INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

INTESTINAL TISSUE ACID-BASE balance has been investigated by many as a detector of inadequate or "dysoxic" intestinal perfusion (8). However, interpretation of tissue acid-base balance remains controversial (31, 39). Early proponents (8) proposed calculation of intestinal tissue pH from measured intestinal PCO₂ and arterial plasma HCO₃. Our laboratory (23, 26) proposed interpreting just the intestinal PCO₂. The idea was that critical or "maximum respiratory" PCO₂ should be that which would occur if all the O₂ in a given arterial blood sample were replaced with CO₂, thereby simulating maximum O₂ extraction and CO₂ stagnation in the absence of strong acid. Larger than maximum respiratory PCO₂ would indicate tissue HCO₃ degradation to CO₂ by strong or "metabolic" acid. However, our prediction of critical PCO₂ turned out to be an underestimate when later applied prospectively (22). The problem, we suspected, had to do with countercurrent CO₂ exchange in vivo, which is difficult to predict (33).

The intent of the current investigation was to consider tissue acid-base balance as a potential detector of dysoxia from the perspective of the respiratory vs. metabolic distinction conventionally employed during the past 50 years in blood (9, 27, 35-38). This understanding is consistent both conceptually and quantitatively. Although several different sets of acid-base nomenclature are used according to preference, all are based on the same fundamental reasoning (24, 27) and all yield practically identical conclusions when properly translated (28, 29).

In blood, H⁺ is the acid-base variable regulated by biological processes, but it is a dependent variable physicochemically. H⁺ homeostasis of an organism is therefore produced by one or more of the three independent physicochemical variables: PCO₂ (respiratory), strong ion difference (SID; metabolic), and total noncarbonic weak acid buffer (A_tot). CO₂ is an acid because it combines with H₂O to form H⁺ and HCO₃. It is a "weak" acid because the net reaction is reversible, with its final equilibrium dependent on prevailing H⁺ availability. The second independent variable, SID, is the net charge of strong (38) or "fixed" (37) ions that do not bond with any other regardless of pH. SID is typically Na⁺ + K⁺  Cl  lactate (La). The third variable, A_tot (nonvolatile or noncarbonic weak acid buffer), is also a weak acid. A_tot is mainly histidine residues on protein (e.g., hemoglobin, albumin). Like CO₂, A_tot can donate or accept H⁺ depending on prevailing H⁺ availability. Unlike CO₂, however, A_tot does not change with lung ventilation or perfusion so that it typically does not change quickly enough in physiological systems to produce substantive acid-base changes (24).

In blood, the distinction between respiratory and metabolic disturbances can be made using the buffer base principle (37). Buffer base is the sum of the two buffers that account for virtually all of H⁺ buffered, HCO₃, and anion (A). Buffer base is a physicochemical "mirror image" of SID (metabolic). The sum of HCO₃ and A is negative and must equal the net charge of the strong ions (SID), which is positive, to preserve electroneutrality. HCO₃ is estimated from pH and PCO₂, and A is estimated from the same pH and an estimate of A_tot. Buffer base (like SID) stays constant during respiratory disturbances because the H⁺ coming from carbonic acid cannot be buffered by its by-product, HCO₃, so that the decrease in A equals the increase in HCO₃. During metabolic acidosis, however, both HCO₃ and A buffer the H⁺, and buffer base (like SID) decreases.

During progressively decreasing perfusion in tissue, the acidosis is initially respiratory. The CO₂ produced by aerobic metabolism stagnates in the decreased blood flow, causing PCO₂ to increase and pH to decrease. Buffer base (or SID) remains constant because no appreciable strong acid is being produced. As flow decreases to critical, however, lactic acid is produced by anaerobic metabolism, causing buffer base (or SID) to decrease.

The goal of the current investigation was to examine how respiratory (aerobic) and metabolic (anaerobic) acid-base disturbances might be distinguished in tissues by applying this common acid-base scheme. We packed freshly harvested intestine of anesthetized swine into steel pipes and sequentially measured PCO₂ and lactate concentration. The behavior of the resulting PCO₂ vs. lactate relations, analyzed from the perspective of common acid-base chemistry, provided estimates of the concentration and lumped negative logarithm of the equilibrium constant (pK_a) value of intestinal tissue A_tot. Application of this scheme to distinguish respiratory from metabolic acid-base balance in tissue would require that both tissue pH and PCO₂ be measured.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Surgical preparation. In a protocol approved by the Institutional Animal Care and Use Committee, seven domestic juvenile swine weighing ~20 kg each were fasted for 24 h, sedated with 10 mg/kg im ketamine, and anesthetized with 30 mg/kg iv pentobarbital sodium injected through an ear vein. Tracheas were cannulated for mechanical lung ventilation, and lungs were ventilated at a tidal volume, rate, and inspired O₂ fraction sufficient to maintain arterial PO₂ at ~150 Torr and arterial PCO₂ at ~40 Torr. The right internal jugular vein was cannulated for continuous infusion of pentobarbital sodium at 2-4 mg · kg¹ · h¹, and a catheter was advanced into the carotid artery to monitor arterial blood pressure and to sample arterial blood gases. All animals were volume loaded with lactated Ringer sufficient to maintain systolic arterial blood pressure >110 mmHg and were given 250 ml of 10% dextrose 30 min before data collection so as to maximize substrate available for anaerobic glycolysis. The intestines were exposed via a midline laparotomy.

