For my own part, I have never before engaged in any study that so totally engrossed my attention and my time as this has lately done; for what with making experiments when I can be alone, and repeating them to my Friends and Acquaintance, who, from the novelty of the thing, come continually in crouds [sic] to see them, I have, during some months past, had little leisure for anything else...

Benjamin Franklin

Letter to Peter Collinson on Electricity

March 28, 1747

6.1 Introduction

In order to demonstrate the simultaneous collection of optical and NMR spectra as a useful tool, we employed the instrumentation designed in this research to simultaneously collect optical fluorescence spectra and ³¹P NMR spectra from the in vivo, in situ heart ventricle of the western painted turtle Chysemys picta bellii during normoxia, anoxia, and subsequent normoxic recovery. As described in Chapter I, this animal is particularly tolerant to prolonged anoxia, and while comparative physiologists studying this phenomenon have used NMR spectroscopy and fluorescence spectroscopy separately, the tools have yet to be combined into a single experiment. Further, these data have never been collected (simultaneously or otherwise) from the in situ heart of a living animal.

This section presents the results of several in vitro and preliminary in vivo studies employed to verify and qualify the instrumentation and preparatory techniques developed for this study. These preliminary results are described first and are followed by the results of in vivo dual spectroscopy studies demonstrating combined NMR and optical spectroscopy. The results of some additional preliminary in vivo NMR experiments are also included in APPENDIX F.

6.2 Preliminary NMR Results

Recall that in Section 3.5 a double-tuned, inductively-coupled solenoidal resonator was developed to allow for ¹ H shimming and ³¹P NMR acquisition over the same sample volume. In order to assess the performance of the doubly-tuned resonator as a sample coil, experiments were performed on an in vitro phosphorus phantom. The phantom consisted of a small 1 mL volume of aqueous solution containing 1M NaH₂ PO₄ and .075 M Na₄ P₂O₇ . This phantom was placed inside of the doubly-tuned resonator and shimming was performed on the phosphorus signal directly. Spectra were collected with 256 acquisitions, a 90° tip angle pulse, and a ± 3000 Hz SW. The SNR of these spectra were calculated by ratioing peak height against the RMS noise value for real spectra. The results were used to compare identical experiments in which a single-tuned solenoid coil identical to the sample coil of the resonator and a commercial GE 2 cm surface coil were used. The results of the experiments are presented in Table 6.1 and demonstrate that no loss in SNR is experienced with the doubly-resonated coil. The unloaded Q of the coil was estimated by tuning the coil in the S ₁₁ mode on an HP4195 network analyzer and noting the points where the reflectance magnitude |G| was ½.

Table 6.1: Assessment of Double-Tuned Resonator.

6.3 Preliminary Optical Fluorescence Results

Two sets of preliminary experiments were performed to evaluate the performance of optical calcium measurement. The first was designed to verify performance of the CCD-based optical fluorescence spectrometer. The second was designed to verify the efficacy of the pericardial fluo-3/AM loading technique developed as part of this work.

6.3.1 In vitro fluo-3 Results

A fluo-3 calibration experiment was performed inside of a high-field magnet in order to verify operation of the CCD based instrument and to ensure that the fluo-3 fluorescence resonance was not significantly altered by the magnetic field. Such effects can occur when electron spins are preferentially aligned by a polarizing field and can cause shifts in observed optical resonance frequencies [94]. A 500 mM fluo-3 stock solution was prepared and 100 mL of this solution was alternately added to 2.5 mL of precision calcium calibration standards to yield a net fluo-3 concentration of approximately 20 mM. Samples contained concentrations of 0 mM, 0.15 mM, 0.35 mM, 1.35 mM, and 39.8 mM, of free ionic Ca²⁺. Samples were placed into test tubes with white paint coatings on the outside to enhance optical reflection back into the fiber probe. The test tubes were alternately placed into the bore of a 2.05 Tesla resistive magnet and optical fluorescence spectra were recorded when an excitation wavelength of 505 nm was used. The experimental setup is diagramed in Figure 6.1. Two spectra for each calibration standard were taken and averaged together in Matlab. In order to calibrate CCD pixels to wavelength values, the excitation source was varied from 490 nm to 650 nm in 10 nm increments. The pixel value location of the resulting reflection peak was used in a linear regression model to correlate pixel numbers to the wavelength values which are reported in Figure 6.2.

