Cardiorespiratory function is altered by picomole injections of 5′-N-ethylcarboxamidoadenosine into the nucleus tractus solitarius of rats
Robin A. Barraco, Cynthia A. Janusz, Eugene P. Schoener and Lori L. Simpson
Department of Physiology,Wayne State University School of Medicine, Detroit, MI (U.S.A.)
(Accepted 20 June 1989)
Key words:Adenosine; Cardiorespiratory control: Nucleus tractus solitarii; 5′-N-Ethylcarboxamidoadenosine
A limited occipital craniotomy was conducted on urethane-anesthetized,spontaneously breathing rats to expose the caudal medulla in the region of the obex.Microinjections of 5′-N-ethylcarboxamidoadenosine (NECA), an adenosine analog, were made into the medial region of the caudal nucleus tractus solitarius (NTS) at the level of the caudal tip of the area postrema, an area of the NTS in which there is known to be a functional co-existence of cardiovascular and respiratory-related neuronal elements. Cardiorespiratory responses were subsequently recorded for a 60 min test period.Microinjections of NECA, in the dose range of 0.35-350 pmol per rat, produced significant dose-related reductions in respiratory rate which were accompanied by dose-dependent increases in tidal volume and these pronounced effects on respiration persisted throughout the test period. In contrast, the effects of NECA microinjections on cardiovascular parameters in this region of the NTS were bidirectional and elicited considerably more complex responses during the test period. During the initial period (2-5 min) following injection,NECA elicited significant hypotension (at lower doses) and pressor responses (at higher doses) in addition to significant bradycardia (at lower doses) whereas by the end of the 60 min test period, almost all doses of NECA had resulted in hypertension and tachycardia. Multivariate analysis of variance (MANOVA) and correlation statistics indicated that the effects of NECA on blood pressure during the initial 2-5 min were dose-dependent and unlikely related to depression of respiratory frequency.A further examination of the data by MANOVA indicated that the pharmacological effects of NECA during the 60 min test period exhibited a highly significant and specific dose-dependent and time-related response pattern for the respiratory,but not the cardiovascular,parameters.Taken together,these manifold response patterns suggest that the respiratory effects of NECA may be mediated by different intrinsic mechanisms in the NTS than are the cardiovascular effects of NECA. At the end of the 60 min test period following the administration of NECA,the respiratory rate remained profoundly depressed. In view of previous studies showing that microinjections of cyclic AMP analogs,forskolin,isoproterenol and adenosine into the same NTS sites elicit a similar depression of respiration,the results with NECA in the present study further support the notion that cyclic AMP may serve as a second messenger in NTS respiratory control regions and these respiratory depressant effects may be mediated by a single adenosine receptor subtype. On the other hand,in contrast to previous studies wherein adenosine itself produced a relatively brief hypotension and bradycardia following microinjections into the same region of the NTS,the data from this study show that microinjections of NECA,a metabolically stable analog of adenosine, exert prolonged and multiplex modulatory influences on cardiovascular parameters in the intact rat. The data further suggest the possibility that NECA,a mixed agonist for adenosine receptor subtypes,may thereby influence cardiovascular response patterns by activating more than one population of adenosine receptor subtypes on neuronal elements in the NTS. Finally, in view of the potent dose-dependent depressant effects of NECA on respiration, the data from this study further support the notion that adenosine plays a neuromodulatory role in brainstem cardiorespiratory control regions.The findings from this study showing the multiplex cardiorespiratory responses of NECA over time also lend support to the case for using in vivo models to assess thepharmacological actions of neuroactive agents.
During the past 50 years,respiratory neuronal activity has been associated with the dorsomedial and ventrolat-eral aspects of the medulla and it is now apparent that respiratory rhythmicity results from the interactions between sensory respiratory inputs and the intrinsic activity of neuronal networks in the brainstem’9. How-ever,subsequent anatomical and physiological studies have more specifically identifed the location and types of

respiratory neurons in the brainstem and the extent of respiratory neuronal activity with the medulla has been enlarged by the recent identification of other concentra-tions of neurons which appear to be part of the respiratory system4. Specifically, much recent work has focussed on a concentration of inspiratory neurons on the dorsomedial aspect of the caudal medulla, associated with the nucleus tractus solitarius (NTS), and occasion-ally referred to as the dorsal respiratory group (DRG)31. 44.Moreover,it has been shown that the caudal region of
Correspondence:R.A.Barraco,Department of Physiology,Wayne State University,School of Medicine,540 E.Canfield,Detroit, MI 48201. U.S.A.
0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division) 
the NTS is the major site of termination of afferent fibers from arterial and cardiopulmonary mechanoreceptors and chemoreceptors38 and it is also known that a pool of inspiratory neurons in this region not only responds to vagal inputs but also profusely projects its axons into the spinal cord37.Thus,in light of the extensive connections with other respiratory-related regions in the brainstem, the NTS has been considered to play an important role in the generation of rhythmic breathing and it is the probable site for integration of the many influences which affect the amplitude and frequency of breathing.How-ever, cardiorespiratory homeostasis in mammals requires not only the regulation of alveolar gas exchange via the respiratory system but also the coordination of respira-tory rhythm with sympathetic and parasympathetic mod-ulation of blood-gas transport via cardiovascular activity32.Thus,from the perspective of central control mechanisms,especially those involving the brainstem,it has been proposed that the concept of cardiorespiratory homeostasis can be viewed organizationally as a unitary system wherein the afferent and efferent pathways subserving cardiorespiratory control follow a parallel course in the brain; this allows for multiple sites where coordination of cardiorespiratory function can take place32. In light of the fact that the NTS,particularly the caudal part together with the commissural nucleus,forms the primary medullary center for multisynaptic cardio-vascular reflexes and since it is richly connected with virtually all CNS areas involved in cardiovascular control’7,it is not surprising,therefore,that the caudal NTS has been considered to be a major integrating site for coordinating cardiopulmonary information and for the consequent production of appropriate patterns of cardiorespiratory homeostatic responses2. Indeed, nu-merous stimulation studies have confirmed the coexist-ence of respiratory and cardiovascular neurons within various subnuclei of the caudal NTS25.38.47-49.
Furthermore,this region of the dorsal medulla,espe-cially the caudal region comprising the area postrema, the NTS and the dorsal motor nucleus of the vagus,has attracted considerable attention recently due to the impressive array of neurotransmitters and putative neu-romodulators occurring within the somata and fibers of this nuclear complex*1. Thus it is likely that one or more of the endogenous neurochemicals, synthesized and released by neurons in the NTS and related structures, may exert modulating effects on intrinsic neuronal networks generating respiratory rhytms and integrating cardiovascular response patterns. Indeed,along with numerous monoamines,adenosine has been considered to be a possible modulator of central respiratory control in a number of physiological and pathophysiological conditions28-30.46. It has been shown. for example.that

adenosine analogs can markedly depress respiration following either intraventricular administration295 or microinjections directly into the NTS”. and that the respiratory stimulant effect of methylxanthines is due to blockade of central adenosine receptors involved in respiratory control3.Moreover, adenosine has also been implicated in the paradoxical ventilatory response to hypoxia seen in neonates21.054. On the basis of these studies and other findings,it has been proposed that adenosine acts as a tonic neuromodulator of respiration 29. Similarly,other studies have implicated adenosine in the modulation of central cardiovascular control-5.7.56 For example,microinjections of adenosine into the area postrema or caudal NTS of rats have been shown to elicit potent cardiovascular responses which can be blocked by methylxanthines7.56.
