- Email: [email protected]
31P NMR studies of energy metabolism in perfused rat kidney
JOURNAL
OF SURGICAL
RESEARCH
35, 373-382
(1983)
31P NMR Studies of Energy Metabolism ROBERTS.
in Perfused
RHODES, M.D.,JoYcEE.JENTo~,PH.D., AND ANN V. ROBINSON
Rat Kidney
RICHARD G. BARR,PH.D.,
Departments of Surgery and Biochemistry, Case Western Reserve University School of Medicine, 2074 Abington Road, and the University Hospitals of Cleveland, Cleveland, Ohio 44106 Submitted
for publication
April
6, 1983
Conventional biochemical analyses have demonstrated significant alterations in high-energy phosphate metabolism during shock, but the time course of these changes cannot be followed in individual animals because these analyses are invasive and destructive. This study sought to evaluate the utility of “P NMR as a means of following phosphorus metabolites under various conditions, including those designed to model the shocked state. Twenty adult albino rats were subject to a modified Wiggee’ model of hemorrhagic shock lasting from 5 to 140 min. ATP was determined on extracts of the kidneys of each animal both by a biochemical assay and by integration of ,‘P NMR resonance signals. The equation for renal ATP content plotted versus time for enzymatically determined ATP was 1.79 - 0.0097x (r = 0.83, P < 0.01) as compared to 1.76 - 0.0093x (r = 0.69, P < 0.01) for NMRdetermined ATP. Isolated, normal rat kidneys perfused with oxygenated, modified Krebs’ solution while in the NMR spectrometer maintained normal ATP levels for several hours. ATP/ADP ratios were greater than those observed by conventional enzymatic analysis. Temporary anoxia, induced by substituting 100% N2 for 95% 0*:5% COI, resulted in decreases in ATP content, which reverted to normal with reinstitution of oxygenation. Intracellular pH changed in accordance with perfusate pH during anoxia. It is concluded that “P NMR studies of the perfused rat kidney have immediate application for the nondestructive study of energy metabolism in shock and &hernia.
INTRODUCTION
high-energy metabolites in heart [6], skeletal muscle [ 141,and kidney [ 151.They have confirmed previous observations using classical biochemical techniques that the levels of these metabolites diminish or disappear with ischemia. However, there have been few attempts to quantitate the relationship between conventional and NMRdetermined tissue levels of ATP and other phosphorus-containing metabolites under these circumstances. This study aimed to further assessthe applicability of NMR to studies of shock and &hernia. Specifically, it sought to (1) quantitatively assessthe sensitivity of 31PNMR to low concentrations of ATP observed in shock; and (2) develop, using 31PNMR analysis of the perfused rat kidney as a model, a system to further study the effects of shock and ischemia.
Nuclear magnetic resonance (NMR) spectroscopy is a powerful tool that has many advantages over conventional biochemical techniques: it is noninvasive and on-line; it does not require freeze-clamping; it can detect compartmentation of metabolites; it detects only “mobile” compounds; and it can be used to assesssteady-state kinetics [3]. The disadvantages of NMR are that it is insensitive, resonance signal assignment may not always be clear, there are logistical problems with the magnet, and it is expensive. Alterations in bioenergy metabolism following shock or ischemia have profound implications regarding metabolic processesand/ or cell viability. The ability of “P NMR spectroscopy to quantitatively analyze adenine nucleotides, intracellular pH, creatine phosMATERIALS AND METHODS phate, and sugar phosphatescould be of great value as a research tool in this area. Previous Cell extracts. Male Sprague-Dawley rats, investigators have used 31PNMR to identify weighing 180-350 g, were anesthetized by in373
0022-4804/83 Copyright All
rights
0
$1.50 1983
of reproduction
by
Academic in any
hers. form
Inc. reserved.
374
JOURNAL
OF SURGICAL
RESEARCH: VOL. 35, NO. 5, NOVEMBER
traperitoneal injection of pentobarbital, 4 mg/ 100 g body wt. A PE 50 polyethylene catheter was inserted into the left femoral artery and 500 units of heparin administered through this line. The cannula was connected to a mercury manometer and a sterile syringe via a threeway stopcock. Body temperature was monitored with a telethermometer and maintained at 35.5-37.5’C. Hemorrhagic shock was induced by slowly withdrawing blood until a mean arterial pressure of 50 mm Hg was obtained. The pressure was held at 50 mm Hg for 30 min and then additional blood was withdrawn to reach and maintain a mean arterial pressure of 30 mm Hg. At intervals ranging from 5 to 140 min following the onset of hemorrhage the abdomen was opened and the kidneys were exposed. The kidneys were freeze-clamped in vivo using Wollenberger tongs cooled in liquid nitrogen and then excised and stored in liquid nitrogen. Twenty animals were subject to this shock procedure and two similarly prepared but nonshocked animals served as controls. Perchloric acid extracts of the kidneys were prepared by the method of Garlick et al. [6]. The frozen tissue was ground to a powder and homogenized in 3% perchloric acid (3 ml perchloric acid/g kidney). The precipitate was removed by centrifugation and the supematant neutralized to pH 7 .Owith saturated potassium carbonate. The same extracts were used for both enzymatic and NMR determination of ATP. Biochemical determination of ATP content in the extracts was performed using a commercially available, enzymatic method (Sigma Chemical Co., St. Louis, MO.). For 3’P NMR analysis, extracts from both kidneys of a rat were suspended in 9.0 ml of a 10% DzO solution made 5 mM in EDTA. Samples were stored frozen to prevent hydrolysis of ATP. Perfusion experiments. Male SpragueDawley rats, weighing 180-350 g, were anesthetized by intraperitoneal injection of pentobarbital, 4 mg/ 100 g body wt. The abdomen was opened and the aorta and right kidney were exposed. Three ligatures were placed around the aorta, one proximal to the origin of the superior mesenteric artery and two distal
1983
to the left renal vein. Both the right adrenal artery and the superior mesenteric artery were exposed and ligated. A PE 190 cannula was inserted up the aorta to the origin of the renal artery, heparin (500 USP units) was injected to prevent clotting, and the proximal and distal ligatures were then tied. On occasion it was necessary to also ligate the right spermatic artery. The right kidney was immediately flushed with cold perfiisate and excised. Less than 2 min elapsed from aortic ligation to introduction of perfusate. The cannulated, perfused kidney was transferred to a Teflon holder designed to hold the kidney in a 20-mm NMR tube (Fig. 1). The renal vein was not cannulated but allowed to drain freely. The total operative procedure from incision to initiation of perfusion was less than 20 min. The perfusion fluid was high-bicarbonate Krebs-Henseleit buffer (pH 7.4) containing 5 mM D-glucose. Final concentrations of salts were NaCl, 118.1 mM; KCl, 5.2 mM; KI-12P04, 1.2 mM, MgS04, 2.6 mA$ CaC&, 2.5 mM; and NaHC03, 24.8 mM. Immediately prior to perfusion the perfusate was hltered through a 0.22~pm Millipore filter and then gassed for at least 20 min with 95% 02:5% C02. This solution was then augmented with 50 g/liter of previously filtered bovine serum albumin (fraction V) [4, 131. The perfusion apparatus (Fig. 1) was a modification of the jacketed oxygenator described by Krebs [ 81 for liver perfusion. It consists of a water-jacketed oxygenator, a constant-temperature circulator, gas humidifier, nylon mesh filter, two stopcocks for venous and arterial sampling of the perfusate, and a Masterflex pump (Model No. 7565) with two heads. One head provides pulsatile arterial inflow and the other facilitates venous return. Flow rates were 35-40 ml/min [ 131. The perfused kidneys were maintained at 37°C. 3’P NMR spectra were obtained at 72.896 MHz on a Bruker WH180/270 Fourier transform NMR spectrometer equipped with a Nicolet 1180 computer. For extracts, the pulse width was 32 psec (70” flip angle) and the relaxation delay was 5 sec. All spectra of extracts were taken at room temperature and
375
RHODES ET AL.: “P NMR IN PERFUSED RAT KIDNEY WATER JACKETED OXYGENATOR
TWO CHANNEL ROTARY WLSATILE PUMP
GAS
TANKS
PERFUSATE RESERVOIR
FIG. 1. Schematic diagram of a perfused kidney in a 20-mm NMR tube. Neither the renal vein nor ureter were cannulated but rather were allowed to drain freely. A j-mm tube containing DZO (not shown) was placed between the kidney and the NMR tube wall.