Spectra were saved and analyzed off-line in Matlab. Baseline corrections were performed based on the observed intensity of the excitation peak in order to account for differences in path lengths and reflectivities of the samples. Figure 6.2 shows the resulting fluo-3 spectra. The curve peaks were integrated between 520 nm and 530 nm and were used as raw fluorescence values in the equation below:

Since the [Ca²⁺] concentration was known in each case, the resulting fluorescence intensities could be used to extract the value of K_D for the experiment. The quantity [F-F_min]/[F_max -F] was computed for each sample (excluding the 0 mM minimum and 39.8 mM maximum samples) and used along with the known sample concentrations to calculate values for K_D . The average computed value of K_D was 325 nM which is in good agreement with the published value of 320 nM in the product literature [95], and hence we concluded that the presence of a large magnetic field had no significant bearing on the performance of fluo-3 as a calcium measurement agent.

6.3.2 Pericardial Loading of fluo-3/AM

As was described in ChapterV, the dye fluo-3/AM was loaded into the ventricular cardiomyocytes of the heart via a pericardial bathing technique. In Chapter IV, it was explained that this dye, when excited with light at ~500 nm, emits a broad 525 nm optical fluorescence signal whose intensity increases as the fraction of the dye bound to free calcium increases. Thus, from the returned intensity signal of the dye, the intracellular calcium concentration in the cells of the ventricle can be determined.

Preliminary experiments were performed to assess the effectiveness of fluo-3/AM in vivo and to see if fluo-3/AM could effectively be loaded into the specific cells under investigation. The heart of an anesthetized turtle was carefully removed (so as not to disrupt the pacing center located in the sinus venosus) and placed immediately into a test tube filled with TES Ringer's solution. The heart continued to beat and was oxygenated by bubbling room air through the bath solution at ten minute intervals. A fluo-3/AM TES Ringer's solution was prepared according to the procedure described in Chapter V and the heart was placed into this solution. Oxygenation was continued as above for an additional two hours. At the end of this loading period, the heart was removed from the fluo-3 bath, washed twice in fresh Ringer's solution and placed into a 1 cm cuvette filled with fresh TES Ringer's solution. This cuvette was placed into a SPEX fluorolog-2 fluorescence spectrometer with a DM1B spectroscopy computer. Arc-lamp excitation at 500 nm was used and the fluorescence emission at 525 nm was monitored at 100 msec intervals (the fastest the machine would allow) for 200 seconds. During this time, the heart was not oxygenated and scans were taken every 15 minutes. Figure 6.3 shows the results of several scans concatenated together. These scans indicate fluorescence transients which correlated exactly with cardiac excitation and are as expected for a typical heart. As time wore on and hypoxia began to set in, a decrease in heart rate along with an apparent increase in baseline fluorescence (and hence resting calcium level) was seen. These data are as expected and indicate that fluo-3 could successfully be loaded into in vivo cardiac cells through a TES Ringer's based bath.

It is important to mention the fact that these experiments were performed on isolated hearts contained in a relatively small volume of bath solution and were void of perfusion. The result is that significant acidosis (due to accumulated lactate produced by anaerobic metabolism) occurs. This acidosis has been shown to cause large rises in intracellular calcium [18]. In the in vivo and in situ heart preparations used later, this acidosis and concomitant rise in intracellular calcium would not be expected, since the heart continues to be perfused and waste products like lactate can be cleared via the circulatory system.

Figure 6.3: Real-time In vitro Fluorescence Data.