Thus,in light of the many lines of evidence implicating the caudal NTS as a major site of integration for coordinating cardiopulmonary information and the con-sequent production of cardiorespiratory homeostatic responses,it is conceivable that adenosine may play a neuromodulatory role in this region of the NTS. This possibility is further supported by the intriguing findings that the caudal NTS exhibits the highest density of adenosine uptake sites of any region of the brain10. 11.22. Consequently,the aim of the present study was to examine the effects of 5′-N-ethylcarboxamidoadenosine (NECA), a metabolically stable and potent analog of adenosine. on cardiorespiratory parameters following discrete microinjections into the caudal NTS of sponta-neously breathing intact rats. Some preliminary results from this study have been reported”.
Male Sprague-Dawley rats (275-300 g) were anesthetized with urethane(1 g/kg),the trachea was intubated and the femoral artery and vein cannulated for recording of blood pressure and systemic drug administration, respectively. In some animals,the other femoral artery was cannulated to obtain samples for blood gas analysis.Thereafter,the subject was mounted in a cranial stereo-taxic apparatus (Kopf) in a prone position.Blood pressure record-ings were made with a Gould Statham transducer adapted to a Grass Model 7 polygraph.The transducer and polygraph recorder were calibrated with a sphygmomanometer prior to blood pressure recordings. Heart rate was electronically derived from blood pressure via a Grass Model 7P4F Tachograph. Body temperature was maintained at 37.5℃ (±1℃) via electric heating pad and monitoring by rectal thermocouple.Respiratory parameters were measured using a modified Fleisch pneumotachograph. The device was manufactured in our laboratory to provide minimum dead space and for convenient attachment to the tracheostomy tube so that the pattern of breathing was not disturbed. The device was an 18 mm stainless-steel tube with 40 x SS wire mesh across the 3 mm cross-sectional diameter half way down the tube.One mm ports on either side of the screen permitted attachment of a Statham PM 15E differential pressure transducer.Differences in pressure across the resistance were linearly related to air flow in the range needed for breathing in rats. The index of air flow was integrated on the 
polygraph(Grass Model 7P10B Signal Conditioner and Integrator) for inspiration only and tidal volume (V,) was thereby obtained.The measurement was calibrated through the system with a glass syringe (accuracy:±1.0%)for V,in the range of 1-2 ml/breath prior to each experiment.The respiratory rate (R,) was provided by a Grass Model 7P4 tachograph. Samples (0.1 ml) for arterial blood gas analysis were obtained from the second femoral line at various times during some of the experiments and analyzed by a clinical blood gas analyzer(Radiometer,Model ABL 2) in order to periodically assess the integrity of the preparation and also to ascertain whether blood gases were affected as a result of altered breathing patterns induced by NECA microinjections.
A limited occipital craniotomy was conducted after electrocautery of neck muscles and then delicate dissection was employed to reveal the caudal medulla in the region of the obex.Microinjections were made stereotaxically into brainstem sites with beveled glass micro-pipettes with a tip diameter of 60-80 μm.The entire microinjection procedure was conducted under a dissecting microscope with a micrometer disc calibrated to 0.01 m at 7x. The rat skull was adjusted to a 45° angle from the horizontal plane with the micropipette barrel held at a 28° angle from the vertical plane. The coordinates for microinjections into the medial portion of the caudal NTS were: AP:-0.2 mm (from caudal tip of the area postrema); ML: 0.3 mm; DV:0.3 mm from dorsal surface of brainstem.The pipettes were filled by suction with drug solutions containing varying concentrations of NECA (5′-N-ethylcarboxamidoadenosine) (Sigma) freshly prepared in artificial cerebrospinal fluid (ACF) and the pH was adjusted to the pH of the ACF (pH 7.2-7.3).ACF was used for control injections.Drug or control solutions were injected in 50 nl volume over 15 s period.Only one injection of drug was made per animal.
Cardiovascular and respiratory parameters were recorded for a period of 60 min. Experimental values were calculated as the mean (± S.E.M.) percent of control values determined prior to injection for each animal.Mean arterial blood pressure (P.),pulse pressure (PP), heart rate (HR), tidal volume (V,), and respiratory rate (R,) were measured prior to injection (control values),then at 2,5.10, 15,30 min following NECA administration (or ACF in control animals) and at the end of the 60 min test period.Dose-response data at 2, 5, 10. 15. 30 and 60 min were analyzed by analysis of

Fig.1.The effects of varying doses of NECA on respiratory rate at 6 separate time points (2, 5, 10, 15,30 and 60 min) during a 60 min test period following microinjection into the caudal NTS.Values are given as the mean(±S.E.M.) percent of control values which were established prior to injection. The NECA dosages are shown in the legend at the upper right corner. The number in parenthesis at the 60 min time point indicate the number of animals tested for each dose of NECA for the test period. Significance levels at each time point for individual doses were determined by post hoc comparisons (two-tailed) with ACF control animals (n = 13) at the corresponding time point following injection of only vehicle:*differs from vehicle controls for a given time point at P <0.01. variance(ANOVA) and post-hoc,pair-wise comparisons were made using Student's t-test(two-tailed)with time-matched ACF control animals at each time point. Statistics were generated with the SPSS/PC+ (V 1.0);ANOVA and analysis of covariance statistics were produced using the MANOVA module in SPSS/PC+ Ad-vanced Statistics with a full-factorial repeated measures design.The Pearson product-moment correlation coefficients between pairs of variables at 2,5,10,15,30 and 60 min following microinjection were generated with the correlation module in SPSS/PC+ which uses the I-statistic for calculating the significance levels of the correlation. At the completion of each experiment, brains were perfused transcardially with 10% buffered formalin and subsequently pro-cessed histologically for microscopic verification of injection sites as described previously. Parameters affecting microinjection of drugs into the NTS, details for marking injection sites,histological identification of specific NTS regions and subsequent diagrammatic mapping of NTS sites have been described elsewhere RESULTS The control values (n = 57) for cardiovascular and respiratory parameters established prior to injections were:mean arterial blood pressure(Pa) =89. 