ytosolic calcium in the ischemic rabbit heart : assessment by pH-and temperature-adjusted rhod-2 spectrofluorometry

21 Objectives: Cytosolic calcium ([Ca ] ) mediates ischemia–reperfusion (I /R) injury, but magnitude and time course of I /R-induced i 21 21 [Ca ] overload remain unclear. Fluorescent indicators are commonly used to measure [Ca ] , and have a dissociation constant ( K ) i i d 21 that depends on pH and temperature. We hypothesized that changes of K during I /R lead to misleading interpretations of [Ca ] d i recordings.Methods: (1) In isolated rabbit hearts ( n54 each), intracellular pH was measured during I /R at 37 8C, 208C, and 48C with 31 and without cardioplegic arrest by P-NMR-spectroscopy. (2) K for rhod-2 and calcium was determined at varying pH and temperature d 21 in in vitro experiments. (3) Isolated rabbit hearts were subjected to I /R, and [Ca ] was recorded by surface rhod-2 spectrofluorometry. i 21 21 Finally, [Ca ] was computed using either the conventional K , or the pHand temperature-adjusted K . Results: K (Ca Rhod-2) i d d d remained stable between pH 7.1 and 6.8, but increased exponentially with lower pH and/or temperature. Calculations using a static Kd 21 21 indicated that [Ca ] rose only mildly during warm ischemia and did not rise during cardioplegic arrest, while a large Ca influx i 21 appeared to occur during early reperfusion. When the pH and temperature-adjusted K was used for calculation, [Ca ] rose significantly d i during ischemia (431 637% during 20 min 378C ischemia, and 78 619% during 20 min cardioplegic arrest at 20 8C). During early 21 21 21 reperfusion, [Ca ] decreased rapidly, without significant further [Ca ] elevation. C clusions: In contrast to previous reports, [Ca ] i i i 21 accumulation occurs during unprotected ischemia as well as hypothermic ischemia with cardioplegic arrest, without further net Ca influx on reperfusion. This finding has important implications for timing of protective strategies during myocardial ischemia.  2003 European Society of Cardiology. Published by Elsevier B.V. All rights reserved.


. Introduction
acerbated by uncontrolled influx from the extracellular space, but controversy remains about extent and time It is generally acknowledged that myocardial ischemia-course of ischemia-reperfusion-induced changes in calreperfusion injury provokes cytosolic free calcium cium homeostasis. The pathophysiologic consequences 21 ( [Ca ] ) accumulation [1,2]. During ischemia, several include energy-depleting futile calcium cycling, impaired i 21 mechanisms result in failure of intracellular Ca compart-relaxation, and systolic contractile dysfunction. Fluoresmentalization and extrusion from the cell against high cence indicators are commonly used to assess intracellular 21 concentration gradients. At the onset of reperfusion, in-Ca handling, and the respective techniques are straight- 21 tracellular Ca overload is thought to be further ex-forward in single cells or tissue preparations such as ventricular trabeculae or atrial muscle strips. However, true  intracellular Ca is much more difficult to measure. The The investigation conforms with the Guide for the Care complex methodology problems include light absorbance and Use of Laboratory Animals published by the US by chromatic molecules (i.e. myoglobin, hemoglobin), National Institutes of Health (NIH Publication No. 85-23, 1 tissue autofluorescence (i.e. NAD ), motion artefacts of revised 1996). New Zealand white rabbits (2-2.5 kg) were the beating heart, and the impact of temperature and pH on euthanized by intravenous injection of ketamine (100 mg / 21 Ca -induced signal emission. Signal quality greatly de-kg), and heparin (500 U / kg). Hearts were rapidly excised pends on the fluorescence indicator used, and each in-and placed in 4 8C cold buffer solution. After cannulation dicator has particular advantages and disadvantages. We of the aorta, hearts were perfused in the Langendorff mode 21 have recently introduced the long-wavelength Ca in-at 80 mmHg constant perfusion pressure with modified 21 dicator rhod-2 for recording of beat-to-beat Ca trans-Krebs-Henseleit (KH) buffer (115 mmol / l NaCl, 26 ients, because rhod-2 possesses several advantages over mmol / l NaHCO , 11 mmol / l glucose, 1.8 mmol / l 3 other fluorescence indicators that are of particular impor-MgSO , 1.8 mmol / l KH PO , 4.7 mmol / l KCl, 1.25 4 2 4 tance in the heart. However, the signal of any cation-mmol / l CaCl , and 10 U / l insulin), that had been equili-2 sensitive indicator is affected by changes in intracellular brated with a 95% O / 5% CO gas mixture and passed 2 2 pH as well as temperature, both of which can vary greatly through a 0.2 micrometer nylon filter. The final buffer pH in clinically relevant models of myocardial ischemia and was 7.35-7.45, PO was 550-600 mmHg, and PCO was 2 2 reperfusion [3]. The extent of pH-or temperature-induced 30-40 mmHg. A fluid-filled latex balloon connected to a signal changes depends on the physical properties and micromanometry catheter (Millar Instruments, Houston, 21 binding affinity for Ca and protons, and is unknown for TX) was placed in the left ventricle via the left atrium. most of the indicators used today.