6.4 Unidirectional Ventilation Experiments

In Chapter V, the details of a unidirectional ventilation technique developed for this work were presented. In order to investigate the efficacy of this unidirectional ventilation technique, two experiments were performed during which an animal so prepared was also fitted with a small occlusive catheter inserted into the right subclavian artery. Small (1-2 mL) arterial blood samples were taken approximately every 15 minutes during ventilation by the unidirectional method. These samples were analyzed on a Radiometer ABL-5 blood gas analyzer. Though this machine is designed for mammalian samples at 37° C and with mammalian electrolyte compositions, approximate temperature corrected values can be obtained which are reasonable estimates of the gas partial pressures for the turtle. The results of one such experiment are shown below in Table 6.2.a for breathing gas mixtures of 97%N₂ /3%CO₂ (anoxic) and 97%O₂/3%CO₂ (hyperoxic). These results indicate that the unidirectional ventilation method is effective and allows rapid equilibrium of blood gasses.

Table 6.2.b contains similarly collected results for a separate experiment where the ventilation conditions were 1) tidal breathing of room air through the tracheostomy tube, and 2) subsequent unidirectional breathing with a 95%O₂ / 5%CO₂ breathing gas mixture. These results indicate that some decrease in oxygen tension during nonventilated intervals does occur and that even upon subsequent tidal ventilation, recovery may not always be complete. They also indicate that unidirectional ventilation with a hyperoxic gas mixture does replenish blood oxygen levels.

Table 6.2.a: Blood Gas Measurements during Unidirectional Ventilation.

Table 6.2.b: Blood Gas Measurements during Tidal Ventilation.

The advantages of this unidirectional ventilation technique are two-fold. First, as can be seen in Table 6.2.b, tidal breathing does not establish a steady-state oxygenation level in the blood, and even upon subsequent ventilation with a short burst of room air, recovery is not complete. This is a problem in prolonged NMR experiments where data collection may take up to 20 minutes, since the metabolic state will not be constant over this time interval, and the animal may actually become anoxic during what is supposed to be a normoxic control period. The second advantage is that no respiratory gating is required and ventilation may be performed continuously during NMR acquisitions. We found that the lungs of the animal inflated bilaterally and uniformly to a fixed and constant volume roughly determined by the gas flow rate.

6.5 Dual Spectroscopy Experiments

Simultaneous collection of ³¹P NMR spectra was employed in order to monitor the energy state of the cells of the in vivo turtle ventricle. As was described in Chapter II, the relative areas of peaks in the phosphorus spectrum can be used to assess metabolic energy state. Of specific interest are the areas of the PCr peak and the b-ATP peak and the changes in these peaks during anoxia. Also, the position of the P_i peak relative to the PCr peak can be used to determine pH since it is well known that the P _i chemical shift is pH dependent [2, 3]. Examination of consecutively collected spectra allows for determination of the metabolic energy state as a function of time. A description of the experiments performed and the resulting information and interpretation follows.

6.6 Experimental Protocol

Animal selection and special preparatory techniques were described in detail in ChapterV and will only be briefly reviewed here.

6.6.1 Preparation

Animals were anesthetized by intramuscular injections of diazepam and ketamine and induction took approximately 20 minutes. The plastron (lower shell) of the turtle was removed to expose the heart and internal organs. A tracheostomy was performed superior to the bifurcation of the bronchi and the lingula of the left lung was fitted with a cannula for the unidirectional breathing technique employed. Next, the animal was connected to an oxygen rich (97%O₂ / 3%CO₂ ) gas mixture and ventilated unidirectionally with input through the tracheostomy tube and output through the cannula. At this time, the pericardial fluid was withdrawn from the intact pericardium and replaced with a fluo-3/AM in TES Ringer's solution as described in Chapter V. Pericardial bathing lasted for approximately two hours after which the pericardium was resected and the excess dye washed away with fluo-3-free TES Ringer's. At this time, the animal was placed into the plexiglass holder and NMR coil positioned. The heart was lifted into the coil and cemented in place. The animal was temporarily disconnected from the gas supply and the cannula and tracheostomy tube were sealed to keep the lungs inflated. The subject was moved to an HP 4195 network analyzer, and the ¹ H and ³¹P coil channels were each tuned in the S₁₁ mode. After coil tuning, the optical fiber was attached and positioned just over the apical portion of the ventricle. The animal was inserted into the 18 cm gradient set of an Oxford 2.0 T, 30 cm horizontal bore superconducting magnet. The oxygen rich mixture was reconnected to the tracheostomy input, and the ¹ H channel was connected to the NMR system.

