Correlation of vasomotor- and respiratory-controlling mechanisms around the caudal ventrolateral medulla in cats

Neuroscience Letters 269 (1999) 79±82

Correlation of vasomotor- and respiratory-controlling mechanisms around the caudal ventrolateral medulla in cats Kun-Ta Yang a, b, Chun-Kuei Su a, Chok-Yung Chai a,* a

Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan 115, ROC b National Defense Medical Center, Taipei, Taiwan 100, ROC

Received 1 April 1999; received in revised form 6 May 1999; accepted 6 May 1999

Abstract We examined the involvement of caudal ventrolateral medulla (CVLM) in respiratory control. Microinjection of glutamate (Glu) into CVLM decreased systemic arterial blood pressure (SAP) and altered phrenic nerve activities (PNA). Among 143 depressor sites, 55% (78/143) increased respiratory frequency (Rf), while 72% altered PNA amplitude (36% increased and 36% decreased). A small but signi®cant positive correlation was observed between the magnitudes of depressor responses and inhibition of PNA amplitude (r ˆ 0:1718, n ˆ 143), indicating a substantial cross-talk between depressor and PNA inhibitory neurons. Furthermore, microinjections of acetylcholine (ACh) mimicked the Glu-induced depressor responses. However, ACh did not alter Rf, but still reduced PNA amplitude. Our ®ndings suggest that Rfregulating and depressor neurons are two separate neuronal populations, coexisting in CVLM. The PNA inhibitory and depressor neurons, in contrast, could have stronger correlation. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Cardiorespiratory integration; Depressor neurons; Ventral respiratory group; Pre-BoÈtzinger complex; Acetylcholine

The caudal ventrolateral medulla (CVLM) indirectly receives glutaminergic baroreceptor inputs and regulates vasomotor tone [4,5,7,19,20]. Via their inhibitory axonal projections to the sympathoexcitatory neurons in the brainstem or directly to the sympathetic preganglionic neurons, CVLM neurons tonically inhibit sympathetic nerve activities (SNA) and depress systemic arterial blood pressure (SAP) [1,2,8,18]. Anatomically, CVLM overlaps with the ventral respiratory group, which contains the most crucial respiratory-controlling mechanism, the pre-BoÈtzinger complex (pre-BoÈtC) [15]. In both rats and cats, the preBoÈtC is critical for respiratory rhythmogenesis [3,16,17]. In brain slice preparations, a blockade of non-NMDA (Nmethyl-d-aspartate) receptors in pre-BoÈtC reduces the frequency and ultimately abolishes respiratory network oscillations [6]. Similar ®ndings have also been observed in in vivo studies. After kainic acid lesion or a broad-spectrum blockade of glutaminergic receptors in this area, the respiratory rhythmogenesis is severely disturbed [9,10]. Our attempt here is to unravel the correlation of neuronal populations located in the CVLM underlying vasomotor- and respiratory-controlling mechanisms. Using microinjection * Corresponding author. Tel.: 1886-2-2789-9105; fax: 1886-22782-9224.

techniques, we demonstrated that glutamate (Glu) stimulation of CVLM neurons decreased SAP and elicited respiratory responses. Some of these vasomotor/respiratory responses could be differentiated by a further challenge with acetylcholine (ACh) stimulation. As described previously [9], cats were anaesthetized with a-chloralose (40 mg/kg) and urethane (400 mg/kg). Whole bundle phrenic nerve activities (C5 or C6) and SAP were recorded. A two-barrel micropipette (outside tip diameter: ~30 mm) was used for microinjection. One barrel was ®lled with Glu (100 mM) containing 1% pontamine sky blue, the other with either 0.9% NaCl (for exploring depressor sites by electrical stimulation) or ACh (100 mM). Areas of CVLM were explored within: 0±2mm rostral to obex, 3.5± 4.5 mm lateral to the midline, and 2±4.5 mm ventral to the horizontal plane at the obex. In some experiments, the depressor sites were ®rstly identi®ed by Glu and followed by microinjecting ACh into the same site (at least 20 min after Glu stimulation). Frozen sections of the brain in 50 mm thickness were used to con®rm the chemical injection sites. Data were analyzed using Student's t-test. A P-value , 0.05 was considered signi®cant. Microinjecting Glu (100 mM, 30 nl) into discrete regions of CVLM elicited a decrease of SAP, mostly, concomitant with a change of PNA (Fig. 1). Among 143 depressor sites

0304-3940/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S03 04 - 394 0( 9 9) 00 42 8- 0

80

K.-T. Yang et al. / Neuroscience Letters 269 (1999) 79±82

Fig. 1. Effects of stimulating the CVLM depressor (A,B) or non-depressor sites (C,D) on SAP and PNA. (A) Glu stimulation decreased SAP while increased Rf. ACh produced a comparable depressor response, but failed to alter Rf. The stimulated sites was indicated by the ®lled circle (middle). Data from the same type of responses were pooled (right). (B) Glu stimulation decreased SAP while abolished PNA for more than 30 s. Note the immediate reduction of PNA amplitude after Glu, indicating an effect due to the inhibition of PNA amplitude rather than its rhythmogenesis. ACh, although less potent than Glu, still decreased SAP and PNA amplitude. (C) Glu did not change SAP but still increased Rf (compared with (A)). ACh did not elicit any respiratory responses. (D) Glu did not change SAP but still reduced PNA amplitude (compared with (B)). ACh was also effective to reduce PNA amplitude, but did not consistently decrease SAP (see the statistical comparisons in the right panel). The symbols in the right panels indicate a signi®cant difference (P , 0:05) of the changes induced by Glu (asterisks) or that between Glu- and ACh-induced responses (cross).