2.1mm e(pa)=89.3±2.1 Hg; heart rate (HR) = 346.1 ± 6.8 bpm; tidal volume (V1)=0.32±=0.32±0.02 ml/100g;respiratory rate(te(Ry)=144.7 ± 4.1 breaths/min;minute volume (Vm)=46.3±2.4 ml/min/100 g. Control values (n= 9) for arterial blood gases after the surgical preparation and the craniotomy but prior to microinjections were:pCO2=38.9±1.4 Fig.2.The effects of varying doses of NECA on tidal volume at 6 separate time points (2, 5, 10,15,30 and 60 min) during a 60 min test period following microinjection into the caudal NTS.Values are given as the mean (± S.E.M.) percent of control values which were established prior to injection.The NECA dosages are shown in the legend at the upper right corner. The number in parentheses at the 60 min time point indicate the number of animals tested for each dose of NECA for the test period.Significance levels at each time point for individual doses were determined by post hoc comparisons (two-tailed) with ACF control animals (n = 13) at the corresponding time point following injection of only vehicle:*differs from vehicle controls for a given time point at P <0.01.  mm Hg;pO2=93.7±4.1mm Hg;pH=7.36±0.01. Control injections of ACF (n = 13) had no observable effect on any of the cardiorespiratory parameters exam-ined throughout the 60 min test period. The effects of microinjections during the 60 min test period of varying doses of NECA on respiratory rate (R,),represented as the mean(± S.E.M.) percent of preinjection control values, are shown in Fig. 1. ANOVA indicated that NECA not only produced significant dose-related reduc-tions in respiratory rate by the end of the 60 min test period (F4,39 = 15.31;P <0.001) but that NECA produced significant dose-related reductions in R, even at the earliest time point examined (i.e.2 min after injection):F4.39=3.32;P<<0.02). In fact, ANOVA indicated NECA produced significant dose-related reduC-tions in at all times examined(i.e.at 5 min:F4.39= 9.51;P<0.0<0.001;at 10 min:F4.39=13.41;P<0 3.41;P<0.001;at 15min:n:F4.39=12.07;P<0.<0.001;at 30 min:F4.39=13.25; F4.39=13 P <0.001)during the 60 min test period. Significance levels,determined for individual doses during the 60 min test period by post hoc comparisons with time-matched ACF controls,are indicated by asterisks in Fig.1.The time of onset (T.) for depression of respiration was rapid for all doses with a range of 15.00 ± 0.67 s for the highest dose to 37.22 ± 4.01 s for the lowest dose. The effects of microinjections during the 60 min test period of varying doses of NECA on tidal volume (V.), represented as the mean(± S.E.M.) percent of prein-jection control values, are shown in Fig. 2. ANOVA indicated that NECA produced significant dose-related increases in V, by the end of the 60 min test period(F4.27 =11.25;P<0.001). Moreover, ANOVA indicated that NECA produced significant dose-related increases in V, as soon as 5 min after injection(F427=2.64;P<0.05) and at every time examined thereafter (at 10 min:F4.27 = 4.61;P <0.006; at 15 min: F4.27=6.39;P< at 30 min:F4.27=1F4.27=10.96;P<0.001).Significance levels, determined for individual doses during the 60 min test period by post hoc comparisons with ACF controls,are indicated by asterisks in Fig.2. Moreover,arterial blood gas values were not profoundly affected following NECA microinjections although there were slight differences in comparison to control values for pCO2at the end of the test period for only the highest doses.For example.60 min following microinjection of the highest dose of NECA (1.0 nmol/kg),at a time when R, was depressed on the average by more than 50%(Fig.1),the blood gas values were:pCO2=33.1±1.2mm Hg(t=3.15;P< TABLE I Effects of varying doses of NECA on cardiovascular parameters at 6 time points during the 60 min test period Values are expressed as the mean (±S.E.M.) percent of control values established prior to injection: n, number of animals tested at cach dose; Pa mean arterial blood pressure;HR.heart rate;PP,pulse pressure. ANOVA indicated that NECA produced significant dose-related effects on p at 2min(F4.39=2.84;P<0.05),5min(F4.39=10.96;P<0.001),10min(F4.39=5.17;P<0.002), and 15 min(F1_3=3.99:P<0.01)but not at 30 and 60 min.NECA produced significant dose-related effects on HR at 10 min(F4.3y=3.29;P<0.02),15min(F43)=3.79:P<0.01),and 30 min (F4.39=2.79 = 2.79; P <0.05) but not at 2.5 or 60 min. NECA produced significant dose-related effects on PP at 5min(F4_w=4.93;P<0.005)and 10 min (F4.39=4.43;P<0.005) but not at 2,15,30 or 60 min.Significance levels for PP at each time point determined by post hoc comparisons (two-tailed) with ACF controls (time-matched):*differs by P<0.05:**differs by P<0.01.Post-hoc comparisons for p. and HR are provided in Figs.3 and 4.respectively. Dose(nmol/kg) n Time(min)fol 2 5 10 15 30 60 Cardiovascular par p.0.01 9 78.5±7.2 83.5±5.6 93.9±7.2 99.4±6.9 85.1±5.7 115.1±9.3 0.05 10 79.4±3.7 87.1±5.4 93.4±6.6 88.9±6.0 87.8±7.9 97.8±3.4 0.10 9 97.4±6.4 112.3±4.8 124.1±6.2 125.6±10.6 109.5±9.6 116.2±11.3 0.50 7 105.6±9.4 133.3±8.6 127.0±12.0 115.1±10.8 114.6±13.4 127.9±9.7 1.0 9 88.3±7.2 101.9±5.3 92.6±6.8 86.6±8.3 88.9±7.9 120.0±8.7 HR 0.01 9 87.5±6.9 85.2±6.6 89.1±1.4 90.4±1.9 91.8±3.3 105.5±6.2 0.05 10 91.7±1.9 86.7±2.8 86.8±2.5 87.1±2.3 89.4±2.4 95.5±4.1 0.10 9 92.7±4.4 90.1±5.3 95.5±4.6 97.3±4.3 102.4±4.6 107.4±5.1 0.50 7 96.7±3.0 101.9±2.9 104.2±6.6 106.1±6.1 102.4±3.6 114.0±6.4 1.0 9 96.3±2.3 97.8±2.5 95.8±2.5 93.9±3.1 99.4±4.2 113.7±4.5 PP0.01 9 79.9±5.0* 86.2±2.6* 90.9±3.1 95.1±3.8 91.4±4.8 115.8±7.6* 0.05 10 82.5±5.1* 94.3±6.3 98.1±6.2 96.9±6.3 97.0±7.6 108.6±5.4* 0.10 9 98.5±4.4 104.8±2.4 116.4±5.6** 122.6±11.1** 113.8±13.1 120.8±14.1* 0.50 7 101.1±4.2 118.6±8.0** 115.9±5.3** 112.3±7.3** 120.6±15.3 128.3±10.8** 1.0 9 90.8±7.8 100.7±5.4 104.7±5.1 99.8±6.9 101.5±6.9 120.9±5.9** *P<0.05,**P<0.01;see above.  Fig.6.Tracing of respiratory rate,blood pressure,tidal volume and heart rate from a rat following microinjection of NECA into the caudal NTS at the 0.1 nmol/kg dose. trols, are indicated by asterisks in Figs. 3 and 4;for PP. significance levels for post hoc comparisons are provided in Table I.In contrast to R, and V,, at the end of the 60 min test period. ANOVA indicated that NECA did not produce any significant dose-related effects on pa(F4.39= 2.19;N.S.),pulse pressure (PP) (F4.39 =0.62;N.S.)or HR(F4.39= 1.64; N.S.), although post-hoc comparisons(F4.39 with time-matched ACF controls at 60 min showed that pa and HR were significantly increased following injec-tions of the highest doses of NECA (Figs. 3 and 4). Throughout the 60 min test period, the effects of varying doses of NECA on p.. PP and HR were quite complex. at times producing significant increases or decreases depending on the dosage (Figs. 3 and 4).The results of ANOVA for pa, HR and PP following microinjections of NECA are provided in Table I. As can be seen from the F values given in the legend for Table I, ANOVA indicated that NECA produced significant dose-related effects on pa. HR and PP predominantly at the earlier time points during the test period.Indeed,particularly during the initial 10 min period following microinjections, NECA exerted quite pronounced effects on cardiovas-cular parameters (Figs. 3 and 4). NECA exerted bidi-rectional responses on p. at the earlier time points; at the higher dosages (0.5 and 0.1 nmol/kg),NECA elicited pressor responses whereas hypotensive responses were Fig.7.Tracing of respiratory rate,blood pressure, tidal volume and heart rate from a rat following microinjection of NECA into the caudal NTS at the 1.0 nmol/kg dose.  