6.6.2 NMR Acquisition

Using a 9 ms 90° ¹H RF pulse-width, standard, single acquisition, one-pulse shimming was performed until an optimal linewidth (usually about .5-.7 ppm) was obtained. Next, the NMR spectrometer input was changed to the ³¹P channel and phosphorus acquisition was initiated. A one-pulse experiment with a 90° pulse width of 25 ms, a repetition time of 2.3 seconds, spectral width of ± 3000 Hz, 4096 points, and 256 acquisitions took just under 10 minutes for a complete scan. Raw FID data were saved prior to any apodization or transformation for later off-line analysis. All acquisitions were performed on a TecMag Libra MR imaging/spectroscopy console using the MacNMR 5.3 interface. A GE broadband 100 Watt amplifier was used to generate hard pulses. The MacNMR pulse sequence is included in APPENDIX D.

6.6.3 Optical Fluorescence Acquisition

The instrument designed to collect optical spectra was described in Chapter IV and consisted of a TE-cooled CCD and fiber-optically coupled imaging spectrograph. During optical spectral collection the WinSpec version 1.4.3 software interface was used to control camera exposure and spectrograph movement. For all acquisitions, the spectrograph was set to 515 nm which translates to a spectral imaging region on the camera of approximately 470 nm to 570 nm. The image of the collection fiber was located on the CCD by first taking a full 512x512 image. The image was displayed on the screen and back reflected light was used to locate the extent of the fiber image on the CCD. The upper and lower pixel limits of the vertical extent of the fiber were used to set up the binning such that a composite 1 (vertical) by 512 (horizontal) spectrum was taken. In most cases the vertical extent of the fiber selected was approximately 100 pixels. With excitation illumination monochromator wavelength set to 500 nm, a total of three separate 20 second exposures were taken over the course of ten minutes (which corresponds to the length of a single NMR acquisition.) These exposures were saved to separate ASCII table files for later analysis in either Matlab or OCTAVE.

6.6.4 Anoxia Protocol

The goal of these experiments was to demonstrate simultaneous measurements of intracellular calcium and phosphorus energy state during anoxia by correlating results to expected results based on experiments on isolated hearts or ventricular cardiomyocytes. The anoxia protocol was similar to those used by other researchers [18, 20, 21, 32, 33]. Baseline values were established while the animal was on the oxygen rich breathing gas mixture (97%O₂ / 3%CO₂ ) and usually consisted of two to three NMR spectral acquisitions (about 30 minutes total). Next, anoxia was initiated by changing the breathing gas mixture to an anoxic one (97% N₂ / 3% CO₂). Anoxic conditions were typically maintained for between one and three hours. At the end of this time, the experiment was either terminated, or a recovery period was initiated during which the breathing gas mixture was switched back to the oxygen rich mixture. If the animal was reoxygenated, this recovery period was typically carried out for between 30 and 60 minutes. At the end of each experiment, the animal was euthanized by removing the heart.

6.7 Data Analysis

All of the experimental results shown here represent off-line processing and analysis of the raw data collected during experiments. Typically, data were exported as ASCII table files and then imported into either Matlab or OCTAVE for manipulation and analysis. Source code for program files (.m-files) specifically written for this analysis are included in APPENDIX E.

6.7.1 NMR Analysis

Analysis of NMR spectra was fairly straight forward. First, the ASCII FID data files were loaded (in Matlab using load -a, in OCTAVE using the loada.m program.) Next, the real and imaginary column data were converted into a complex FID, offset correction (mean subtraction) was performed, exponential apodization (line broadening) was applied, and the FIDs were transformed using a built-in fast Fourier transform (FFT) routine into complex NMR spectra. All of these steps were accomplished using the linbrd.m program. Next, first and second order phase correction was performed using the autophas.m program. This routine analytically determines a first and second order phase correction term based on maximization of the cumulative sum real-valued portion

of the spectrum. Finally, phased, real spectra were plotted either singly or together (using the waterfall.m command.) Spectra were plotted versus frequency shift in ppm (frequency in Hz divided by 34.635) and a horizontal offset was added so that the PCr peak was centered at zero ppm (reference).

