator Award from the Department of Veterans Affairs.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 1998 the American Physiological Society