were proton decoupled. Nuclear Overhauser effects were not corrected for but visual comparison of a proton-decoupled and a protoncoupled spectrum showed no major intensity differences. At least 1000 scans were collected for each extract spectrum. For each perfused kidney spectrum, 400 scans were collected using a relaxation delay of 2Y2 sec. Diethyl ethyl phosphonate (7.14% v/v in D20) was used as a secondary external reference and integration standard for extracts. Approximately 10 ~1 of reference in a Wilmad 520-2 coaxial inner cell capillary tube was placed in a 5-mm NMR tube containing D20. The 5-mm tube was held coaxially in the 20mm NMR tube by use of Teflon spacers. The reference sample was always totally contained in the receiver field. The chemical shift of the reference was determined by obtaining a spectrum against 85% phosphoric acid. The phosphoric acid reference was taken as 0.00 ppm; diethyl ethyl phosphonate was found to be +38.5 1 ppm downfield. For whole organs a 5-mm NMR tube containing DzO for external lock was placed between the kidney and the 20-mm NMR tube wall. This method allowed the NMR to be
locked without addition of a deuterium source to the perfusing fluid. The spin-lattice relaxation times (Ti) of the phosphorus metabolites in a perfused kidney were measured using the inversion recovery technique. Routine spectra taken before and after collection of the T1 data demonstrated that the metabolite levels were constant during the experiment. Assignment of resonances was based on published chemical-shift data. Resonance areas were calculated with the aid of a planimeter. The line widths at half height, corrected for line-broadening weighting factors, are included for reference to other systems (Table 1). Intracellular pH was calculated from the position of the inorganic phosphate peak in the “P NMR spectra using the pKa and chemical shift limit values of Navon et al. [lo]. Chemical shift values are relative to the unidentified phosphodiester peak, which was identified to have a chemical shift of +0.668 ppm downfield of 85% H3P04. This compound, which remains present after shock and anoxia even after ATP has disappeared, has also been seen by Radda’s group [ 121. Its
316
JOURNAL
OF SURGICAL
TABLE PARAMETERS F’ROM ATION
TIMES,
PERCENTAGE OUR
1
OF THE "P NMR
PERFUSED
RAT
RESEARCH:
SPECIXU OBTAINED SPIN-LATTICE RELAX-
KIDNEYS:
LINEWIDTHS,
ANDESTIMATES
OF TRUE AREA EXPERIMENTAL
DETERMINED CONDITIONS
OFTHE UNDER
Resonance Sugar phosphates Inorganic phosphate Phosphodiester y ATP a ATP j3 ATP “Immobile” phosphates
0.8 + 0.1 1.1 f 0.1
2.5 0.8 0.8 0.8 1.3
k 0.4 f 0.1 + 0.1
+ 0.1 * 0.1
38.5 32.5
87 72
8.7 32.5 42.3 32.5 1728
43 103 103 89 70
Note. Errors in measurement are approximately 10%. ’ Area of resonance with a delay time of 2 set/area of resonance with a delay time of 4 set, all for spectra obtained with a 90” RF pulse. Errors are estimated to be 10%.
VOL.
35, NO.
5, NOVEMBER
1983
decreased progressively with the duration of shock. ATP content determined by NMR spectroscopy followed a similar time course to that determined biochemically (Fig. 3). The equation for the line-of-best-fit for renal ATP content plotted versus time for enzymatically determined ATP was 1.79 - 0.0097x (standard error of the estimate -t 0.33, I = 0.83, P < 0.01) as compared to 1.76 - 0.0093x (standard error of the estimate f 0.49, r = 0.69, P < 0.01) for 3’P NMR-determined ATP. It can be seen that the lines essentially correspond, the differences between the slopes being insignificant. Perfusion Experiments
The second phase of this study demonstrated that prolonged perfusion of isolated organs could be achieved within the magnet and that 3’P NMR could be used to follow chemical shift is pH insensitive from pH 4 to the energetic state of the kidney. Three isolated, normal rat kidneys were perfused for 3 11, well within the physiologic range. hr while in the magnet of the NMR spectrometer. Spectra with a signal-to-noise ratio RESULTS of approximately 20 for the phosphate peak Cell Extracts could be obtained every 15 min. A typical A typical 31P NMR spectrum obtained from spectrum is shown in Fig. 4. Relative ATP an extract of normal rat kidney is shown in and ADP concentrations were calculated from the areas under the @and the y ATP peaks Fig. 2A. The p ATP peak contains contributions from ATP alone, the y ATP peak as described in the previous section. In this system, the AMP resonance overlaps the sugar contains contributions from ATP and ADP, phosphate resonance and cannot be specifiand the cz ATP peak contains contributions from ATP, ADP, and NAD. The fact that the cally resolved. The calculated values for T, for the phos/3 ATP peak and the y ATP peak encompass comparable areas indicates that little ADP is phorus atoms in the metabolites in a perfused present. The amount of NAD present is ob- kidney are listed in Table 1. Our T1 values, tained by subtracting the area under the y except that for inorganic phosphate, agree with data reported for whole cells studies by 31P ATP peak from the area under the cy ATP peak. This spectrum compares favorably with NMR [5, 171. The T, value for the inorganic those reported from other laboratories [3, 161. phosphate resonance which we and others have measured was obtained from the comFigure 2B demonstrates a spectrum of a kidney extract after 70 min of shock. Note that the posite resonances of intracellular and perhisate phosphate. We estimate the contribution of /3 ATP peak is now smaller than the y ATP peak. The difference indicates the amount of perfusate phosphate to the total phosphate signal in our system to be considerably less ADP present. The ATP content of the nonshocked kidthan half of the total, whereas the situation neys determined biochemically was similar to was reversed for several of the whole-cell studthat reported previously [7, 111. Among 20 ies. Thus, the T, value we obtained probably animals subject to shock, renal ATP content reflects that for intracellular phosphate. On
RHODES ET AL.: ,‘P NMR IN PERFUSED RAT KIDNEY
311
NORMAL
A
KIDNEY Ref
‘i
EXTRACT
Sugar Phosphoie
PDE \
BI
70
mm
dATP n!ADP NAD
SHOCK
KIDNEY EXTRACT
FIG. 2. “P NMR spectra of extracts from a normal (A) and a shocked kidney (B). These spectra were obtained under conditions described under Materials and Methods. The y, (I, and /3 labels refer to the individual phosphates of ATP and ADP. Pi is the resonance peak for inorganic phosphate. PCr is the resonance peak for phosphocreatine, and PDE is the resonance peak for the unidentified phosphodiester compound.