Steel pipe. A 3-m-long steel pipe (Fig. 1), internal diameter = 1.3 cm and external diameter = 1.7 cm, which was threaded at one end, was used to store intestinal tissue segments anaerobically during ischemic measurements. A steel cap was screwed securely onto the threaded end of the pipe. A tight-fitting hole was drilled into the steel cap for the PCO₂ electrode. When inserted into this hole for measurements, the PCO₂ electrode protruded ~2 cm into the lumen of the pipe. Intestines were pulled into this pipe with a 4-m length of umbilical tape that passed through this same hole. To seal the opposite end of the pipe anaerobically and also to advance tissue segments for biopsy, a 0.5-cm-diameter steel "ramrod" with a tight-fitting washer at its tip was used.

View larger version (39K): [in this window] [in a new window]   Fig. 1.   Depiction of the experimental apparatus used.

Experimental protocol. After the laparatomy was completed, a 7-cm segment of small intestine was isolated by tying umbilical tape strictures at both ends. Mucosal PCO₂ was then measured between these strictures through a small antimesenteric incision. The arterial and venous mesenteric blood vessels supplying the segment were then quickly ligated, and the intestinal segment was removed, cut longitudinally, and frozen in liquid N₂. About 45 s elapsed from the time that mucosal PCO₂ was measured until the segment was vascularly isolated and placed in liquid N₂. The ends of the harvested segments adjacent to the umbilical tape strictures were amputated before freezing to isolate perfused from unperfused tissue. This procedure was performed three to four times for each animal.

Measurements and calculations. PCO₂ electrodes (Microelectrodes, Londonderry, NH) and PCO₂ meters (Orion) were calibrated in 40 mM NaHCO₃ baths (40°C) as previously described (22) to establish linearity. Bath fluid (40°C) was injected into our Radiometer ABL 330 analyzer where PCO₂ was measured at 37°C. We interpreted PCO₂ electrode measurements from this 37°C frame of reference. Because the sensitivity of these electrodes tended to drift over time, two-point calibrations were performed before each tissue PCO₂ measurement, and this two-point calibration was used to correct measured PCO₂ by interpolation or extrapolation.

Data analysis. We assumed that changes in SID were equal to changes in La because La is a strong anion and because concentrations of other strong ions must have remained constant. Hence, increases in La were taken as decreases in SID, the independent (metabolic) acid-base variable.

(1)

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Fundamental acid-base physicochemistry in blood. Top: fraction of negatively charged buffer as a function of pH for 3 different buffers with pKa values of 5.1, 6.1, and 7.1. pKa is the pH value at which one-half of the buffer is ionized. Middle and bottom: PCO2 vs. strong ion difference (SID) relations for 3 different closed systems containing 1 of the noncarbonate buffers shown at top in addition to HCO3. If the pKa of noncarbonate buffer equals 6.1, i.e., the same pKa as carbonic acid, the resulting PCO2 vs. SID relation is linear. If the pKa of the noncarbonate buffer is greater than or less than 6.1, the resulting relations are convex downward or upward, respectively. Middle: PCO2 vs. SID relations for a fluid with a total noncarbonate buffer (Atot) concentration of 20 mM and total CO2 (CO2tot) concentration of 25 mM, typical values for plasma. In bottom, Atot = 105 mM and CO2tot = 13 mM, the values for erythrocyte fluid (29, 35). BB, buffer base.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Slope of PCO2 vs. SID relations as a function of the CO2tot fraction of total buffer for any closed system in which the pKa of the noncarbonate buffer equals that of carbonic acid, i.e., 6.1 pH units.