TABLE II Pearson product-moment correlations between respiratory parameters and cardiovascular parameters at 6 time points during the 60 min test period following microinjection of NECA All significance levels are 2-tailed.R,.respiratory rate;V..tidal volume:pn,mean arterial blood pressure;HR, heart rate:PP.pulse pressure. Parameters Time(min)fo correlated 2 5 10 15 30 60 R1xPa -0.188 -0.303* -0.310* -0.201 -0.231 -0.049 R,xHR 0.104 -0.197 -0.312* -0.297* -0.265 -0.136 R,xPP -0.189 -0.390** -0.492*** -0.354** -0.258 -0.141 R,xV1 -0.233 -0.407** -0.505** -0.569*** -0.626*** -0.643*** P1xPP 0.771*** 0.696*** 0.603*** 0.695*** 0.651*** 0.707*** P』xHR 0.675*** 0.597*** 0.434** 0.432** 0.493*** 0.479*** elicited by the lower doses of NECA (0.05 and 0.01 nmol/kg)(Fig. 3). A similar multiplex response was seen with PP at theearlier time points for the lower and higher doses of NECA (Table I). Moreover, NECA elicited significant bradycardia at the lower dosages (0.05 and 0.01 nmol/kg) which persisted for 30 min whereas the higher doses of NECA apparently caused no significant effects on heart rate during the initial period following injection (Fig.4). In contrast to the bidirectional cardiovascular effects at earlier time points, post hoc comparisons with time-matched ACF controls at 60 min show that almost all doses of NECA resulted in significant hypertension and tachycardia by the end of the test period (Figs. 3 and 4). Similarly,PP was significantly increased with all doses of NECA by the end of the 60 min test period (Table I). Interestingly,MANOVA further indicated that there were only two time intervals during the 60 min test period wherein pa was significantly affected across all doses of NECA;for the 2 min vs 5 min interval: F1.39 = 14.45;P < 0.001; and for the 30 vs 60 min interval:F1.39=37.26; P<0.001.These results indicate that the most significant changes in pa following NECA microinjections occurred at the very beginning and at the end of the 60 min test period. In contrast to the results with pa,MANOVA indicated that R, and V, were significantly affected following NECA microinjections for all time intervals during the 60 min test period except for the 30 vs 60 min interval.Thus, for the 2 vs 5 min interval(R:F1,39=96.49;P< =96.49;P<0.01; V1:F1.27=31.42;P<0.001);0.001);for the 5 vs 10 min interval (Rf:F1.39=86.04;P<0.001;V1:F1.27=29.2 0.001); for the 10 vs 15 min intervalal(Ry:F1.39=85.37; P<0.01;V1:F1.27=44.2;P<0.001)1);for the 15 vs 30 min intervalrval(R1:F1.39=38.76;P<0.001;V1:F1.27= 45.95;P<0.001);fofor the 30 vs 600 vs 60 min(R4:F1,39=3.29; P=0.077;V1:F1.27=0.78;P=0.385).These results indicate that unlike the effects of NECA on Pa,the time-related changes in R, and V,across all doses of NECA occurred at almost all intervals of the test period. Moreover,since there were no significant time-related changes in R, and V, across all doses of NECA for the 30 vs 60 min interval, it appears that the depression of respiratory frequency reached an asymptote in magnitude at 30 min and this depression persisted without much change until the end of the 60 min test period.In fact,as can be seen from the post hoc comparisons for R, in Fig. 1,only the 0.05nmol/kg dose of NECA obtained a level of significance during the 30-60 min interval of the test period whereas the 3 highest dosages had already significantly depressed R,at 30 min and earlier. In Figs. 5-7 are shown tracings of cardiorespiratory parameters during the 60 min test period following microinjections of NECA at the 0.05,0.10 and 1.0 nmol/kg dose, respec-tively. The multiplex effects of NECA microinjections on cardiovascular parameters raises the possibility that some of the changes in cardiovascular parameters were corre-lated with the pronounced depression of respiratory frequency exerted by NECA. To examine the degree of association between respiratory parameters and cardio-vascular parameters at a given time point,correlation coefficients between pairs of variables were calculated (Pearson product-moment correlation coefficients) across all doses at each time point. The correlation results for each time point are shown in Table II. As can be seen from Table II,there were only marginally significant correlations between R,and pa at 5 and 10 min and with HR at 10 and 15 min and the correlations between R, and PP were significant at 5, 10 and 15 min.In contrast,the correlations between Pa and the other cardiovascular parameters (PP and HR) were highly significant at all time points.Moreover,although R, was only occasionally correlated with pa and HR, at the intermediate time  caudal NTS,elicited potent cardiorespiratory responses. The possibility that adenosine plays a modulatory role in mechanisms of cardiorespiratory regulation by the NTS29,has important physiologic and pathophysiologic implications. It is known, for example,that adenosine levels in the extracellular fluid of the rat brain are elevated during hypoxia and ischemias2 and there is mounting evidence that adenosine levels are elevated in the NTS itself following brief exposure to hypoxia. Much of the recent work on adenosine relates to its modulating actions on adenylate cyclase in which two extracellular adenosine receptor subtypes have been postulated.Those designated A, receptors are associated with decreases in cyclic AMP(cAMP)while A2 receptors are associated with increased cAMP levels16.27.Addition-ally, in some tissues, there is evidence that adenosine receptors are directly coupled to potassium channels13 or calcium channels53. It has been proposed that A,recep-tors on presynaptic terminals mediate inhibition of neurotransmitter release in CNS tissues2 but there is also evidence that A, receptors are localized postsynap-tically2.Although the functional role of A2receptors has been less apparent, recent work from ligand binding assays has suggested the existence of A2 receptor sub-types15.16.33. In addition to biochemical criteria,adenosine receptor subtypes are often distinguished by their relative affinities for specific agonists53. As a metabolically stable analog of adenosine,NECA was selected as the preferred agonist in this study for several reasons. Firstly,although NECA has been considered to show greater affinity for A2 receptors,it is actually a mixed A,-A2 agonist like adenosine itself'6. Since little is known about the role of adenosine receptor subtypes in the NTS,the approach in the present study was to use an analog, such as NECA, with mixed agonist properties, which thereby more closely resembles the receptor-mediated actions of aden-osine,the naturally-occurring,endogenous nucleoside. Secondly,there is substantial evidence that some aden-osine analogs commonly used for in vivo studies (i.e. N°-cyclohexyladenosine (CHA), No-phenylisopropylade-nosine (PIA),2-chloroadenosine (CADO)) are also potent inhibitors of the adenosine transporter,showing higher affinities for transport sites in some tissues than adenosine itself', and some analogs may even be transported and accumulate intracellularlys'. This is not the case with NECA which exhibits poor affinity for the adenosine transport site18.35. Thus it is likely that the in vivo receptor-mediated effects of NECA are not con-founded by interaction with the nucleoside transporter site.Finally,it has recently been shown that CADO. CHA and PIA may not be good adenosine receptor agonists in vivo since they are phosphorylated into active derivatives by adenosine kinase whereas NECA is a superior receptor agonist in vivo since it is not phosphorylated43. In any event, in the present study,microinjections of NECA into the