the other hand, Ackerman et al. [I] have shown that the presence of bovine serum albumin causes the T, of the perfusate phosphate to decrease from 5.5 set [ 171 to a value equivalent to that of renal tissue. Thus, we cannot distinguish the relative contributions of perfusate phosphate from intracellular phosphate in our system on the basis of their T, values. Both the decrease in perfusate T,, due to the addition of bovine serum albumin, and the relatively large contribution of intra-
cellular phosphate contribute to the short T, value which we measured for this resonance. To estimate the relative concentrations of all phosphorus metabolites in our system, we chose a 2%~set delay, even though this increased the signal-to-noise ratio at a constant accumulation time per spectrum relative to the rapid pulse sequences used by other investigators [9]. The last column in Table 1 gives the percentages of the areas of phosphorous resonances obtained using a delay
378
JOURNAL
OF SURGICAL
RESEARCH: VOL. 35, NO. 5, NOVEMBER
-
2.4 ;: t s E
2.2 .
NMR
---o--ENZYMATIC
(r=
69.pc.01)
(r=
83,
1983
p < 01 )
2.0 I 8
t is 1 P e
16
g
I .o
14 I .2
01
0
,,,,,,',,,,y,, 0
20
40
DURATION
60
00
OF
SHOCK
100
120
140
(minutes)
FIG. 3. The effect of the duration of shock on the ATP content of renal extracts. Individual values for enzymatically determined ATP are shown by open circles and the dashed line-of-best-fit. Individual values for NMR-determined ATP are shown by solid circles and the solid line-of-best-fit.
time of 2 set compared to a delay time of 4 set using a 90” pulse. These values demonstrated that, with the exception of the inorganic phosphates and phosphodiester resonances, our spectra, obtained with a delay of 21/2 seconds and pulse width of 70”) reliably reflected the concentrations of the phosphorus metabolites. ATP levels remained relatively constant throughout the perfusion. The fi and y ATP
peaks were of equal area, given the error of determining peak areas. ATP/ADP ratios, one measure of the state of energy metabolism, were, therefore, usually greater than 10 to 1. The constancy of the integrated areas for each resonance during the 180 min of perfusion are shown in Table 2. The initial scan, the final scan, and the mean -C the standard deviation for all scans during the perfusion are listed for each of the three normal rat kid-
PDE ‘i Ref.
Sugar Phosphatb
YATP
FIG.
aATP aADP NAD
,
4. “P NMR spectrum from a normal perfused rat kidney.
379
RHODES ET AL.: “P NMR IN PERFUSED RAT KIDNEY TABLE 2 CHANGES
IN THE AREAS OF
PHOSPHORUSRESONANCESOVER THE TIME COURSEOF A PERFUSION EXPERIMENT IN THREE RAT IQDNEYS Time of scan Kidney
15 min
180 min
All (mean k SD)
Sugar phosphates
A B C
141 93 55
95 77 64
112 f 14 19 f 14 51 f 10
Inorganic phosphate
A B C
143 116 72
154 82 12
153 f 13 86k 17 67+ I
PDE
A B C
36 37 13
35 23 26
31 + 10 26 f 11 17k 6
y ATP
A B C
40 26 23
26 49 44
42 f 14 42k 7 31+ I
(YATP
A B C
132 67 39
78 71 56
92 k 22 78 + 14 57 + 11
/3 ATP
A B C
58 37 26
31 33 30
482 40+ 35k
Resonance
15 8 7
Note. Areas were determined with the aid of a planimeter and are presented in arbitrary units. Values of each of the three kidneys (A, B, and C) are presented in the same order for each “P resonance.
neys. Note that the variations between the kidneys are larger than any time-dependent changes for individual kidneys. The intracellular pH for perfused rat kidneys was constant at 7.33 f 0.07 and reproducible for the three separate kidneys (Fig. 5). This value agrees with the value of 7.19 f 0.1 80 -
6.5 60 0
I 20
I 40
I 60 TIME
I 80
I 100
, 120
I 140
(m!n)
FIG. 5. Intracellular pH during the course of perfusion of three normal rat kidneys.
previously reported as the internal pH for perfused rats [12]. In three additional perfusions, temporary anoxia was created by switching the gas mixture from 95% 02:5% CO* to 100% nitrogen. During these anoxic periods ATP decreased and intracellular pH increased.. The way in which intracellular pH follows extracellular pH during anoxic challenges is illustrated for one of the kidneys in Fig. 6. In two anoxic challenges, when oxygen was resumed within 20 min, ATP and both internal and external pH rapidly returned to near normal levels. Figure 7 illustrates the phosphorus metabolite levels for this same kidney as displayed in the 3’P NMR spectra, which were present before anoxia (7A), after 20 min of anoxia (7B), and, finally, after two cycles of anoxia and reoxygenation (7C).
380
JOURNAL
OF SURGICAL
RESEARCH: VOL. 35, NO. 5, NOVEMBER
1983
dent by maintenance of energy metabolism. ATP/ADP ratios were usually estimated to be 10 or greater, much larger than the usual values of 1.52,O reported with conventional techniques. These greater ratios are consistent 6.
1% 0
1 20
Nz
I 40
“2
I 60
Ne I 80
TIME
(mln)
02 I 100
I 120
A
FIG. 6. The relationship of intracellular pH to pet&ate pH during periods of anoxia in a single kidney. Intracellular pH values are shown with solid circles. Perfusate pH values are shown with open circles. The low initial value for intracellular pH is probably due to insufficient time for stabilization after initiation of perfusion.
DISCUSSION
The current study demonstrates the utility of “P nuclear magnetic resonance (NMR) spectroscopy to study metabolic effects of shock and ischemia. The major reservation in using NMR for such studies is its relative insensitivity to low concentrations of metabolites. During ischemia, tissue ATP content decreases significantly. If the threshold for NMR detection of ATP under these circumstances is too high, it might be difficult to quantitate effects of therapeutic interventions. Correlation between conventionally and NMR-determined ATP in cardiac muscle appears satisfactory [ 121, but such studies have not been previously reported for the kidney. To establish “P NMR as a method for quantitative determination of ATP levels during ischemic states, we first compared ATP levels, determined by both NMR and classic enzymatic methods, in extracts from kidneys of shocked rats. A highly significant correlation does exist between the two methods, as shown by the near correspondence of the lines-ofbest-fit (Fig. 3). Therefore, NMR appears to have sufficient sensitivity to be useful in following the metabolic consequences of shock and other ischemic states. Having established that 3’P NMR was sufficiently sensitive for metabolic studies during ischemia, we proceeded to determine whether the perked rat kidney would serve as a model for dynamic studies of energy metabolism via 3’P NMR. Excellent tissue viability was evi-
B
C
I
20 mmutes
o*
20 mtnutes
NE
30
mmutes
0,
20
mmutes
N2
20
mmutes
O;! I
FIG. 7. “P NMR spectra from (A) a normal perfused rat, (B) the same kidney after 20 minutes of 100% nitrogen, and (C) the same kidney after a total of two cydes of nitrogen and reoxygenation.
