Classical Aversive Conditioning of Heart Rate in Curarized Rats at Different Blood Gas Levels CLAUDE J. GAEBELEIN, JAMES L. HOWARD, RICH~D A. GALOSY, ANY PAUL A. OBmST

Department of Psychiatry and the Biological Sciences Research Center of the Child Development Institute, University of North Carolina, . Chapel Hill, North Carolina

Abstract-In an effort to exa,ninc whether normal blood gas tensions were essential for conditioning, paralyzed rats received a classical aversive heart rate (HR) conditioning session while respirated at different peak expired CO.,_ values. After the session, arterial blood was drawn for analysis. That peak expired CO2 was effective in manipulating Pco., was indicated by a significant correlation (r=0.594, dr--17, P < 0.0.5). In addition, only rats with blood gas values similar to those of anesthetized controls displayed a discriminated HR CR. These animals also had lower baseline HRs and greater HR variability. Further, 7 of the 9 rats with normal blood gas values were respirated at peak expired CO2 values from 5.0-5.1 per cent, and no animal ventilated within this range displayed abnormal values. These findings suggest that previous difficulties in obtaining classical and operant conditioning in paralyzed animals may, in part, be attributable to inadequate ventilation.

OPTIMISTIC REPOIITS Of instrumcntally-acquired changes in visceral responses in curarized rats (DiCara, 1970; Miller, 1969) have been replaced by cautions following recent reports of unsuccessful attempts to replicate this phenomenon (Miller & Dworkin, 1974). In addition, investigations of the efficacy of Pavlovian techniques in the paralyzed organism have reported mixed results. For example, although Black, Carlson and Solomon (1962) reported being able to elaborate a discriminative CR in dogs immobilized with d-tubocurarine chloride, some animals failed to condition. In eats, Howard This investigation was supported in part by research grant MH-07995 NIH and Institutional Grant HD-03110 NICND, USPHS.

76

Volume 11 Number 2

LEARNING AT VABIOUS BLOOD GAS LEVF-J~

77

and Obrist (1971) found that d-tubocurarine abolished a CR previously established in the noncurarized state. Ray (1972) confirmed this in rats and further demonstrated that curarization inhibited later conditioning in the non-paralyzed state (Ray & Brener, 1973). On the other hand, CRs have been successfully elaborated in dogs (Black, et al., 1962), cats (Hein, 1969; Howard, et al. 1975) and rats (Hahn & Slaughter, 1970; Roberts, Wright & Saletta, 1974). One explanation for these diverse findings may be related to procedures used to maintain the curarized animal. One important aspect of such maintenance concerns proper respiration of the paralyzed animal, particularly since subsequent work has suggested that the respiratory parameters typically used in studies of operant learning may have resulted in aphysiological respiratory levels (Brener, Eissenberg, & Middaugh, 1974; Hahn, 1972). Adequate ventilation may be assured by adjusting respiratory parameters to maintain blood gas tensions of Pco.~ and Po._,, as well as pH, at normal levels. However, in small animals such as the rat, withdrawal of multiple blood samples may not be desirable because of deleterious cardiovascular effects, e.g., reduction of blood volume. In a previous report, a respiratory system was described which not only seemed to preclude ,multiple blood gas determinations, but incorporated other improvements over earlier systems (e.g., the use of an endotrachcal tube) (Gaebelcin & Howard, 1974). With this system, end-tidal peak expired CO.. is monitored continually, and is maintained at a desired level by alterations in rate and pressure of respiration. In the present paper, the relationship among expired CO., and blood gas values was further studied. In addition, a Pavlovian, discriminative, heart rate (HR) conditioning session was conducted to provide an index of the rats' sensitivity to exteroceptive stimuli at different blood gas levels. Finally, criteria used by others as indicative of adequate ventilation, specifically in terms of baseline HR (Miller & DiCara, 1967) and HR variability (Brener, et al. 1974), were examined with respect to blood gas levels. Method Subjects Twenty male Sprague-Dawley rats, 60-120 days of age and 350-550 grn, were housed in wire mesh cages in groups of five with ad lib access to food and water. Apparatus Conditioning was carried out with the animal placed in a sound ~ttenuated chamber. The CSs were 500 and 1000 Hz tones of an

.~.,'7'R

GAEBELEIN, ET AL.