dorsomedial,medial and ventromedial regions of the caudal NTS. potently depressed respiration and this was accompanied by marked increases in tidal volume. In light of a previous study showing that central injections of NECA were more potent in depressing respiration than R-PIA, an A,-specific ligand,the potent depressant effects of NECA on respiration seen in the present study,point to the possibility that cAMP may function as a second messenger in brainstem respiratory control areas and thereby mediates the modulating effects of a number of neuroactive agents implicated in central respiratory regulation6.8.9.28. This possibility is further supported by recent findings from our laboratory which showed that microinjections of cAMP analogs8, forskolin®, isoproterenol' and adenosinel27 into the same NTS sites,produced similar depressant effects on respiratory frequency as seen in the present study with NECA.On the other hand,although these drugs exerted very similar effects in depressing respiration,overall they elicited somewhat different cardiorespiratory response patterns. For example,although NECA and isoproterenol pro-duced significant increases in V1, cAMP and forskolin caused irregular breathing and periodic apneas*. Also, whereas cAMP elicited a transient hypotension which was then followed by a prolonged bradycardia',isopro-terenol produced bidirectional effects on blood pressure with little effect on heart rate'. Microinjections of adenosine itself into the same NTS sites produced dose-dependent hypotension and bradycardia and these effects were relatively short-acting at the lower doses?. Interestingly,the threshold dose of adenosine (0.05 nmol/kg) required to elicit significant hypotension' is quite comparable to the threshold doses of NECA (0.01 and 0.05 nmol/kg)which elicited hypotension at 2 and 5 min in the present study. The similarity in threshold hypotensive dosages for NECA and adenosine would be expected if, in fact, they acted on similar receptor populations in the NTS. In view of the high density of adenosine uptake sites in the caudal NTS10-12,it is also not surprising that the cardiorespiratory actions of aden-osine(the time for recovery of Pa to control values was 2.44± 0.14 min at the 0.05 nmol/kg dose)were much shorter acting than NECA, a metabolically stable analog. which is not a good substrate for the adenosine transporter18.43. Finally, it is worthwhile to note that the doses of adenosine and NECA required to elicit signifi-cant hypotension and bradycardia following microinjec-tions into the NTS are 1000- to 3000-fold lower than  doses eliciting similar responses following intravenous administration3. On the other hand, the multiplex cardiovascular response pattern elicited by NECA microinjections over the 60 min test period in the present study has not been observed previously with other drugs. Namely,during the initial period after injection,NECA exerted bidirectional effects on blood pressure,eliciting significant hypoten-sion (at lower doses) and pressor responses (at higher doses) in addition to significant decreases in heart rate (at lower doses).Further,the changes in pa during the initial 2-5 min interval were shown to be dose-related and not correlated with depression of R,suggesting,therefore, that the effects on pa during this period were likely mediated by a pharmacologic action of NECA,presum-ably via adenosine receptors. The only other period in which there were significant time-related changes in Pa across all doses was the 30-60 min interval.Indeed,by the end of the 60 min test period, most NECA doses had resulted in significant pressor responses. The hyperten-sion seen at 60 min was not related to variability in experimental conditions (i.e. anesthetic plane, body temperature,etc.)since post hoc comparisons were made with time-matched animals at 60 min following ACF control injections and subjected to identical experimental conditions (i.e.pa(mean ± S.E.M.percent of preinjec-tion control values) for ACF control animals at 60 min= 94.4±0.51).Thus there is no obvious explanation for the NECA-induced hypertension at the end of the test period. The results for heart rate were remarkably similar to the pattern seen with pa. At the lowest doses.NECA elicited significant bradycardia at earlier time points whereas significant tachycardia was produced by the end of the 60 min test period with the higher doses of NECA. Since R, was not significantly correlated with HR at the earlier time points, the bradycardia elicited with the lower doses probably also reflects a pharmacologic action of NECA at adenosine receptors, similar to the brady-cardic effects of adenosine itself.However,as is the case for the increases in pa,there is no obvious explanation for the significant tachycardia associated with the higher doses which is seen by the end of the 60 min test period. In any case, it should be kept in mind that the magnitude for depression of R, approaches an asymptote at 30 min and, as indicated by the correlation statistics, it is unlikely,although still a possibility,that the hypertension and tachycardia observed at 60 min is consequent to depression of Rf. The findings from this study,showing a depression of R,and the multiplex, bidirectional effects on Pa,suggest that the respiratory and cardiovascular effects of NECA microinjections may involve the activation of separate adenosine receptor populations on different neuronal elements in the NTS. This possibility exists since NECA is a mixed A,~A2 receptor agonist16,similar to adenosine itself,and the brainstem possesses both receptor sub-types'5. Indeed,the mixed agonist actions of NECA on A,and A2 receptors have also been shown in adenylate cyclase assays wherein NECA exhibited biphasic effects on the activity of forskolin-stimulated adenylate cyclase activity;inhibition was seen with nanomolar levels of NECA while potent stimulation occurred at micromolar concentrations of NECA12.Interestingly,the highest dose of NECA elicited neither hypotension nor hyper-tension,nor affected heart rate,during the initial portion of the test period. This type of 'saturation' phenomenon at high doses of adenosine analogs has been reported in cultured neuroblastoma cells possessing both A, and A2 receptors wherein the net effect of adenosine agonists was the sum of opposing actions at A,and A2 recep-torss. On the other hand, in contrast to the bidirectional Pa and HR responses, the highest dose of NECA did elicit the largest effect on R, and V, which supports the notion that NECA's effects on respiratory parameters were mediated through actions on a different receptor population than were NECA-induced cardiovascular responses. Thus it is possible that NECA may elicit specific cardiorespiratory response patterns by activating multiple adenosine receptor subtypes on separate neuronal ele-ments in the caudal NTS. Immediately following injec-tion,NECA produced significant dose-related bidirec-tional effects on Pa, pressor and depressor responses, whereas for R, NECA exerted only depressant effects. Multiple adenosine receptor interactions in vivo of this sort have been recently reported in the rat caudate nucleus wherein adenosine analogs exerted postsynaptic actions on A2 receptors which could be distinguished behaviorallyfrom the presynaptic inhibition of dopamine release via