RHODES ET AL.: “P NMR IN PERFUSED
with those previously reported with NMR in viva [2]. Energy charge could not be calculated because of the inability to differentiate AMP from the sugar phosphate resonance and because, in a healthy kidney, ADP could not be distinguished from errors in integration of the @and y ATP signals. Phosphocreatine levels were low in the kidneys, a finding also noted by others [2]. Anoxia, induced by substituting 100% nitrogen for oxygen:carbon dioxide, resulted in the anticipated decrease in high-energy phosphates. Similar alterations have been noted by NMR following ischemia of the heart [6] and kidney [ 161. The great advantage of NMR is its ability to follow these changes sequentially in a single, isolated organ. An additional advantage of NMR is the ability to analyze metabolic interrelationships. Thus, previous investigators observed an increase in the sugar phosphate resonance during anoxia that decreased toward normal values when ATP levels were restored postanoxia. Similar changes were noted in our model except that we observed an increase rather than the anticipated decrease in intracellular pH during anoxia. This was probably caused by the use of 100% nitrogen rather than 95% nitrogen:5% carbon dioxide to induce anoxia. This loss of gaseous CO2 caused the pH of the perfusate to rise. Radda et al. [ 121 previously demonstrated that the intracellular pH of the perfused kidney closely follows changes in the external pH during acidosis. This report demonstrates that this is also the case during alkalosis (Fig. 6). Although the more physiologically relevant experiment would have been to create anoxia with 95% N2:5% C02, the purpose of this experiment was simply to demonstrate on-line changes in metabolic parameters, including intracellular pH, in perfused organs using “P NMR spectroscopy. Nuclear magnetic resonance appears to be an excellent tool for evaluating metabolic alterations associated with shock and ischemia. The data clearly demonstrate the sensitivity of this method. NMR analyses can quantify alterations in tissue content of high-energy phosphates throughout the range of values
381
RAT KIDNEY
previously seen only with conventional analysis. Since only 15 min is required to obtain well-resolved spectra, a given kidney can be subjected to several different conditions over the several hours of perfusion. Thus, it is possible to assessthe effect of multiple therapeutic interventions in a single experiment. The timedependent changes within a single kidney appear to be less than the variations between kidneys. Furthermore, each NMR spectrum provides information on the interrelationships between ATP, creatine phosphate, sugar phosphates, and intracellular pH. All of these parameters can be followed as a function of time and/or perturbation during the perfusion. Thus, 3’P NMR of perfused kidney is a very powerful probe for studies of energy metabolism under conditions such as those associated with shock and ischemia. REFERENCES 1. Ackerman, J. J. H., Lowry, M., Radda, G. K., Ross, B. D., and Wong, G. G. The role of intrarenal pH in regulation of ammoniagenesis: “P NMR studies of the isolated perfused rat kidney. J. Physiol. 319: 65, 1981. 2. Balaban, R. S., Gadian, D. G., and Radda, G. K. Phosphorous nuclear magnetic resonance study of the rat kidney in vivo. Kidney Int. 20: 515, 1981. 3. Balaban, R. S. Nuclear magnetic resonance studies of epithelial metabolism and function. Fed. Proc. 41: 42, 1982. 4. Bowman, R. H. Methodology for study of isolated perfused rat kidney in vitro. In M. Martinez-Maldonado (Ed.), Methods in Pharmacology, Vol. 4B, Renal Pharmacology. New York: Plenum, 1978. pp. 385-399. 5. Evans, F. E. “P nuclear magnetic resonance studies on relaxation parameters and line broadening of intracellular metabolites of HeLa cells. Arch. B&hem. Biophys. 193: 63, 1979. 6. Garlick, P. B., Radda, G. K., Seeley,P. J., and Chance, B. Phosphorous NMR studies on perfused heart. Biochem.
Biophys.
Res. Commun.
74: 1256,
1977.
7. Hems, D. A., and Brosnan, J. T. Effect of &hernia on content of metabolites in rat liver and kidney in vivo. B&hem.
J. 120: 105, 1970.
8. Krebs, H. A., Cornell, N. W., Lund, P. et al. Isolated liver cells as experimental material. In F. Lundquist and N. Tygstrup (Eds.), Regulation of Hepatic Metabolism. Copenhagen: Munksgaard, 1914. Pp. 726750. 9. McLaughlin, A. C., Takeda, H., and Chance, B. Rapid
382
10.
11. 12.
13. 14.
JOURNAL
OF SURGICAL
RESEARCH: VOL. 35, NO. 5, NOVEMBER
ATP assays in perfused mouse liver by “P NMR. Proc. Nafl. Acad. Sci. USA 76: 5445, 1919. Navon, G., Ogawa, S., Shulman, R. G., and Yamano, T. “P nuclear magnetic resonance studies of Ehrlich ascites tumor cells. Proc. Natl. Acad. Sci. USA 14: 81, 1971. Needleman, P., Passonneau, J. V., and Lowry, 0. H. Distribution of glucose and related metabolites in rat kidney. Amer. J. Physiol. 215: 655, 1968. Radda, G. K., Ackerman, J. J. H., Bore, P., Sehr, P., Wong, G. G., Ross, B. D., Green, Y., Bartlett, S., and Lowry, M. “P NMR studies on kidney intracellular pH in acute renal acidosis. Int. J. Biochem. 12: 217, 1980. Ross, B. D. The isolated perfused kidney (an editorial review). Clin. Sci. Mol. Med. 55: 513, 1978. Seeley, P. J., Sehr, P. A., Gadian, D. G., Garlick,
1983
P. B., and Radda, G. K. Phosphorous NMR in living tissue. In R. A. Swek, I. D. Campbell, R. E. Richards, and R. J. P. Wiiams (Eds.), NMR in Biology. London: Academic Press, 1977. 15. Sehr, P. A., Radda, G. K., Bore, P. J., and Sells, R. A. A model kidney transplant studied by phosphorous nuclear magnetic resonance. Biochem. Biophys. Rei. Commun. 17: 195, 1917. 16. Sehr, P. A., Bore, P. J., Papatheofanis, J., and Radda, G. K. Non-destructive measurement of metabolites and tissue pH in the kidney by 3’P nuclear magnetic resonance. Brit. J. Exp. Pathol. 60: 632, 1979. 17. Ugurbil, K., Guernsey, D. L., Brown, T. R., Glynn, P., Tobkes, N., and Edelman, I. S. “P NMR studies of intact anchorage-dependent mouse embryo fibroblasts. Proc. Nat. Acad. Sci. USA 78: 4843, 1981.