Pay. 1. Biol. Sol. At~ril-lune 1976

intensity clearly audible over background noise produced by two tone generators and delivered through two speakers located next to the animal. The UCS was delivered from a constant-current shock source through a tail electrode (Weiss, 1967) at an intensity which produced an HR acceleration (X = 0.4 mA). Tone and shock presentations were controlled with solid-state circuitry. Succinyldloline chloride (Ancctine, l~urroughs-Wcllcome) at a cmlcentration of 5 mg/cc was infused with a ttarvard infusion pump (Model 600-000). To ensure that overdosage did not result, infusion rates were altered during the session to maintain minimal EMG activity. Catheters for infusion and sampling of blood gas were similar to those described by Popovic and Popovic (1965), consisting of 12-cm lengths of polyethylene tubing (PE 20, Clay Adams). One end of the catheter was cut obliquely, and the other was heat-sealed after the catheter had been filled with a sodium-heparin solution. The respiratory system has been described elsewhere (Gaebelein & Howard, 1974). Briefly, air was delivered to the animal with a Harvard rodent respirator (Model 681) modified to provide an inspiration: expiration ratio of 1:3 (Gaebelein, Howard & Hutcheson, 1974). A pressure transducer (Statham P-23) allowed continuous monitoring of respiratory pressure. The minimum expiratory pressure was adjusted as needed to 0 cm tI~O by altering the magnitude of a small negative pressure exerted with the sample pump of a Beckman Medical Gas Analyzer (Model CB-1). Expired air was passed through the CO~ monitor of the gas analyzer. Physiological measures were recorded with a Beckman Dynograph. These included respiratory pressure calibrated in cm H:O; and expired CO._, expressed as per cent content of expired air. Stainless steel needles placed subdermally in the thorax were used to record EKG, which was converted to beats per minute with a Microtronics cardiotachometer. Electrodes were placed in the cervical neck muscles to record EMG activity at a sensitivity of 1 ~,V/mm. Rectal temperature was monitored with a Yellow Springs Telethermometer (Model 44-TA) and maintained at 37• 2 C with a heating pad. Blood gas tensions were determined with an Instrumentation Laboratories (Model $113) blood gas analyzer.

Procedure Catheters were implanted in the cervical portion of the right jugular vein and the left common carotid artery through ventral incisions under ether anesthesia. The catheters were secured with ligatures and externalized through the incision, which was closed with wound dips. Immediately following surgery an endotraeheal tube was in-

Volume 11 Number 2

L E A R N I N G AT VARIOUS BLOOD GAS LEVELS

79

serted into the rat using a method described elsewhere (Gaebelein & Howard, 1974), and the rat was attached to the respiratory system. As the ether anesthesia wore off, signs of struggling appeared. Succinylcholine chloride was then infused tlarough the venou~ catheter until obvious struggling ceased. Then the EKG, EMG, and rectal electrodes were attached to the animal, and the infusiol~ continued at a rate sufficient to suppress EMG. During the next 10 min, respiratory parameters were adiusted to maintain the rats at peak expired CO., values between 4.7-4.9 pet cent (n = 5), 5.0-5.1 per cent (n -- 7), or 5.2-5.6 per cent (11 = 8). These intervals were selected to explore a range of values, although it was expected that the 5.0-5.1 per cent interval would lead to normal blood gas values, as has been previously shown with cats (Howard, unpublished observations). During conditioning, rate and volume of respiration were altered to maintain peak expired CO._, at the desired level. Conditioning was carried out using a method similar to that suggested by Seligman (1969) to provide within-subiect control for sensitization effects. That is, two tones, each of I0 sec duration, were presented to the animal. One tone (CS-~-) was consistently associated with shock, while the other (CS--) was randomly distributed and was followed with shock only when it occurred cuncomitantly with CS+. The average intertrial interval for CS+ a1r CS-- tones was 70 sec (range 50-90 sec). This resulted in 30 CS+, 3(1 C S - , and 5 combined presentations during the 40 min conditioning period. At the end of conditioning, blood gas tensions were determined from a 0.5 cc sample obtained from the arterial catheter. Prior to data quantification, polygraph records were visually examined to exclude trials from analysis if (a) there was obviou~ evidence of incomplete paralysis as reflected in EMG, (b) the respiratory measures suggested a ventilatory deficiency such as a a partially occluded endotracheal tube, or (c) the pattern of HR was clearly abnormal as compared to HR pattenas observed during other segments of the session, e.g., HR shifts resulting from alteration of the rate or volume of respiration. In addition, CS--trials were excluded from analysis if they occurred within 20 sec of a UCS, to prevent artifactual HR changes resulting from the UCR. Trials in which both tones were presented together were also excluded. Trials included in the analysis were quantified by averaging HR for the 5 see prior to tone onset (Preperiod) and in 2-see intervals during the 10 sec of tone presentation. For each trial the difference between the average HR of the Preperiod and the average HR value of each interval following tone onset was obtained. Then, scores during