The following figures are representative of the NMR data collected and the intermediate processing steps. Figure 6.4 is the real portion of a typical raw FID before baseline (average) subtraction. Figure 6.5 is the real portion of the same FID after baseline removal and apodization by a negative exponential of the form e^-15pt (15 Hz line broadening). After Fourier transformation, the spectrum appears as in Figure 6.6 which shows separately the real, imaginary, and magnitude spectra. Finally, Figure 6.7 shows the real portion of the spectrum after zero and first order phase correction terms have been added. Also shown are peak assignments for P_i, PDE, PCr, and g-, a-, and b-ATP. The fully analyzed results of some preliminary in vivo NMR experiments are included in APPENDIX F and demonstrate the response of the phosphorus energetic state in both anoxia and anoxia/recovery experiments.

6.7.2 Optical Spectra Analysis

When using fluorescent indicator dyes, the measurand of interest is not measured, but rather is quantified by observing the optical response of the dye and deducing quantitative information about the measurand from this response. In this case, the quantity of interest is the concentration of intracellular calcium, however, the quantity which is measured is the intensity of fluo-3 fluorescence (which increases with the concentration of free Ca²⁺.) One of the problems of this type of measurement is that the amount of the indicator solution present (and other factors such as pH and temperature) can effect the intensity of the fluorescence peak in addition to changes in the value of the measurand [8]. For this reason ratiometric dyes like fura-2 and indo-1 are attractive since a method of determining dye concentration is ``built in.'' However, in the case of dyes like fluo-3, this luxury does not exist and either calibration against known standards must be performed, or assumptions must be made about the dye which was loaded.

This does not mean, however, that reasonable conclusions about the concentration of intracellular calcium can not be made from single peak fluorescence data. Recall equation (4.1) which relates calcium concentration to fluorescence intensity

In order to successfully use this equation, the maximum (when the dye is saturated with calcium) and minimum (when no calcium is present) fluorescence intensities must be known as well as the binding constant K_D . Then, a particular observed fluorescence intensity may be quantitatively related to a particular calcium concentration. Since F _Max and F_Min will in general depend on the amount of dye present, direct correlation of an observed fluorescence intensity value to a calcium concentration is impossible without performing a calibration experiment.

However, since some information about the intracellular calcium concentration is known a priori, useful information about calcium concentrations can be derived without specific knowledge of F_Min or F_Max. For example, when F is significantly less than F_Max , and F_Min is zero, the above equation is nearly linear with F. In fact, if these conditions hold (which is reasonable to assume), we can even make a guess as to when the calibration curve will become significantly non-linear and render measurements without calibration invalid. Consider the case when F_Min is zero. (This is a reasonable assumption if instrument baseline correction is possible.) Equation (6.2) can be rewritten as

Further, if we normalize F_Max to one, we can write the pair of equations:

These equations and some a priori knowledge of our physiologic sample allow sensitivity analysis between fluorescence intensity and calcium concentration. For example, it was shown in Section 6.2.1 that the presence of a large magnetic field did not significantly alter the K_D for fluo-3 dye from the published value of 325 nM [95], and it is known that normoxic, diastolic intracellular calcium level in healthy cardiomyocytes is approximately 100 nM. We can use these values to establish an ``operating point'' on the normalized calibration curve. Thus, at normoxic levels, the observed fluorescence F should roughly correspond to (1+100/325) ^-1 = .235 of the maximum fluorescence intensity F_Max. Therefore, if an increase in fluorescence intensity of say 10% is observed, equation 6.4.b shows that the corresponding [Ca²⁺] concentration is [Ca ²⁺] = (325)(1.1)(.235)/(1 - (1.1)(.235)) [nM]= 113 nM, a 13% increase.