OF SURGICAL
RESEARCH
35, 373-382
(1983)
31P NMR Studies of Energy Metabolism ROBERTS.
in Perfused
RHODES, M.D.,JoYcEE.JENTo~,PH.D., AND ANN V. ROBINSON
Rat Kidney
RICHARD G. BARR,PH.D.,
Departments of Surgery and Biochemistry, Case Western Reserve University School of Medicine, 2074 Abington Road, and the University Hospitals of Cleveland, Cleveland, Ohio 44106 Submitted
for publication
April
6, 1983
Conventional biochemical analyses have demonstrated significant alterations in high-energy phosphate metabolism during shock, but the time course of these changes cannot be followed in individual animals because these analyses are invasive and destructive. This study sought to evaluate the utility of “P NMR as a means of following phosphorus metabolites under various conditions, including those designed to model the shocked state. Twenty adult albino rats were subject to a modified Wiggee’ model of hemorrhagic shock lasting from 5 to 140 min. ATP was determined on extracts of the kidneys of each animal both by a biochemical assay and by integration of ,‘P NMR resonance signals. The equation for renal ATP content plotted versus time for enzymatically determined ATP was 1.79 - 0.0097x (r = 0.83, P < 0.01) as compared to 1.76 - 0.0093x (r = 0.69, P < 0.01) for NMRdetermined ATP. Isolated, normal rat kidneys perfused with oxygenated, modified Krebs’ solution while in the NMR spectrometer maintained normal ATP levels for several hours. ATP/ADP ratios were greater than those observed by conventional enzymatic analysis. Temporary anoxia, induced by substituting 100% N2 for 95% 0*:5% COI, resulted in decreases in ATP content, which reverted to normal with reinstitution of oxygenation. Intracellular pH changed in accordance with perfusate pH during anoxia. It is concluded that “P NMR studies of the perfused rat kidney have immediate application for the nondestructive study of energy metabolism in shock and &hernia.
INTRODUCTION
high-energy metabolites in heart [6], skeletal muscle [ 141,and kidney [ 151.They have confirmed previous observations using classical biochemical techniques that the levels of these metabolites diminish or disappear with ischemia. However, there have been few attempts to quantitate the relationship between conventional and NMRdetermined tissue levels of ATP and other phosphorus-containing metabolites under these circumstances. This study aimed to further assessthe applicability of NMR to studies of shock and &hernia. Specifically, it sought to (1) quantitatively assessthe sensitivity of 31PNMR to low concentrations of ATP observed in shock; and (2) develop, using 31PNMR analysis of the perfused rat kidney as a model, a system to further study the effects of shock and ischemia.
Nuclear magnetic resonance (NMR) spectroscopy is a powerful tool that has many advantages over conventional biochemical techniques: it is noninvasive and on-line; it does not require freeze-clamping; it can detect compartmentation of metabolites; it detects only “mobile” compounds; and it can be used to assesssteady-state kinetics [3]. The disadvantages of NMR are that it is insensitive, resonance signal assignment may not always be clear, there are logistical problems with the magnet, and it is expensive. Alterations in bioenergy metabolism following shock or ischemia have profound implications regarding metabolic processesand/ or cell viability. The ability of “P NMR spectroscopy to quantitatively analyze adenine nucleotides, intracellular pH, creatine phosMATERIALS AND METHODS phate, and sugar phosphatescould be of great value as a research tool in this area. Previous Cell extracts. Male Sprague-Dawley rats, investigators have used 31PNMR to identify weighing 180-350 g, were anesthetized by in373
0022-4804/83 Copyright All
rights
0
$1.50 1983
of reproduction
by
Academic in any
hers. form
Inc. reserved.
374
JOURNAL
OF SURGICAL
RESEARCH: VOL. 35, NO. 5, NOVEMBER
traperitoneal injection of pentobarbital, 4 mg/ 100 g body wt. A PE 50 polyethylene catheter was inserted into the left femoral artery and 500 units of heparin administered through this line. The cannula was connected to a mercury manometer and a sterile syringe via a threeway stopcock. Body temperature was monitored with a telethermometer and maintained at 35.5-37.5’C. Hemorrhagic shock was induced by slowly withdrawing blood until a mean arterial pressure of 50 mm Hg was obtained. The pressure was held at 50 mm Hg for 30 min and then additional blood was withdrawn to reach and maintain a mean arterial pressure of 30 mm Hg. At intervals ranging from 5 to 140 min following the onset of hemorrhage the abdomen was opened and the kidneys were exposed. The kidneys were freeze-clamped in vivo using Wollenberger tongs cooled in liquid nitrogen and then excised and stored in liquid nitrogen. Twenty animals were subject to this shock procedure and two similarly prepared but nonshocked animals served as controls. Perchloric acid extracts of the kidneys were prepared by the method of Garlick et al. [6]. The frozen tissue was ground to a powder and homogenized in 3% perchloric acid (3 ml perchloric acid/g kidney). The precipitate was removed by centrifugation and the supematant neutralized to pH 7 .Owith saturated potassium carbonate. The same extracts were used for both enzymatic and NMR determination of ATP. Biochemical determination of ATP content in the extracts was performed using a commercially available, enzymatic method (Sigma Chemical Co., St. Louis, MO.). For 3’P NMR analysis, extracts from both kidneys of a rat were suspended in 9.0 ml of a 10% DzO solution made 5 mM in EDTA. Samples were stored frozen to prevent hydrolysis of ATP. Perfusion experiments. Male SpragueDawley rats, weighing 180-350 g, were anesthetized by intraperitoneal injection of pentobarbital, 4 mg/ 100 g body wt. The abdomen was opened and the aorta and right kidney were exposed. Three ligatures were placed around the aorta, one proximal to the origin of the superior mesenteric artery and two distal
1983
to the left renal vein. Both the right adrenal artery and the superior mesenteric artery were exposed and ligated. A PE 190 cannula was inserted up the aorta to the origin of the renal artery, heparin (500 USP units) was injected to prevent clotting, and the proximal and distal ligatures were then tied. On occasion it was necessary to also ligate the right spermatic artery. The right kidney was immediately flushed with cold perfiisate and excised. Less than 2 min elapsed from aortic ligation to introduction of perfusate. The cannulated, perfused kidney was transferred to a Teflon holder designed to hold the kidney in a 20-mm NMR tube (Fig. 1). The renal vein was not cannulated but allowed to drain freely. The total operative procedure from incision to initiation of perfusion was less than 20 min. The perfusion fluid was high-bicarbonate Krebs-Henseleit buffer (pH 7.4) containing 5 mM D-glucose. Final concentrations of salts were NaCl, 118.1 mM; KCl, 5.2 mM; KI-12P04, 1.2 mM, MgS04, 2.6 mA$ CaC&, 2.5 mM; and NaHC03, 24.8 mM. Immediately prior to perfusion the perfusate was hltered through a 0.22~pm Millipore filter and then gassed for at least 20 min with 95% 02:5% C02. This solution was then augmented with 50 g/liter of previously filtered bovine serum albumin (fraction V) [4, 131. The perfusion apparatus (Fig. 1) was a modification of the jacketed oxygenator described by Krebs [ 81 for liver perfusion. It consists of a water-jacketed oxygenator, a constant-temperature circulator, gas humidifier, nylon mesh filter, two stopcocks for venous and arterial sampling of the perfusate, and a Masterflex pump (Model No. 7565) with two heads. One head provides pulsatile arterial inflow and the other facilitates venous return. Flow rates were 35-40 ml/min [ 131. The perfused kidneys were maintained at 37°C. 3’P NMR spectra were obtained at 72.896 MHz on a Bruker WH180/270 Fourier transform NMR spectrometer equipped with a Nicolet 1180 computer. For extracts, the pulse width was 32 psec (70” flip angle) and the relaxation delay was 5 sec. All spectra of extracts were taken at room temperature and
375
RHODES ET AL.: “P NMR IN PERFUSED RAT KIDNEY WATER JACKETED OXYGENATOR
TWO CHANNEL ROTARY WLSATILE PUMP
GAS
TANKS
PERFUSATE RESERVOIR
FIG. 1. Schematic diagram of a perfused kidney in a 20-mm NMR tube. Neither the renal vein nor ureter were cannulated but rather were allowed to drain freely. A j-mm tube containing DZO (not shown) was placed between the kidney and the NMR tube wall.