80

GAEBELEIN, E T AL.

TABLE 1.

Exp~ed CO2 Pco2

Pay. ]'. Biol. Sol. Al~ril-.lune 1976

Product-moment Correlations Among Respiratory Measures Pco2

Po2

pH

.594

-.565

-.480

-.694

-.746

Po~

.606

CS+ and C S - trials during the last third of the session were averaged individuaUy, and the absolute differences between these latter means used to assess conditioning effects. For all comparisons of HR data~ animals were discriminated into optimum and non-optimum groups on the basis of blood gas analysis. That is, rats were included in the optimum group if Pco~ exceeded 35 mm Hg, andpH was greater than 7.30. These values are within the range reported for normal rats (Gaebelein & Howard, 1974) and resulted in 9 animals with normal blood gas values (XPco._,= 38.7, XpH = 7.40, XPo., = 55.9). Animals displaying abnormal blood gas values could bc grouped according to expired CO.,. Five rats respirated at less than 5 per cent expired CO., displayed average values of Pco._,, pH, and Po., of 31.0, 7.47, and 59.5, respectively. Five animals ventilated to produce expired CO., levels greater than 5 per cent showed mean blood gas values of Pco~ = 39.1, pH = 7.39, and Po._,= 48.3. Because these latter groups did not differ behaviorally, their data were combined for comparison with animals displaying optimum blood gas values. The remaining animal was eliminated from the study for reasons unrelated to the procedures, i.e., misplacement of the EMG electrodes precluded an adequate assessment of the depth of paralyzation. For all statistical comparisons, a minimal level of significance of ,, < .05 was selected. Results

The sensitivities of blood gas values to manipulations in expired CO.~ were evaluated with correlations among these variables as well as among the blood gas tensions and pH (Table 1). It is apparent that blood gas values are alterable by changes in expired CO._,: increases in expired CO., were associated with increases in Pco.,, and decreases in Pc).. and pH. Further, both expired CO-, and blood gas Pco., displayed a similar relationship to Po.., and pH. The correlation of Pco., with Po2 was similar to that between expired CO., and Po2. A similar result was obtained when the con-e-

Volume I1 Number 2

L E A R N I N G AT VARIOUS BLOOD GAS LEVELS

81

A

o.

5

v

uJ 4

FIC. 1. Mean differenee in heart rate response to discriminative stimuli for groups with optimum and non-optimum blood gas values.

z -r ~ 3 w ~n- 2 w

-r-

I

~1

0"--"0 9 --

0

O P T I M U M RATS "Q N O N - O P T I M U M R&TS

i

I

I

I

I

I

2

3

4

,5

BLOCKS OF SECONDS

lation between Pco.~ and pH was compared to that between expired CO., and pH (McNemar, 1949). Peak expired CO.- also seemed to be a consistent predictor of 9blood gas values. That is, 7 of 9 rats with adequate blood gas values were respirated between 5.0 and 5.1 per cent peak expired CO._,. Departures from this interval in either direction resulted in non-optimal values. It is also noteworthy that all rats respirated between 5.0 and 5.1 per cent showed adequate blood gas values, while only 2 of the remaining 12 animals did so. Rats with normal blood gases also displayed lower baseline HRs (X = 408.7) than did the other rats (X = 491.6) (t,7 = 4.04), although both groups showed similar tlR variability. Neither tlR nor HR variability, however, were correlated with Pco..,. Group differences in conditioning were assessed with a mixed design analysis of variance, which revealed a significant Group x Blocks interaction (F 4, 69 = 2.74) (Fig. 1 ). Tests of simple effects indicated that the difference in HR response during tone presentation was greater for the optimum rats than for tile non-optimum animals. This difference was most pronounced during the 2nd and 3rd blocks of tone presentation. To insure that the conditioning effects were characteristic of individual animals, and not due to averaging "good" with "poor" learners, the difference scores were summed over the 5 blocks of tone presentation for each animal and ranked. A test of these ranks between the groups (Ferguson, 1966), suggested that animals ill the optilnum group displayed larger differences than rats in the non-optimum group (Z -- 2.09).