After 30 min stabilization, the hearts were placed in the 31 In order to achieve a better understanding of intracellu-respective perfusion chamber for P NMR spectroscopy lar calcium during ischemia in the intact heart, we mea-or rhod-2 spectrofluorometry. Global ischemia was insured the change in intracellular pH during several is-duced by occlusion of the aortic cannula. Cardiac temperachemia / reperfusion protocols that mimic clinically and ture was monitored throughout the experiment, including experimentally relevant situations, determined the dissocia-the period of ischemia, using a temperature probe inserted 21 tion constant K for rhod-2 and Ca at the respective pH in the right ventricle. During stabilization and reperfusion d and temperature, and determined cytosolic calcium by myocardial temperature was adjusted to 37 8C. During the spectrofluorometry in whole hearts, using the appropriate period of ischemia, temperature was maintained at 37 8C or K . lowered to 20 8C by adjusting the ambient temperature in d the perfusion chamber.

31
2 .2. P NMR spectroscopy Our approach to obtain pH and temperature-corrected Hearts were perfused with phosphate-free modified KH measurements of intracellular calcium using rhod-2 inbuffer using a customized perfusion system. The isolated cluded the following steps: heart in its perfusion chamber was positioned within a 20-mm solenoid radiofrequency coil. NMR spectra were 1. Measurement of intracellular pH in isolated rabbit acquired in an 8.45 Tesla vertical bore Bruker spectrometer hearts subjected to ischemia at 37 8C, 20 8C, and 4 8C (Bruker Instruments, Billerica, MA). Spectra were obwithout cardioplegic arrest, and at 20 8C with high-31 tained by signal averaging 120 scans with a 2-s delay, potassium cardioplegic arrest, using P nuclear magresulting in a time resolution of 4.5 min. Intracellular pH netic resonance. 21 was calculated from the shift of the inorganic phosphate 2. Determination of the specific K for rhod-2 and Ca in d (P ) peak using the following equation [4]: series of in vitro experiments at pH 7.1-6.2 in increi ments of 0.1, each at the above described temperatures. 21 3. Recording of intracellular Ca signals in intact per-pH 5 6.9 1 log [(x 2 3.28) /(5.7 2 x)] fused rabbit hearts using rhod-2 spectrofluorometry, during ischemia at 37 8C without cardioplegic arrest and where x is the position of the P peak with respect to the i at 20 8C with cardioplegic arrest: two commonly used PCr peak in p.p.m. experimental protocols that are of particular clinical Following a stabilization period of 20-30 min, hearts relevance.