A, receptors2". Indeed,data from previous work in our laboratory suggest that the depressant effects of NECA on respiration seen in the present study may be predominantly mediated via adenosine receptors posi-tively coupled to adenylate cyclase,presumably A2 receptors,since microinjections of cAMP analogs,fors-kolin, a direct activator of adenylate cyclase, and the βadrenergic agonist, isoproterenol, potently depressed respiratory frequency following microinjections into the same NTS sites6.8. Therefore,it is conceivable that altering intracellular cAMP levels in intrinsic pools of respiratory neurons in the NTS may influence respiratory frequency and breath-ing patterns by affecting the excitability of the interneu-ronal network and/or other pools of inspiratory neurons under phasic inhibition.In the absence of evidence for respiration-related pacemaker cells in the NTS, respira-  tory rhythmicity likely results from synaptic interactions between pools of respiratory neurons in which waves of excitation are interrupted by periods of inhibition. Moreover,a variety of neurotransmitters,some of which may modulate adenylate cyclase through specific receptor subtypes,are presumably involved in the synaptic inter-actions between respiratory neuronal pools which have shared inputs and local processing by short axon inter-neurons16.36.Although there are few studies directly examining the role of cAMP in NTS-mediated respiratory rhythmicity,there are other lines of evidence implicating a role for cAMP in respiratory periodicity in addition to the findings from our laboratory. For example,baclofen, a selective agonist of GABAB receptors which are inhibitory to adenylate cyclase42,stimulates respiratory frequency and phrenic nerve discharge following admin-istration into the NTS39. It has been suggested y-aminobutyric acid (GABA)-induced increases in the large after hyperpolarization seen following antidromic invasion of spontaneously discharging NTS-neurons may contribute to this acceleration of respiration24. Similarly, neurotensin45 and angiotensin II55,neuropeptides with inhibitory action on adenylate cyclase activity2,excite NTS neurons and also increase phrenic nerve discharge. Finally,it has recently been shown that locus coeruleus neurons are excited by both cyclic AMP analogs and forskolin and that a cAMP-activated inward current likely plays a role in regulating the spontaneous activity of pacemaker cells in some mammalian central neu-rons57. In any event, the respiratory depressant effects observed in the present study may reflect a pharmaco-logic action of NECA at A2 receptors in the NTS and closely resemble the respiratory responses elicited by βadrenergic receptor activation with isoproterenol°. On the other hand, the multiplex cardiovascular responses seen with NECA microinjections in the present study may reflect the greater intrinsic complexity of cardiovascular control elements in the NTS. Whereas the respiratory rhythm generator is relatively localized in the medulla and a eupneic pattern of breathing can even be preserved with transections rostral to the caudal edge of the inferior colliculus,cardiovascular regulation,on the other hand, involves a complex network of connections between the NTS, other brainstem regions (the ventral medulla,vagal complex, parabrachial nucleus)and ro-stral brain structures which provide the basis forfore-brain modulation of central autonomic and humoral activity14. Thus, not only does the NTS receive the primary baroreceptor afferents and projections from many afferents of cardiovascular importance from the viscera25.38,but a variety of forebrain sites involved in cardiovascular regulation also project to the NTS14.17.32. Moreover,the caudal portion of the NTS has projections back to the thalamus,several areas of the hypothalamus. the central amygdala and other forebrain structures 17.32. It follows that a large number of cardiovascular-related elements in the NTS,particularly in the caudal region,consist of axons projecting from the periphery and various rostral brain regions in addition to interneu-rons which provide integration for multiple inputs in-volved in cardiorespiratory homeostasis32. This connec-tivity gives the NTS a unique role for integrating neural and humoral cardiovascular responses to a wide variety of stimuli. The release of numerous neurotransmitters from cen-tral tissues has been shown to be potently modulated by adenosine and its analogs27 and many of these neuroche-micals have been implicated in NTS-mediated mecha-nisms of cardiovascular control32,41. Thus, it is possible that microinjections of NECA, an adenosine analog with mixed agonist properties, into the caudal NTS may exert a variety of presynaptic effects,likely mediated by A1 receptors,thereby modulating neurotransmitter release from the array of nerve terminals making synaptic connections in the caudal NTS-from the periphery, other brainstem sites and rostral brain regions. More-over,the cardiovascular responses elicited by NECA's action on a given set of cardiovascular-related nerve terminals may also depend on whether neurotransmitter release is at basal levels at the time or whether this release is under evoked conditions, as a result of modulatory influences from rostral brain regions. At the same time,NECA may also exert other effects on intrinsic NTS cardiovascular-related neurons via A2 receptors.Thus the mixed agonist properties of NECA may account for the prolonged and multiplex cardiovas-cular responses seen in the present study. Indeed,in view of the possibility that NECA may modulate neurotransmitter function of the vast network of nerve terminals within the NTS,the disruption of the numerous reciprocal projections from rostral brain re-gions involved in cardiovascular control could conceiv-ably affect the pattern of cardiovascular responses in-duced by NECA microinjections in the NTS. In fact,we have recently examined this possibility using an acute decerebrate preparation to disrupt rostral brain projec-tions and under the same experimental conditions used in the present study (manuscript in preparation).The preliminary data show that microinjections of NECA into the same NTS sites of a decerebrate rat prouduced a remarkably similar response pattern for the respiratory parameters as that seen in the present study.In contrast, microinjections of NECA in the decerebrate rat pro-duced a qualitatively different cardiovascular response pattern from that seen in the intact animal. In any event, these studies comparing cardiorespiratory response pat-  terns in intact and decerebrate rats are currently in progress as part of an effort to further examine cardio-respiratory responses following microinjections into the same NTS sites of selective agonists and antagonists for adenosine receptor subtypes.However,for the studies using selective A, and A2 ligands, the major method-ological obstacle has been the poor aqueous solubility of many of these drugs for use in microinjection studies. In summary, the results of this study demonstrate the potent effects of NECA on respiratory rate and tidal volume following microinjections into the caudal NTS. The data suggest that neuronal elements in the NTS involved in respiratory rhythmicity do not receive the multitude of presynaptic projections, especially from rostral brain regions, nor do they profusely project back to these sites,as is the case with the longer 'control loops' involved in central cardiovascular regulation.The data further suggest that once NECA influences NTS neuronal elements involved in respiratory rhythmicity,presumably by acting at A2 receptors, the depressant effects on REFERENCES 1 Barraco,R.A., Phillis,J.W., Campbell, W.R.,Marcantonio, D.R.and Salah.R.S.,The effects of central injections of adenosine analogs on blood pressure and heart rate in the rat, Neuropharmacology, 25 (1986) 675-680. 