Statistical considerations. The increase in PCO₂ and the increase in tissue La with time were compared using Pearson's correlation, and significance was taken as P < 0.05. The relations between PCO₂ and intestinal tissue La were analyzed by linear regression.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Initial and final values of PCO₂, La, and glucose are displayed in Table 1. La increased by ~30 mM, whereas PCO₂ increased by ~550 Torr, meaning that HCO₃ decreased by ~0.0306 × 550 or 17 mM. Figure 4, top, shows the increases in PCO₂ and La with time for the group. Dissolved CO₂ (in mM) increased at about one-half the rate that La increased, indicating that about one-half of lactate's H⁺ was taken up by HCO₃ and one-half by A. Our data fit the equation PCO₂ = 597  514^t/32.4. Figure 4, bottom, expresses the same data in terms anaerobic CO₂ production rate (CO₂) (right axis) and in terms of HCO₃ decay rate (left axis). The initial rate of anaerobic CO₂ was ~0.5 mmol · kg¹ · min¹. This compares with a normal aerobic intestinal CO₂ of ~0.8 mmol/min, calculated from intestinal O₂ consumption of 20 ml · kg¹ · min¹ (19, 20) and respiratory quotient of 0.85 (42). HCO₃ decay was consistent with first-order kinetics with a half time of 22.4 min and rate constant of 0.031 min¹.

View this table: [in this window] [in a new window]   Table 1.   Initial values, final values, and total change in values for PCO2, La, and glucose

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Top: increase in dissolved CO2 () and in tissue La () over time for a group of 6 animals. Values represent means ± SE. Brackets indicate concentration. Bottom expresses the same data in terms anaerobic CO2 production rate (CO2; right) and in terms of HCO3 decay rate (left).

Figure 5A shows the relation between PCO₂ and tissue La. Each of these relations was approximately linear, with the exception of subject 3, for which PCO₂ decreased progressively during the final five data collections. We believe these data represented loss of CO₂ gas from the steel pipe and thus did not include them in the analysis. Average r² was 0.94 ± 0. Because the PCO₂ vs. La relations were linear, the pK_a value for noncarbonate buffer (A_tot) should have been near 6.1, as per our methods displayed in Fig. 2, middle and bottom. Average slope of our relations was 16.7 ± 1.6 Torr/mM change in tissue La, suggesting a CO_2tot-to-total buffer ratio of ~50%, as per our methods displayed in Fig. 3. Hence, CO_2tot concentration approximated A_tot concentration. We did not measure CO_2tot but would estimate it roughly at 30 mM, the quantity of HCO₃ degraded in the subject with the largest change in La concentration (subject 2).

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Relation between intestinal tissue PCO2 and La (A) and between intestinal tissue La and glucose (B). The  values were included in statistical analysis; + values were excluded from statistical analysis.

Figure 5B shows the relation between tissue La and tissue glucose. These relations were also linear (average r² = 0.91 ± 0, average slope = 2.75 ± 0.2 mmol tissue La/mmol tissue glucose). This finding of more La produced than glucose catabolized suggests perhaps some pyruvate production by alanine transaminase, i.e., -keto acid + alanine  -amino acid + pyruvate. Average y- and x-intercepts were 44 ± 4 and 16.2 ± 2 mM.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Care of critically ill patients might advance if dysoxia, i.e., insufficient O₂ to support O₂ demand, could be detected in organs other than the beating heart or conscious brain. It would then be possible to reevaluate resuscitation and life support therapies currently in use. Here, we address the question of how tissue acid-base balance might be used to detect dysoxia, suggesting application of an old conceptual framework. Respiratory imbalance means abnormal CO₂ efflux from the system and is quantified as the difference between normal and measured PCO₂. Metabolic imbalance means abnormal accumulation of strong acid or base. Metabolic imbalance can be quantified by measuring pH and PCO₂ if A_tot and K_a are known (9, 27, 29, 35-38). Our findings suggest an intestinal tissue pK_a near 6.1 and A_tot of roughly 30 mmol/l.

What is the intestinal tissue value for K_a? Our method for estimating intestinal tissue K_a from PCO₂ vs. La relations (Eq. 1 and Fig. 3) is new. It is a simplification of an approach earlier used by Katsura et al. (15) to draw conclusions about the buffering characteristics of brain tissue. They had observed linear PCO₂ vs. La relations in anoxic brain and proposed several arbitrary pK_a values for brain proteins to simulate their data (15). Here, we have assumed that tissue protein behaves as if there were a single K_a value. We based this assumption on the fact that this simplification is entirely satisfactory in blood. For example, Figge et al. (9) estimated the K_a value for each individual histidine residue on the albumin molecule and developed a complex buffer base or "estimated SID" equation that used a different K_a value for each residue. However, their equation proved reducible to a simple linear equation that assumed a common K_a value for albumin, sacrificing no accuracy in this process (24). Their equation further proved to be no improvement over Siggaard-Andersen's original base excess equations, which had assumed a single K_a value for plasma of all individuals, regardless of plasma albumin or phosphate concentration (29). We therefore submit that buffering in any given fluid compartment (e.g., plasma, hemoglobin, intestinal tissue, brain tissue) does not differ sufficiently among individuals to require direct measurement of buffer concentration. Only pH and PCO₂ are needed if the typical buffer concentration and its pK_a is known.