were proton decoupled. Nuclear Overhauser effects were not corrected for but visual comparison of a proton-decoupled and a protoncoupled spectrum showed no major intensity differences. At least 1000 scans were collected for each extract spectrum. For each perfused kidney spectrum, 400 scans were collected using a relaxation delay of 2Y2 sec. Diethyl ethyl phosphonate (7.14% v/v in D20) was used as a secondary external reference and integration standard for extracts. Approximately 10 ~1 of reference in a Wilmad 520-2 coaxial inner cell capillary tube was placed in a 5-mm NMR tube containing D20. The 5-mm tube was held coaxially in the 20mm NMR tube by use of Teflon spacers. The reference sample was always totally contained in the receiver field. The chemical shift of the reference was determined by obtaining a spectrum against 85% phosphoric acid. The phosphoric acid reference was taken as 0.00 ppm; diethyl ethyl phosphonate was found to be +38.5 1 ppm downfield. For whole organs a 5-mm NMR tube containing DzO for external lock was placed between the kidney and the 20-mm NMR tube wall. This method allowed the NMR to be
locked without addition of a deuterium source to the perfusing fluid. The spin-lattice relaxation times (Ti) of the phosphorus metabolites in a perfused kidney were measured using the inversion recovery technique. Routine spectra taken before and after collection of the T1 data demonstrated that the metabolite levels were constant during the experiment. Assignment of resonances was based on published chemical-shift data. Resonance areas were calculated with the aid of a planimeter. The line widths at half height, corrected for line-broadening weighting factors, are included for reference to other systems (Table 1). Intracellular pH was calculated from the position of the inorganic phosphate peak in the “P NMR spectra using the pKa and chemical shift limit values of Navon et al. [lo]. Chemical shift values are relative to the unidentified phosphodiester peak, which was identified to have a chemical shift of +0.668 ppm downfield of 85% H3P04. This compound, which remains present after shock and anoxia even after ATP has disappeared, has also been seen by Radda’s group [ 121. Its
316
JOURNAL
OF SURGICAL
TABLE PARAMETERS F’ROM ATION
TIMES,
PERCENTAGE OUR
1
OF THE "P NMR
PERFUSED
RAT
RESEARCH:
SPECIXU OBTAINED SPIN-LATTICE RELAX-
KIDNEYS:
LINEWIDTHS,
ANDESTIMATES
OF TRUE AREA EXPERIMENTAL
DETERMINED CONDITIONS
OFTHE UNDER
Resonance Sugar phosphates Inorganic phosphate Phosphodiester y ATP a ATP j3 ATP “Immobile” phosphates
0.8 + 0.1 1.1 f 0.1
2.5 0.8 0.8 0.8 1.3
k 0.4 f 0.1 + 0.1
+ 0.1 * 0.1
38.5 32.5
87 72
8.7 32.5 42.3 32.5 1728
43 103 103 89 70
Note. Errors in measurement are approximately 10%. ’ Area of resonance with a delay time of 2 set/area of resonance with a delay time of 4 set, all for spectra obtained with a 90” RF pulse. Errors are estimated to be 10%.
VOL.
35, NO.
5, NOVEMBER
1983
decreased progressively with the duration of shock. ATP content determined by NMR spectroscopy followed a similar time course to that determined biochemically (Fig. 3). The equation for the line-of-best-fit for renal ATP content plotted versus time for enzymatically determined ATP was 1.79 - 0.0097x (standard error of the estimate -t 0.33, I = 0.83, P < 0.01) as compared to 1.76 - 0.0093x (standard error of the estimate f 0.49, r = 0.69, P < 0.01) for 3’P NMR-determined ATP. It can be seen that the lines essentially correspond, the differences between the slopes being insignificant. Perfusion Experiments
The second phase of this study demonstrated that prolonged perfusion of isolated organs could be achieved within the magnet and that 3’P NMR could be used to follow chemical shift is pH insensitive from pH 4 to the energetic state of the kidney. Three isolated, normal rat kidneys were perfused for 3 11, well within the physiologic range. hr while in the magnet of the NMR spectrometer. Spectra with a signal-to-noise ratio RESULTS of approximately 20 for the phosphate peak Cell Extracts could be obtained every 15 min. A typical A typical 31P NMR spectrum obtained from spectrum is shown in Fig. 4. Relative ATP an extract of normal rat kidney is shown in and ADP concentrations were calculated from the areas under the @and the y ATP peaks Fig. 2A. The p ATP peak contains contributions from ATP alone, the y ATP peak as described in the previous section. In this system, the AMP resonance overlaps the sugar contains contributions from ATP and ADP, phosphate resonance and cannot be specifiand the cz ATP peak contains contributions from ATP, ADP, and NAD. The fact that the cally resolved. The calculated values for T, for the phos/3 ATP peak and the y ATP peak encompass comparable areas indicates that little ADP is phorus atoms in the metabolites in a perfused present. The amount of NAD present is ob- kidney are listed in Table 1. Our T1 values, tained by subtracting the area under the y except that for inorganic phosphate, agree with data reported for whole cells studies by 31P ATP peak from the area under the cy ATP peak. This spectrum compares favorably with NMR [5, 171. The T, value for the inorganic those reported from other laboratories [3, 161. phosphate resonance which we and others have measured was obtained from the comFigure 2B demonstrates a spectrum of a kidney extract after 70 min of shock. Note that the posite resonances of intracellular and perhisate phosphate. We estimate the contribution of /3 ATP peak is now smaller than the y ATP peak. The difference indicates the amount of perfusate phosphate to the total phosphate signal in our system to be considerably less ADP present. The ATP content of the nonshocked kidthan half of the total, whereas the situation neys determined biochemically was similar to was reversed for several of the whole-cell studthat reported previously [7, 111. Among 20 ies. Thus, the T, value we obtained probably animals subject to shock, renal ATP content reflects that for intracellular phosphate. On
RHODES ET AL.: ,‘P NMR IN PERFUSED RAT KIDNEY
311
NORMAL
A
KIDNEY Ref
‘i
EXTRACT
Sugar Phosphoie
PDE \
BI
70
mm
dATP n!ADP NAD
SHOCK
KIDNEY EXTRACT
FIG. 2. “P NMR spectra of extracts from a normal (A) and a shocked kidney (B). These spectra were obtained under conditions described under Materials and Methods. The y, (I, and /3 labels refer to the individual phosphates of ATP and ADP. Pi is the resonance peak for inorganic phosphate. PCr is the resonance peak for phosphocreatine, and PDE is the resonance peak for the unidentified phosphodiester compound.