In order to measure resting calcium levels in the heart with the current system, the following strategy was employed. During a particular period of interest, three 20 second exposures were taken with the CCD array. These three scans were added together to yield what is essentially the average calcium concentration induced fluorescence spectrum over a period of one minute. If the heart rate is fairly low (this probably would not work with a rat at 250 bpm, but a turtle at ~20 bpm is reasonable) and if the peak height of the systolic calcium transient is relatively constant, then this average value should reflect a reasonable estimate of the diastolic intracellular calcium level. This can be illustrated by fluorimeter data taken on a SPEX fluorolog-2 fluorescence spectroscopy system of isolated hearts. Figure 6.8 is a fluorescence spectrum from an in vitro (but still living) fluo-3 loaded heart which was placed against the side of a methacrylate cuvette in a bath of manually aerated TES Ringer's solution. In order to smooth out variations due to calcium transients and demonstrate the fluo-3 fluorescence, a 10 second integration time was used for each wavelength point. Because large slit widths were used in order to get high intensity, the 525 nm fluorescence peak is superimposed on the shoulder of the 500 nm excitation peak.

Figure 6.9 shows a second scan taken with the fluorimeter on an isolated heart in a similar configuration. In this case, however, the integration time was reduced to 0.1 seconds (the fastest the machine would allow) so that calcium transients could be observed. The machine does not allow for stationary measurement, so the smallest wavelength increment possible (.01 nm) was used to scan the region from 523 nm to 527 nm. Note that the intensity of the fluorescence not only drops off because of the excitation shoulder, but that the height of the transient peak is scaled as well due to the fall off of fluo-3 fluorescence intensity with wavelength. In order to correct this distortion to some extent, the spectra were multiplied by a linear (as a function of wavelength) scaling factor to force the ending peak heights to be the same as the initial peak heights. The resulting spectra were then baseline fitted with a simple linear regression model and slope (but not offset) correction was applied. Figure 6.10 shows a group of spectra taken after different lengths of anoxia (no perfusion) which were acquired and corrected in this way.

Figure 6.11 shows the time averaged fluorescence intensity of each of the calcium transient signals from Figure 6.10. It is evident that this average value is a very good representation of the baseline resting fluorescence intensity. Therefore, it can be concluded that for the slowly beating turtle heart, a time averaged fluorescence intensity should be a good representation of the fluorescence intensity due to diastolic intracellular calcium levels.

It should be noted in all cases that the recorded signal intensity for a particular experiment on this fluorometer depends on a number of factors including excitation wavelength, excitation slit width, emission slit width, integration time, and sample placement. Therefore, the relative magnitudes of recorded fluorescence signal may vary from one particular experiment to another.

Fluorescence scans using the CCD system were collected as described in Section 6.6.3 and were analyzed as follows. First the files were loaded into OCTAVE using the loada.m command. Next the series of calibration scans were plotted versus pixel value and the pixel index of the maxima for each CCD image located. The wavelengths at which the calibration scans were taken were then regressed against the pixel values of the located maxima in order to form a relationship between pixel value and wavelength. In the subsequent figures, data are plotted against these fitted wavelength values.

Data scans were first cleaned with 11-point smoothing using the smooth.m command. Individual spectra were then baseline corrected by forcing them to coincide at a point far away from any spectral features (about 580 nm). The resulting spectra are shown in Figure 6.12. In order to gauge the calcium related fluo-3 fluorescence, the portion of each spectrum between 520 nm and 530 nm (fluo-3 has emission maximum at 525 nm) was integrated numerically. These integrated intensity values were then ratioed against the first intensity value in order to obtain fluorescence intensity as a percentage of baseline (initial) value. Figure 6.13 shows these calculated values for the spectra in Figure 6.12. It is apparent from this figure that calcium-dependent fluorescence levels dropped as much as 40% during normoxic breathing. This change is probably not significant since drops of up to 50% in diastolic intracellular calcium levels have been observed by others during normoxic control protocols [18]. One possible explanation is that during the preparation immediately prior to normoxic ventilation, the animal became slightly anoxic and then recovered. It is also evident that 25 minutes of anoxic ventilation had little or no effect on intracellular calcium which is in line with results observed by others [18].