2 Barraco,R.,Polasek,P.M.,Janusz,C.J.,Campbell, W.R. and Schoener,E.P.,Parameters affecting microinjections of adeno-sine into the brainstem of rats,Med. Sci. Res., 15 (1987) 797-799. 3 Barraco,R.A.,Marcantonio,D.R.,Phillis,J.W.and Cammpbell, W.R.,The effects of parenteral injections of adenosine and its analogs on blood pressure and heart rate in the rat,Gen. Pharmac., 18(1987)405-416. 4 Barraco, R.A.,Polasek, P.M., Janusz, C.J. and Parizon, M.. Cardiovascular effects of microinjections of cyclic AMP into the nucleus tractus solitarius of rats, Med. Sci. Res., 15 (1987) 1123-1124. 5 Barraco,R.A.,Campbell,W.R.,Schoener,E.P.,Shehin,S.E. and Parizon,M.,Cardiovascular effects of microinjections of adenosine analogs into the fourth ventricle of rats, Brain Research,424(1987)17-25. 6 Barraco,R.A. and Janusz,C.A.,Microinjections of isoproter-enol into the nucleus tractus solitarius modulates respiration, Med. Sci. Res., 16 (1988) 977-979. 7 Barraco.R.A.,Janusz.C.A.,Polasek,P.M.,Parizon,M.and Roberts,P.A.,Cardiovascular effects of microinjection of adenosine into the nucleus tractus solitarius,Brain Res.Bull.,20 (1988)129-132. 8 Barraco.R.A.,Schoener,E.P.,Polasek,P.M.,Simpson,L.L., Janusz,C.J.and Parizon,M.,Respiratory effects of cyclic AMP following microinjections into the nucleus tractus solitarius of rats, Neuropharmacology, 27 (1988) 1285-1294. 9 Barraco,R.A. and Janusz,C.A.,Respiratory effects of 5-ethylcarboxamidoadenosine,an analog of adenosine,following microinjections into the nucleus tractus solitarius of rats,Brain Research,480(1989)360-364. 10 Bisserbe,J.C..Patel.J. and Marangos, P.J., Autoradiographic localization of adenosine uptake sites in rat brain using ['H)-nitrobenzylthioinosine,J. Neurosci., 5 (1985) 544-550. 11 Bisserbe,J.C..Deckert, J. and Marangos, P.J., Autoradio-graphic localization of adenosine uptake sites in guinea pig brain respiratory rate persist for a considerable time. In contrast to the respiratory parameters,the data from this study suggest that NECA exerts multiplex pharmacolog-ical effects on neuronal elements in the NTS subserving cardiovascular regulation and these effects may involve activation of both A,and A2 receptors. Finally, the data showing NECA's potent effects on respiratory param-eters provide convincing support for the contention that adenosine plays a neuromodulatory role in the NTS and further suggest that adenosine,released within some NTS regions during pathophysiologic conditions such as hyp-oxia,may have profound consequences for cardiorespi-ratory function. Acknowledgements. The authors gratefully acknowledge the technical assistance of P. Polasek and the meritorious efforts of L. McCraw in preparing the manuscript.The authors also are grateful to M.Kruger for help in statistical analysis.The work was supported by grants from the National Science Foundation (R1186-04084), NIH (RR-08167-10) and the American Heart Association of Michigan. using ['H]dipyridamole,Neurosci.Lett., 66 (1986)341-345. 12 Blazynski,C.,Adenosine A,receptor-mediated inhibition of adenylate cyclase in rabbit retina, J. Neurosci., 7 (1987) 2522-2528. 13 Bohm,M.,Bruckner,R.,Neumann,J.,Schmitz,W.,Scholz,H. and Starbatty,J.,Role of guanine nucleotide-binding protein in the regulation by adenosine of cardiac potassium conductance and force of contraction.Evaluation with pertussis toxin, Naunyn-Schmiedeberg's Arch. Pharmacol., 332 (1986) 403-405. 14 Brody,M.J..Alper, R.H., O'Neill, T.P. and Porter, J.P.. Central neural control of the cardiovascular system.In A. Zanchetti and R.C.Tarazi (Eds.). Handbook of Hypertension, Vol.8:Pathophysiology of Hypertension-Regulatory Mecha-nisms,Elsevier,New York, 1986, pp. 1-25. 15 Bruns,R.F.,Lu,G.H. and Pugsley,T.A.,Characterization of the A2 adenosine receptor labeled by ['HJNECA in rat striatal membranes,Mol. Pharmacol., 29 (1986) 331-346. 16 Bruns, A.F.,Adenosine receptor binding assays. In D.M.F. Cooper and C. Londos (Eds.),Receptor Biochemistry and Methodology,Vol. I1:Adenosine Receptors,Liss,New York, 1988.pp. 43-62. 17 Bystrzycka,E.K.and Nail,B.S.,Brainstem nuclei associated with respiratory,cardiovascular and other autonomic fucntions. In G.Paxinos (Ed.).The Rat Nervous System,Academic Press, New York,1985,pp.95-110. 18 Clanachan,A.S.,Heaton,T.P. and Parkinson,F.E.,Drug interactions with nucleoside transport systems. In E.Gerlach and B.F. Beckers (Eds.),Topics and Perspectives in Adenosine Research,Springer,Berlin, 1987.pp.118-130. 19 Cohen,M.T.,Neurogenesis of respiratory rhythm in the mam-mal,Physiol.Rev.,59 (1979) 1105-1173. 20 Criswell,H.,Mueller,R.A.and Breese.G.R.,Assessment of purine-dopamine interactions in 6-hydroxydopamine-lesioned rats:evidence for pre-and postsynaptic influences by adenosine. J. Pharmacol.Exp. Ther., 244(1988)493-500. 21 Darnall, R.A.,Aminophylline reduces hypoxic ventilatory depression:possible role of adenosine.Pediat. Res., 19 (1985) 706-710. 22 Deckert,J.,Bisserbe,J.C.and Marangos,P.J.,Quantitative ['H]dipyridamole autoradiography: evidence for adenosine transporter heterogeneity in guinea pig brain. Naunyn-Schmie-deberg's Arch. Pharmacol.. 335 (1987) 660-666.  23 Deckert.J. and Jorgensen,M.B.,Evidence for pre-and postsynaptic localization of adenosine A,receptors in the CA1 region of rat hippocampus: a quantitative autoradiographic study, Brain Research, 446 (1988) 161-164. 24 DeFeudis,F.V.,GABA and respiratory function,Gen. Phar-macol.,15(1984)441-444. 25 Donoghue,S., Felder,R.B., Jordan, D. and Spyer,K.M.,The central projection of carotid baroreceptors and chemoreceptors in the cat: a neurophysiological study. J. Physiol. (Lond.),347 (1984)397-409. 26 Dowdy,S.and Wearden, S.,Statistics for Research,1983. Wiley,New York,1983,pp.201-242. 27 Dunwiddie,T.V.,The physiological role of adenosine in the central nervous system,Int.Rev.Neurobiol., 27 (1985) 63-139. 28 Eldridge,F.L. and Millhorn,D.E.,Central regulation of respiratory by endogenous neurotransmitters and neuromodu-lators,Annu. Rev. Physiol., 43 (1981) 121-135. 29 Eldridge,F.L.,Millhorn,D.E. and Kiley,J.P.,Respiratory effects of a long-acting analog of adenosine,Brain Research,301 (1984)273-280. 30 Eldridge, F.L.,Millhorn,D.E. and Kiley, J.P., Antagonism by theophylline of respiratory inhibition induced by adenosine,J. Appl. Physiol., 59 (1985) 1428-1433. 31 Euler,C. von, On the central pattern generator for the basic breathing rhythmicity, J. Appl. Phyciol.:Respirat.Environ. Exercise Physiol., 55 (1983) 1647-1659. 32 Feldman,J.L.and Ellenberger,H.H.,Central coordination of respiratory and cardiovascular control in mammals, Annu.Rev. Physiol.,50(1988)593-606. 33 Florio,C.,Traversa,U.,Vertue,R. and Puppini, P., 5′-N-ethylcarboxamidoadenosine binds to two different adenosine receptors in membranes from the cerebral cortex of the rat. Neuropharmacology,27 (1988) 85-94.