the other hand, Ackerman et al. [I] have shown that the presence of bovine serum albumin causes the T, of the perfusate phosphate to decrease from 5.5 set [ 171 to a value equivalent to that of renal tissue. Thus, we cannot distinguish the relative contributions of perfusate phosphate from intracellular phosphate in our system on the basis of their T, values. Both the decrease in perfusate T,, due to the addition of bovine serum albumin, and the relatively large contribution of intra-
cellular phosphate contribute to the short T, value which we measured for this resonance. To estimate the relative concentrations of all phosphorus metabolites in our system, we chose a 2%~set delay, even though this increased the signal-to-noise ratio at a constant accumulation time per spectrum relative to the rapid pulse sequences used by other investigators [9]. The last column in Table 1 gives the percentages of the areas of phosphorous resonances obtained using a delay
378
JOURNAL
OF SURGICAL
RESEARCH: VOL. 35, NO. 5, NOVEMBER
-
2.4 ;: t s E
2.2 .
NMR
---o--ENZYMATIC
(r=
69.pc.01)
(r=
83,
1983
p < 01 )
2.0 I 8
t is 1 P e
16
g
I .o
14 I .2
01
0
,,,,,,',,,,y,, 0
20
40
DURATION
60
00
OF
SHOCK
100
120
140
(minutes)
FIG. 3. The effect of the duration of shock on the ATP content of renal extracts. Individual values for enzymatically determined ATP are shown by open circles and the dashed line-of-best-fit. Individual values for NMR-determined ATP are shown by solid circles and the solid line-of-best-fit.
time of 2 set compared to a delay time of 4 set using a 90” pulse. These values demonstrated that, with the exception of the inorganic phosphates and phosphodiester resonances, our spectra, obtained with a delay of 21/2 seconds and pulse width of 70”) reliably reflected the concentrations of the phosphorus metabolites. ATP levels remained relatively constant throughout the perfusion. The fi and y ATP
peaks were of equal area, given the error of determining peak areas. ATP/ADP ratios, one measure of the state of energy metabolism, were, therefore, usually greater than 10 to 1. The constancy of the integrated areas for each resonance during the 180 min of perfusion are shown in Table 2. The initial scan, the final scan, and the mean -C the standard deviation for all scans during the perfusion are listed for each of the three normal rat kid-
PDE ‘i Ref.
Sugar Phosphatb
YATP
FIG.
aATP aADP NAD
,
4. “P NMR spectrum from a normal perfused rat kidney.
379
RHODES ET AL.: “P NMR IN PERFUSED RAT KIDNEY TABLE 2 CHANGES
IN THE AREAS OF
PHOSPHORUSRESONANCESOVER THE TIME COURSEOF A PERFUSION EXPERIMENT IN THREE RAT IQDNEYS Time of scan Kidney
15 min
180 min
All (mean k SD)
Sugar phosphates
A B C
141 93 55
95 77 64
112 f 14 19 f 14 51 f 10
Inorganic phosphate
A B C
143 116 72
154 82 12
153 f 13 86k 17 67+ I
PDE
A B C
36 37 13
35 23 26
31 + 10 26 f 11 17k 6
y ATP
A B C
40 26 23
26 49 44
42 f 14 42k 7 31+ I
(YATP
A B C
132 67 39
78 71 56
92 k 22 78 + 14 57 + 11
/3 ATP
A B C
58 37 26
31 33 30
482 40+ 35k
Resonance
15 8 7
Note. Areas were determined with the aid of a planimeter and are presented in arbitrary units. Values of each of the three kidneys (A, B, and C) are presented in the same order for each “P resonance.
neys. Note that the variations between the kidneys are larger than any time-dependent changes for individual kidneys. The intracellular pH for perfused rat kidneys was constant at 7.33 f 0.07 and reproducible for the three separate kidneys (Fig. 5). This value agrees with the value of 7.19 f 0.1 80 -
6.5 60 0
I 20
I 40
I 60 TIME
I 80
I 100
, 120
I 140
(m!n)
FIG. 5. Intracellular pH during the course of perfusion of three normal rat kidneys.
previously reported as the internal pH for perfused rats [12]. In three additional perfusions, temporary anoxia was created by switching the gas mixture from 95% 02:5% CO* to 100% nitrogen. During these anoxic periods ATP decreased and intracellular pH increased.. The way in which intracellular pH follows extracellular pH during anoxic challenges is illustrated for one of the kidneys in Fig. 6. In two anoxic challenges, when oxygen was resumed within 20 min, ATP and both internal and external pH rapidly returned to near normal levels. Figure 7 illustrates the phosphorus metabolite levels for this same kidney as displayed in the 3’P NMR spectra, which were present before anoxia (7A), after 20 min of anoxia (7B), and, finally, after two cycles of anoxia and reoxygenation (7C).
380
JOURNAL
OF SURGICAL
RESEARCH: VOL. 35, NO. 5, NOVEMBER
1983
dent by maintenance of energy metabolism. ATP/ADP ratios were usually estimated to be 10 or greater, much larger than the usual values of 1.52,O reported with conventional techniques. These greater ratios are consistent 6.
1% 0
1 20
Nz
I 40
“2
I 60
Ne I 80
TIME
(mln)
02 I 100
I 120
A
FIG. 6. The relationship of intracellular pH to pet&ate pH during periods of anoxia in a single kidney. Intracellular pH values are shown with solid circles. Perfusate pH values are shown with open circles. The low initial value for intracellular pH is probably due to insufficient time for stabilization after initiation of perfusion.
DISCUSSION
The current study demonstrates the utility of “P nuclear magnetic resonance (NMR) spectroscopy to study metabolic effects of shock and ischemia. The major reservation in using NMR for such studies is its relative insensitivity to low concentrations of metabolites. During ischemia, tissue ATP content decreases significantly. If the threshold for NMR detection of ATP under these circumstances is too high, it might be difficult to quantitate effects of therapeutic interventions. Correlation between conventionally and NMR-determined ATP in cardiac muscle appears satisfactory [ 121, but such studies have not been previously reported for the kidney. To establish “P NMR as a method for quantitative determination of ATP levels during ischemic states, we first compared ATP levels, determined by both NMR and classic enzymatic methods, in extracts from kidneys of shocked rats. A highly significant correlation does exist between the two methods, as shown by the near correspondence of the lines-ofbest-fit (Fig. 3). Therefore, NMR appears to have sufficient sensitivity to be useful in following the metabolic consequences of shock and other ischemic states. Having established that 3’P NMR was sufficiently sensitive for metabolic studies during ischemia, we proceeded to determine whether the perked rat kidney would serve as a model for dynamic studies of energy metabolism via 3’P NMR. Excellent tissue viability was evi-
B
C
I
20 mmutes
o*
20 mtnutes
NE
30
mmutes
0,
20
mmutes
N2
20
mmutes
O;! I
FIG. 7. “P NMR spectra from (A) a normal perfused rat, (B) the same kidney after 20 minutes of 100% nitrogen, and (C) the same kidney after a total of two cydes of nitrogen and reoxygenation.