were subjected to 45 min ischemia at 37 8C (n54), at 21 4. Computation of the intracellular free Ca concen-20 8C (n54), and at 4 8C (n54). In another group of tration using the specific pH and temperature-corrected hearts, cardioplegic arrest was induced by injection of 1 K for a given protocol and time point. perfusion buffer containing 20 mM K (n54). undiluted Ca EGTA buffer solution at a given tempera- 21 Measurement of beat-to-beat intracellular Ca transture and pH were derived using the web-based computer ients was performed as we have previously described and program Webmaxclite v1.00 (Stanford University, http: / validated in detail [5]. During the 30 min stabilization / www.stanford.edu / |cpatton / maxclite.htm). The free 21 21 21 period, hearts were loaded with the Ca -sensitive dye Ca concentration in the Ca dilution series obtained by was infused over a period of 2 min, immediately followed temperature or cardioplegic arrest, pH increased rapidly by calcium ionophore A23187 (calcimycin) in 10 ml 10% during early reperfusion, reaching near-normal levels 21 21 between pH 7.0 and pH 7.1 at 5 min reperfusion in all Ca solution to maximize Ca entry from the extracellugroups. lar space. Fluorescence was recorded with a time resolution of 40 ms during the infusion, and maximum 21 Fig. 2 depicts a typical series of in vitro rhod-2 21 21 [Ca ] 5 K (Ca Rhod-2) i d fluorescence emission scans, recorded at increasing free 21 Ca concentrations (here: pH 7.1, 20 8C). As expected,

.2. K for rhod-2 and Ca
increasing Ca in the solution results in an increase in 21 21 21 where [Ca ] is the free intracellular Ca concentration, fluorescence intensity. The impact of temperature on Ca -  [Ca ] was recorded during the following ischemia / to 4 8C increased the rightward shift moderately. Hence, it i reperfusion protocols: 20 min unprotected ischemia at should be noted that lowering temperature from 37 8C to 37 8C without cardioplegic arrest followed by 10 min room temperature, as easily happens in experimental reperfusion (n56); 20 min ischemia at 20 8C with cardiopsettings studying ischemic and thus non-perfused hearts, 1 legic arrest (20 mM K ) followed by 10 min reperfusion. has a significant effect on intensity of the fluorescence 21 21 The fluorescence signal was then used to calculate [Ca ] signal. The relationship between pH and Ca -induced for every time point. This is the K (Ca Rhod-2) in a evident that lowering pH from pH 7.1 to pH 6.8 had no d 21 simulated intracellular myocardial environment that was significant effect on Ca -induced rhod-2 fluorescence, determined by del Nido et al. when they first described both curves are virtually congruent. At pH 6.5, however, a 21 rhod-2 spectrofluorometry for measurement of [Ca ] in marked rightward shift was apparent, again indicating a i isolated rabbit hearts [5]. Then, the calculation was repeated, but now the adjusted K for a given pH (derived d from the NMR experiments) and temperature (according to the experimental protocol) was used for each time point.

.5. Statistics
Linear and nonlinear regression analysis was utilized to determine the best fitting models for describing the relationship between pH and K at a given temperature. The d Levenberg-Marquardt method was applied to derive non-2 linear models. The coefficient of determination (R ) was used to assess the proportion variability in K accounted d for by pH in each regression model with goodness-of-fit evaluated by the F-test. SPSS statistical package (version 11.0) was used.