What is the intestinal tissue value for A_tot? The slopes (Fig. 3) of the PCO₂ vs. La relations (Fig. 5) indicated that A_tot was approximately equal to CO_2tot. Because we didn't measure CO_2tot, our rough estimate of it was based on HCO₃ degraded in the subject with the largest La accumulation (subject 2, Fig. 5), i.e., 30 mmol/kg. Our methods are more easily applied to the data of Katsura et al., who measured CO_2tot directly in their preparations (15). CO_2tot of their hypocapnic preparations was ~13 mmol/kg, and their slope for brain PCO₂ vs. La was ~ 6.2 Torr/mM, giving a CO_2tot fraction of total buffer equal to 19% (Fig. 3) and an A_tot concentration of ~55 mmol/l. In their normocapnic preparations, CO_2tot was ~18 mmol/l and slope of PCO₂ vs. La was ~7.5 Torr/mM, giving an A_tot concentration of 60 mmol/l. These estimates compare with their A_tot estimate of 58 mmol/l, which they had based on best fit of five arbitrary protein K_a values and concentrations.

What is the underlying acid-base disturbance of dysoxia? Here we have assumed that the primary acid-base disturbance of dysoxia is a lactic acidosis. This assumption is at odds with the current belief (10, 11, 14) that lactate does not acidify because it is generated without a proton. This belief is based on chemical balance equations predicting that anaerobic glycolysis should produce no protons per glucose catabolized. The idea is that H⁺ is added to the system during anaerobiosis only if ATP is hydrolyzed irreversibly. Our data do not address this controversy directly. However, intestinal ATP is only ~2 mM in muscularis (3, 13) and ~2 mM in homogenized whole intestine (16, 18). In our investigation, 17 mM of tissue HCO₃ decayed to CO₂, requiring at least 17 mM of H⁺. This 17 mM of H⁺ could not have come from 2 mM of ATP. In fact, even more than 17 mM of H⁺ must have been produced to account for the HCO₃ decomposed because tissue, like blood, must contain substantial quantities of protein with histidine residues, which also buffer H⁺. We thus submit that lactic acid is the chief anaerobic product.

2

2

2

Why were the PCO₂ values so large? In our gas-tight chamber, intestinal PCO₂ approached atmospheric pressure within ~2 h (Fig. 5). These PCO₂ values may seem quite large but are similar to those originally reported by Bass et al. (2) who measured PCO₂ with a mass spectrometer. In most subsequent investigations, intestinal tissue PCO₂ has been measured with silastic balloon tonometers (8, 26, 40), which have given lower estimates. For example, when our laboratory (26) and Walley et al. (40) used balloon tonometers in small intestine during progressive flow reduction, critical and maximum intestinal PCO₂ values were only ~60 and ~150 Torr, respectively. However, use of PCO₂ electrodes (22) in similar preparations yielded much larger critical and maximum PCO₂ values of ~100 and ~380 Torr, respectively. Tonometers occupy most of the unstressed volume of the small intestinal lumen in medium-sized animals, doubling the space into which CO₂ gas can diffuse and thus diluting CO₂ gas arising from anaerobic metabolism.

What was the range of tissue pH in our preparations? Without having measured pH, we can provide only a rough approximation, using the Henderson-Hasselbalch equation, of our measured PCO₂ values (Table 1) and a guess of initial intestinal HCO₃ in our preparations. If we assume that initial HCO₃ concentration was approximately equal to HCO₃ degraded in the subject with the largest La accumulation (subject 2, Fig. 5), i.e., 30 mmol/l, then initial pH would be 6.1 + log 30/(0.0306 × 84) or 7.17 pH units. Final HCO₃ would be 30 mmol/l minus the average change in HCO₃, 17 mmol/kg, and final pH would be 6.1 + log 14/(0.0306 × 640) or 5.95 pH units. Because we studied tissue, which contains only a small proportion of blood and interstitial fluid, these results would represent average intracellular pH estimates.

How do our findings elucidate intestinal tissue acid-base balance as a potential detector of dysoxia? Our measurements indicated that CO₂ generation and lactate accumulation proceeded at the same rate, consistent with first-order kinetics (Fig. 4). The half time of both their accumulations was ~22 min. The initial rate of anaerobic CO₂ accumulation was substantively less than normal aerobic CO₂, but PCO₂ was much larger than normal because no blood was flowing to carry the CO₂ away. One implication of these findings is that the fundamental acid-base change of intestinal dysoxia is lactate accumulation. Another implication is that flow must be extremely low, probably near zero, for PCO₂ to increase to the large values that we (22) and others (2) have observed in vivo. If dysoxic tissue were to become reperfused, even transiently, the CO₂ ought to be flushed from the tissue quickly. Hence, a large tissue PCO₂ value, e.g., >130 Torr (22), ought to be highly specific for dysoxia but possibly insensitive.