RHODES ET AL.: “P NMR IN PERFUSED
with those previously reported with NMR in viva [2]. Energy charge could not be calculated because of the inability to differentiate AMP from the sugar phosphate resonance and because, in a healthy kidney, ADP could not be distinguished from errors in integration of the @and y ATP signals. Phosphocreatine levels were low in the kidneys, a finding also noted by others [2]. Anoxia, induced by substituting 100% nitrogen for oxygen:carbon dioxide, resulted in the anticipated decrease in high-energy phosphates. Similar alterations have been noted by NMR following ischemia of the heart [6] and kidney [ 161. The great advantage of NMR is its ability to follow these changes sequentially in a single, isolated organ. An additional advantage of NMR is the ability to analyze metabolic interrelationships. Thus, previous investigators observed an increase in the sugar phosphate resonance during anoxia that decreased toward normal values when ATP levels were restored postanoxia. Similar changes were noted in our model except that we observed an increase rather than the anticipated decrease in intracellular pH during anoxia. This was probably caused by the use of 100% nitrogen rather than 95% nitrogen:5% carbon dioxide to induce anoxia. This loss of gaseous CO2 caused the pH of the perfusate to rise. Radda et al. [ 121 previously demonstrated that the intracellular pH of the perfused kidney closely follows changes in the external pH during acidosis. This report demonstrates that this is also the case during alkalosis (Fig. 6). Although the more physiologically relevant experiment would have been to create anoxia with 95% N2:5% C02, the purpose of this experiment was simply to demonstrate on-line changes in metabolic parameters, including intracellular pH, in perfused organs using “P NMR spectroscopy. Nuclear magnetic resonance appears to be an excellent tool for evaluating metabolic alterations associated with shock and ischemia. The data clearly demonstrate the sensitivity of this method. NMR analyses can quantify alterations in tissue content of high-energy phosphates throughout the range of values
381
RAT KIDNEY
previously seen only with conventional analysis. Since only 15 min is required to obtain well-resolved spectra, a given kidney can be subjected to several different conditions over the several hours of perfusion. Thus, it is possible to assessthe effect of multiple therapeutic interventions in a single experiment. The timedependent changes within a single kidney appear to be less than the variations between kidneys. Furthermore, each NMR spectrum provides information on the interrelationships between ATP, creatine phosphate, sugar phosphates, and intracellular pH. All of these parameters can be followed as a function of time and/or perturbation during the perfusion. Thus, 3’P NMR of perfused kidney is a very powerful probe for studies of energy metabolism under conditions such as those associated with shock and ischemia. REFERENCES 1. Ackerman, J. J. H., Lowry, M., Radda, G. K., Ross, B. D., and Wong, G. G. The role of intrarenal pH in regulation of ammoniagenesis: “P NMR studies of the isolated perfused rat kidney. J. Physiol. 319: 65, 1981. 2. Balaban, R. S., Gadian, D. G., and Radda, G. K. Phosphorous nuclear magnetic resonance study of the rat kidney in vivo. Kidney Int. 20: 515, 1981. 3. Balaban, R. S. Nuclear magnetic resonance studies of epithelial metabolism and function. Fed. Proc. 41: 42, 1982. 4. Bowman, R. H. Methodology for study of isolated perfused rat kidney in vitro. In M. Martinez-Maldonado (Ed.), Methods in Pharmacology, Vol. 4B, Renal Pharmacology. New York: Plenum, 1978. pp. 385-399. 5. Evans, F. E. “P nuclear magnetic resonance studies on relaxation parameters and line broadening of intracellular metabolites of HeLa cells. Arch. B&hem. Biophys. 193: 63, 1979. 6. Garlick, P. B., Radda, G. K., Seeley,P. J., and Chance, B. Phosphorous NMR studies on perfused heart. Biochem.
Biophys.
Res. Commun.
74: 1256,
1977.
7. Hems, D. A., and Brosnan, J. T. Effect of &hernia on content of metabolites in rat liver and kidney in vivo. B&hem.
J. 120: 105, 1970.
8. Krebs, H. A., Cornell, N. W., Lund, P. et al. Isolated liver cells as experimental material. In F. Lundquist and N. Tygstrup (Eds.), Regulation of Hepatic Metabolism. Copenhagen: Munksgaard, 1914. Pp. 726750. 9. McLaughlin, A. C., Takeda, H., and Chance, B. Rapid
382
10.
11. 12.
13. 14.
JOURNAL
OF SURGICAL
RESEARCH: VOL. 35, NO. 5, NOVEMBER
ATP assays in perfused mouse liver by “P NMR. Proc. Nafl. Acad. Sci. USA 76: 5445, 1919. Navon, G., Ogawa, S., Shulman, R. G., and Yamano, T. “P nuclear magnetic resonance studies of Ehrlich ascites tumor cells. Proc. Natl. Acad. Sci. USA 14: 81, 1971. Needleman, P., Passonneau, J. V., and Lowry, 0. H. Distribution of glucose and related metabolites in rat kidney. Amer. J. Physiol. 215: 655, 1968. Radda, G. K., Ackerman, J. J. H., Bore, P., Sehr, P., Wong, G. G., Ross, B. D., Green, Y., Bartlett, S., and Lowry, M. “P NMR studies on kidney intracellular pH in acute renal acidosis. Int. J. Biochem. 12: 217, 1980. Ross, B. D. The isolated perfused kidney (an editorial review). Clin. Sci. Mol. Med. 55: 513, 1978. Seeley, P. J., Sehr, P. A., Gadian, D. G., Garlick,
1983
P. B., and Radda, G. K. Phosphorous NMR in living tissue. In R. A. Swek, I. D. Campbell, R. E. Richards, and R. J. P. Wiiams (Eds.), NMR in Biology. London: Academic Press, 1977. 15. Sehr, P. A., Radda, G. K., Bore, P. J., and Sells, R. A. A model kidney transplant studied by phosphorous nuclear magnetic resonance. Biochem. Biophys. Rei. Commun. 17: 195, 1917. 16. Sehr, P. A., Bore, P. J., Papatheofanis, J., and Radda, G. K. Non-destructive measurement of metabolites and tissue pH in the kidney by 3’P nuclear magnetic resonance. Brit. J. Exp. Pathol. 60: 632, 1979. 17. Ugurbil, K., Guernsey, D. L., Brown, T. R., Glynn, P., Tobkes, N., and Edelman, I. S. “P NMR studies of intact anchorage-dependent mouse embryo fibroblasts. Proc. Nat. Acad. Sci. USA 78: 4843, 1981.