.1. Intracellular pH
As depicted in Fig. 1, pH decreased during ischemia. The extent of intracellular acidification depended largely acidification, reaching pH 6.8 after 45 min. Irrespective of 21 the relationship between K (Ca Rhod-2) and pH (Fig. 4)  during the ischemic period. On reperfusion, there is a rapid fluorescence signal increased only by approximately 10%. and pronounced increase in fluorescence, followed by a On reperfusion, there was again a rapid increase in signal 21 rather rapid decrease over the next 5 min. When [Ca ] is intensity, followed by a decline over the next 5 min. When i 21 calculated after determining F , using a constant K [Ca ] was then calculated using a fixed K of 710 nM, max d i d 21 21 (here: 710 nM) (Fig. 5B), mean [Ca ] appears to the plot depicting mean [Ca ] (Fig. 7B) follows the raw i i increase only mildly during ischemia, followed by an fluorescence signal in parallel. However, when the pHimmediate increase on early reperfusion. However, when adjusted K is used in to determine the actual cytosolic d 21 21 21 the pH-adjusted K is used to quantify [Ca ] at each time Ca concentration, it becomes clear that [Ca ] increases d i i 21 point (Fig. 5C), it becomes evident that [Ca ] rises by by approximately 80% during ischemia (78619% in four i almost 400% throughout the period of ischemia experiments), and decreases immediately on reperfusion (431637% in four experiments), followed by a rapid (Fig. 7C). decrease during early reperfusion, without reaching preischemic levels during the first 10 min of reperfusion. 21 Beat-to-beat Ca transients during 15 min unprotected 4 . Discussion 21 ischemia are shown in Fig. 6. As mean [Ca ] increases, i 21 the amplitude of the Ca transient decreases. After 20 As expected, based on the physicochemical properties of 21 21 min ischemia, no rhythmic Ca transients were record-a Ca indicator dye such as rhod-2, we found the 21 able. In order to mimic the cardiac surgical situation of dissociation constant K for rhod-2 and Ca to be d cardioplegic arrest at lowered temperature, we recorded dependent on pH and temperature. At lower pH, ac- 21 21 cytosolic Ca in rabbit hearts subjected to 20 min cumulating protons compete with Ca ions at the binding 21 ischemia at room temperature with high-potassium cardiac sites resulting in a net decrease of Ca affinity of the dye, arrest. As shown in Fig. 7A, injection of the cardioplegic while lower temperature decreases the quantum efficiency 21 solution effectively abolished cytosolic Ca transients. of the dye, i.e. blunting the fluorescence intensity emitted 21 Over the following period of ischemia at 20 8C, the rhod-2 when a Ca ion binds to rhod-2. Both mechanisms  constant for rhod-2 and calcium in an intracellular environment, 710 nM. Note that [Ca ] does not seem to rise during ischemia, but a large increase in i 21 [Ca ] appears to occur during early reperfusion. (C) Here, the cytosolic calcium concentration was computed using the pH-adjusted dissociation constant i for rhod-2 and calcium at 37 8C. Note that cytosolic calcium increases mildly but steadily throughout the period of ischemia, and decreases during early reperfusion. 21 inevitably increase the effective K for rhod-2 and Ca .
reperfusion is caused by the rapidly normalizing intracellud lar pH (and thus decreasing K Various techniques to assess the [Ca ] have been estimated if K is not adjusted according to the changes in i d developed. NMR measurements using BAPTA-based pH or temperature. The pathophysiologic relevance is that, 21 21 Ca -sensitive indicators first allowed for observation of in contrast to previous reports based on Ca recordings 21 21 changes in [Ca ] over time [6], and beat-to-beat recordquantified using a static K , [Ca ]  the period of ischemia and not during early reperfusion. ings of Ca transients became possible using fluorescent Based on our findings, the increase in signal intensity on or luminescent indicators. Measurements of intracellular 21 Ca handling during whole-organ ischemia and reperfu-properties of the protein aequorin are of course not sion have been reported since the late 1980s using fluores-comparable with those of the BAPTA-based fluorescence cent indicators such as Indo-1, Fura-2, Fluo-3, or the indicators, and a systematic investigation of pH and visible light-emitting protein aequorin. Although we fo-temperature dependency of the dissociation constant for 21 cused on rhod-2, it is important to acknowledge that the aequorin and Ca has not been reported. However, it can 21 principles of pH-and temperature depending K apply to be assumed that the affinity of aequorin for Ca is also d all ion-sensitive indicator molecules. Following Lattanzio pH-sensitive, and that its quantum efficiency is a function 21 and Pressman's recordings of intracellular Ca transients of the ambient temperature. If this is indeed the case, one 21 using the first-generation indicator Quin-2 [7,8], Lee et al.
can infer that the true increase of [Ca ] in these studies i 21 established the measurement of [Ca ] in intact rabbit was also systematically underestimated, while the apparent i 21 hearts using Indo-1 [9]. They described a rapid increase of rapid [Ca ] increase on reperfusion is in fact a function i 21 systolic and diastolic [Ca ] during the first 30 s of global of the normalizing intracellular pH. i normothermic ischemia that reached a plateau after 90 s. In the majority of the studies investigating post-ischemic 21 Subsequently, the same group described the changes in Ca homeostasis, unprotected normothermic ischemia is