34 Gallman,E.A.and Millhorn,D.E.,Adenosine involvement in a long-lasting brainstem-mediated inhibition of phrenic activity following hypoxia,Soc.Neurosci.Abstr., 14(1988) 1182.
35 Geiger,J.D.,Johnston,M.E.and Yago,V.,Pharmacological characterization of rapidly accumulated adenosine by dissociated brain cells from adult rat, J. Neurochem., 51 (1988) 283-291.
36 Grelot,L.,Iscoe,S.and Bianchi,A.L.,Effects of amino acids on the excitability of respiratory bulbospinal neurons in solitary and paraambiqual regions of medulla in cat, Brain Research, 443 (1988)27-36.
37 Jiang,C.,Gao,L.I.,Shen,E.H.and Wei,J.Y.,Respiration related neurons in the region of the nucleus tractus solitarius of the rabbit, Brain Research, 377 (1986) 190-193.
38 Jordan,D.and Spyer,K.M., Brainstem integration of cardio-vascular and pulmonary afferent activity.In F.Cervero and J.F.B.Morrison(Eds.),Progress in Brain Research,Vol.67. Elsevier, New York, 1986, pp. 295-314.
39 Lalley,P.M.,Effects of baclofen and y-aminobutyric acid on different types of medullary respiratory neurons. Brain Re-search, 376(1986) 392-395.
40 Lagercrantz,H.,Fredholm,B.B.,Irestedt,L.,Runold,M.and Sollevi,A.,Adenosine-a neuromodulator released during asphyxia and a mediator of some hypoxic effects in the newborn?,Cardiovasc.Resp.Physiol.Fetus Neonate,433(1985) 153-160.
41 Leslie,R.A.. Neuroactive substances in the dorsal vagal com-

plex of the medulla oblongata:nucleus of the tractus solitarius, area postrema, and dorsal motor nucleus of the vagus.Neuro-chem.Int., 7(1985)191-211.
42 Limbird. L.E., Receptors linked to inhibition of adenylate cyclase:additional signaling mechanisms,Fed. Proc., 2(1988) 2686-2695.
43 Lin,B.B.,Hurley,M.C.and Fox,I.H.,Regulation of adenosine kinase by adenosine analogs, Mol. Pharmacol.,34(1988) 501-505.
44 Long,S.and Duffin,J.,Theneuronal determinants of respira-tory rhythm,Prog. Neurobiol., 27 (1986) 101-182.
45 Morin-Surdun,M.P.,Marlot,D.,Kessler,J.P.and Denavit-Saubie,M.,The excitation by neurotensin of nucleus tractus solitarius neurons induces apneustic breathing,Brain Research, 384(1986)106-113.
46 Moss,I.R.,Denavit-Saubie,M.,Eldridge,F.L., Gillis,R.A., Herkenham,M.and Lahiri, S., Neuromodulators and transmit-ters in respiratory control, Fed. Proc., 45 (1986)2133-2147.
47 Onai,T.and Miura,M.,Projections of supraspinal structures to the phrenic motor nucleus in cats studied by a horseradish peroxidase microinjection method, J. Auton. Nerv. Syst., 16 (1986)61-77.
48 Onai,T.,Saji,M.and Miura, M., Functional subdivisions of the nucleus tractus solitarii of the rat as determined by circulatory and respiratory responses to electrical stimulation of the nucleus, J.Auton. Nerv. Syst., 21 (1987) 195-202.
49 Onai, T., Takayama, K. and Miura, M., Projections to areas of the nucleus tractus solitarii related to circulatory and respiratory responses in cats, J. Auton Nerv. Syst.. 18 (1987) 163-175.
50 Palmour,R.M.and Ervin,F.R., Adenosine receptor subtypes: toward an understanding of in vivo effects,Soc. Neuroci. Abstr., 13(1987)1488.
51 Perez,M.T.R.and Ehinger,B.,Adenosine uptake and release in the rabbit retina. In E. Agardh and B. Ehinger (Eds.),Retinal Signal Systems,Degenerations and Transplants,Elsevier,New York,1986, pp. 163-172.
52 Phillis,J.W.,Walter,G.A.,O’Regan,M.H.and Stair,R.E., Increases in cerebral cortical perfusate adenosine and inosine concentrations during hypoxia and ischemia, J. Cereb. Blood Flow Metab., 7 (1987)679-686.
53 Ribeiro,J.A.and Sebastiao,A.M.,Adenosine receptors and calcium: basis for proposing a third (A3)adenosine receptor, Progr. Neurobiol., 26(1986)179-209.
54 Runold,M.,Lagercrantz,H.and Fredholm,B.B.,Ventilatory effect of an adenosine analogue in unanesthetized rabbits during development, J. Appl. Physiol., 61 (1986) 255-259.
55 Sessle,B.J.and Henry, J.L.,Angiotensin II excites neurons in cat solitary tract nuclei which are involved in respiration and related reflex activities,Brain Research, 407 (1987) 163-167.
56 Tseng,C.,Biaggioni,I.,Appalsamy, M. and Robertson, D.. Purinergic receptors in the brainstem mediate hypotension and bradycardia,Hypertension,11(1988) 191-197.
57 Wang,Y.Y.and Aghajanian, G.K.,Excitation of locus coeruleus neurons by an adenosine 3′,5′-cyclic monophosphate-activated inward current: extracellular and intracellular studies in rat brain slices,Synapse, 1 (1987) 481-487.
58 Wessberg,P.,Hedner,J.,Hedner,T.,Persson,B.and Jonason, J..Adenosine mechanisms in the regulation of breathing in the rat,Eur. J. Pharmacol., 106 (1985) 59-67.