In Figure 6.14, the spectra from Figure 6.12 have been mean-centered and replotted. In order to do this, an ``average'' spectrum is created by calculating the mean of all spectral data at each wavelength and subsequently subtracting this ``average'' spectrum from each individual spectrum. The result is a set of data which highlight the differences between spectra [90]. As can be seen in the figure, mean centering the data has the effect of removing the CCD anomaly at 540 nm and extracting the broad shape of the fluo-3 525 nm fluorescence peak.

6.8 Dual Spectroscopy Results

The following are the results of a dual spectroscopy experiment performed to demonstrate simultaneous collection of optical and NMR spectra for monitoring of intracellular calcium concentration and phosphate metabolism in vivo. Figure 6.15 is a stack plot of the ³¹ P NMR spectra obtained. Information from these spectra was used to calculate the relative levels of both PCr and b-ATP as a function of time. Figure 6.16 shows the calculated values as percent of the initial normoxic value. In this particular experiment no recovery period was used.

Figure 6.17 shows the processed fluorescence spectra associated with the corresponding NMR acquisitions. The 525 nm fluorescence intensity is used to calculate the relative fluorescence as a percentage of baseline. This plot is shown in Figure 6.18.

Figure 6.14: Mean-centered fluo-3 CCD Fluorescence Spectra.

6.9 Discussion of Dual Spectroscopy Results

The data from Figures 6.16 and 6.18 are shown together in Figure 6.19. From this plot it can be seen that as anoxia progresses the cellular energy state becomes depleted as evidenced by a decrease in PCr stores. This decrease takes place fairly rapidly in about 10-20 minutes and agrees well with results observed by other investigators of this species [21, 32, 33]. Concomitant with the depletion of PCr stores is a very slight rise in the diastolic intracellular calcium level. This is evidenced by the increased fluo-3 fluorescence intensity observed. However, this fluorescence rise is slight, and even after 100 minutes of anoxia only an 8% fluorescence rise is observed. This 8% increase would be expected to correspond to a rise in intracellular calcium concentration of 10 nM or so, an amount which is not physiologically significant [17, 18].

This ability to maintain [Ca²⁺] homeostasis even during prolonged anoxia and significant depletion of phosphate energy stores can be explained by examination of b-ATP levels. It is evident in Figure 6.19 that although PCr stores have become severely depleted, b-ATP levels remain relatively constant during the course of the anoxic assault. This same result has also been observed by other investigators, and it is postulated that the explanation lies in this organisms remarkable ability to switch to anaerobic metabolic pathways with few deleterious effects [21, 32, 33, 96]. The ability to maintain a steady supply of b-ATP facilitates continued regulation of diastolic calcium levels since the ATP required for operation of active Ca²⁺ -ATPase pumps is apparently still available even in the absence of oxidative phosphorylation.

It is important to note that these data come from an in vivo heart preparation where normal circulatory function is being maintained. The ability to clear the waste products (like lactate) associated with anaerobic metabolism is important for continued calcium regulation during anoxia. This is evidenced by work on this species in which isolated cardiomyocytes subjected to lactic acidosis and prolonged anoxia did exhibit significant rises in intracellular calcium levels [18], and in the results presented in Section 6.3.2 where a rise in intracellular calcium was apparent when the heart was not afforded perfusion. Thus the distinction between respiratory anoxia or hypoxia and ischemia is important in terms of its effects on calcium regulation.

Finally, while the results presented here are similar to those observed by other investigators, they are important for two reasons. First, they are derived from an in vivo and in situ heart preparation and, as such, can be used to verify and validate results observed in in vitro preparations involving isolated, hearts perfused with crystalloid buffer. Also, the suggestion by some researchers that circulating plasma factors may play a role in the anoxia tolerance and maintenance of calcium regulation in the brains of this same animal points to the importance of such in vivo studies [22, 97]. Secondly, the phosphorus metabolism and calcium regulation data have been obtained simultaneously. Therefore relationships extrapolated from these data do not suffer from subject or treatment variability. The importance of obtaining simultaneous data from a single experiment has been pointed out by Camacho et al. [16] and is practiced here for the first time in an in vivo animal model.