8 "~

GAEBELEIN, ET AL.

Pay. J. Biol. Scl. Avril-Junc 1976

Discussion These findings suggest that, at least in rats immobilized with suecinylcholine, monitoring end tidal peak expired CO., provides an adequate means of preserving Pco., and ptI within the normal physiological range. This method also enables, in the paralyzed rat, a relatively practical manipulation of inspired O._, for experimental control of certain blood gases essential for regulating respiratory acidosis or alkalosis, as reflected in the magnitude of the correlations reported in Table 1. These results also suggest that there is a narrow interval of peak expired CO., which will consistently produce normal blood gas levels: all rats with peak expired CO., values between 5.0 and 5.1 per cent displayed normal Pco~ and pH levels, while normative values were obtained in only two of the renmining rats. Criteria used by other workers, however, do not seem to be as sensitive as alterations in peak expired CO._.. Miller and DiCara (1967), for example, adiusted respiratory parameters until baseline HR fell within a pre-described range, usually 300-500 beats per minute. All rats with normal blood gases, and some abnormally respirated rats, displayed HRs within this range. However, with one exception, all optimally-respirated rats had lower HRs at the end of conditioning than did the remaining animals. In addition, little support for the use of HR variability as a criterion of ventilatory adequacy was obtained, since this measure did not discriminate group differences in blood gas levels. Further, neither HR nor HR variability seemed to be related intimately to Peon., as reflected by the correlations. Our respiratory system, however, is not without di~culties. For example, the percentages of peak expired CO_~ levels reported here are not expressed in absolute units, because the normal respiratory rate of rats exceeds that required by the gas analyzer to display 0.0 per cent expired CO., at the end of each respiratory cycle. However, pilot work suggested that maintaining rats at a peak expired CO., level of 5 per cent resulted in more normal blood gases than did other peak expired CO.., values. Given that the presently available expired gas analyzers have too long a time constant to completely follow the rapidity of shifts of CO., in rats' expired air because of their high respiratory rate, it seems that each laboratory should calibrate different peak expired CO: levels against blood gas values in order to establish the proper respiratory setting. Once this is done, peak expired CO2 should be a reliabl~ and easy means of maintaining rats in a normal physiological state.

Volume Number 11 2

LEARNING AT VARIOUS BLOOD GAS LEVELS

~:3

Another difficulty with the respiratory system is the apparent inability to maintain proper levels of Po:. It seems possible that gas exchange efficiency was reduced as the duration of paralysis increased due, perhaps, to a progressive reduction in lung compliance or to the histamine-releasing properties of the.drug (Howard, et al. 1974a). The levels of P~r_,may be increased by adding O~ to the inspired air, or by pre-treatment with an anti-histamine. In any case, animals respirated at peak expired CO., values of 5.0-5.1 per cent showed greater differentiation of HR responses between CS+ and C S - tones than did the remaining animals. The direction of the HR response, however, was inconsistent across animals. For example, 4 optimum animals showed HR acceleration to CSq--, while the remaining 5 animals decreased HR to this stimulus. Using non-curarized rats in an identical paradigm, Howard, et al. (1974b) reported HR acceleration as a CR with tail shock, and ttR deceleration with foot shock UCS. Since this study used tail shock, an increase in HR would be expected to the tone. Goesling and Brener (1972) have recently demonstrated that the direction of HR change observed in an operant conditioning paradigm was determined more by previous experience (i.e., active or passive avoidance) than by tile reinforcement contingencies used during curarization. Since activity levels in this study were not systematically cwduat

Classical aversive conditioning of heart rate in curarized rats at different blood gas levels.

Classical Aversive Conditioning of Heart Rate in Curarized Rats at Different Blood Gas Levels CLAUDE J. GAEBELEIN, JAMES L. HOWARD, RICH~D A. GALOSY,...
577KB Sizes 0 Downloads 0 Views