(SAP decrease .10%), 89% (127/143) also altered either the rhythmic discharge or the amplitude of PNA, or both. A higher incidence of the depressor responses (55%, 78/143) concurred with increasing respiratory frequency (Rf). Only 10% (14/143) revealed a decrease of Rf, while 35% (51/ 143) were unchanged. However, the amplitude of PNA was either increased (36%, 52/143), decreased (36%, 51/ 143) or unchanged (28%, 40/143). Fig. 2 shows the histological distribution of these depressor sites. In relation to their respiratory responses, most Rf-increased sites (78%) were distributed in the vicinity of nucleus ambiguus at the frontal plane 1 mm rostral to the obex. In the rostral CVLM (2 mm rostral to the obex), Glu stimulation was more likely

to reduce PNA amplitude (67%). On the basis of these more frequent and focalized Glu-induced responses, we compared the effects induced by microinjecting Glu and ACh into the Rf-increased and PNA amplitude-decreased sites. Microinjecting the same amount of Glu or ACh (3 nmol) into the CVLM elicited comparable depressor responses (n ˆ 28, Fig. 1A,B). However, the effects on respiratory responses were different. Unlike Glu, ACh did not affect Rf (n ˆ 18, Fig. 1A). ACh, although less potent than Glu, still signi®cantly reduced PNA amplitude (n ˆ 10, Fig. 1B). Interestingly, another 21 stimulated sites did not respond to microinjection of Glu in decreasing SAP but succeeded in increasing Rf (Fig. 1C, n ˆ 18) or decreasing PNA ampli-

K.-T. Yang et al. / Neuroscience Letters 269 (1999) 79±82

Fig. 2. Histological distribution of Glu-induced CVLM depressor sites in relation to their respiratory responses. In each frontal section, the sites related to the responses of Rf and PNA amplitude were reconstructed on the left and right side, respectively. Symbols represent an increase (®lled triangle), decrease (open triangle), or no response (circle). The lower right panel shows the incidence ratio of Glu-induced respiratory responses at each level rostral to the obex. Note the highest incidence ratio of Rfincreased sites at A1 (78%). The incidence of PNA amplitudedecreased sites had a monotonic increase along the levels and reached the highest ratio at A2 (67%).

tude (Fig. 1D, n ˆ 3). In these CVLM non-depressor sites, ACh evoked similar responses as those from the depressor sites, i.e. not affecting Rf but still reducing PNA amplitude (Fig. 1C,D). In an attempt to elucidate the quantitative relationship between depressor and respiratory responses, we further examined the magnitudes of changes in these cardiorespiratory parameters by correlation analysis. Fig. 3 shows a small but signi®cant linear correlation between changes of SAP and PNA amplitude (r ˆ 0:1718, n ˆ 143). However, the correlation between changes of SAP and Rf induced by Glu stimulation was not apparent (r ˆ 0:0387, n ˆ 143). Our results indicate a coexistence of vasomotor- and respiratory-controlling neural mechanisms in the CVLM. Intriguingly, Glu-induced increase of Rf could not be replicated by ACh stimulation, which otherwise mimicked the Glu-induced responses in reducing PNA amplitude. The magnitudes of Glu-induced changes between SAP and PNA amplitude also revealed a positive correlation. These ®ndings suggest differential mechanisms in the CVLM controlling respiratory rhythmogenesis and amplitude, whereas the neural mechanism underlying the modulation of inspiratory amplitude has a closer relation to that regulating vasomotor functions. We used microinjection techniques to probe the func-

81

tional subsets of CVLM neurons underlying vasomotor and respiratory regulation. Using a two-barrel micropipette, we assume the effective scopes resulting from the Glu and ACh stimulation are similar, i.e. both stimulations would perturb the same subset of neurons near the microinjection sites. Thus, the observation that ACh stimulation failed to mimic the Glu-induced responses would allow us to segregate the functional difference within the subset of neurons by their pharmacological characteristics. Endogenous release of excitatory amino acid to the preBoÈtC is essential for respiratory rhythmogenesis [6]. Using exogenous application of Glu, we clearly demonstrated that the CVLM contains subsets of neurons that can increase Rf. These neurons could be the pre-BoÈtC neurons responsible for respiratory rhythmogenesis. Most of these neurons were distributed in the vicinity of nucleus ambiguus and about 1 mm caudal to the sites where most of the Glu stimulations reduced the PNA amplitude (Fig. 2), which presumably resulted from a direct inhibition of phrenic motoneurons by stimulating the BoÈtzinger neurons [11]. Thus, both our histological and functional data indicate that these Gluinduced Rf-increased sites are equivalent to the pre-BoÈtC in cats [3,16]. Activation of central cholinergic receptors is believed to accelerate Rf [13,14]. In our studies, ACh stimulation did not replicate the Glu-induced responses with respect to accelerating Rf, even though a higher proportion of Rfincreased sites concomitant with depressor responses was located in the pre-BoÈtC. This observation indicates that Rfregulating and depressor neurons are co-localized in the CVLM and have distinct pharmacological characteristics.

Fig. 3. Correlation analysis of the magnitudes of changes between depressor and respiratory responses. There is a significant positive correlation of the changes between SAP and PNA amplitude (solid circles and line; y ˆ 0:4143x 1 114:49, r ˆ 0:1718), while the correlation of changes between SAP and Rf is not signi®cant (open squares and dashed line; y ˆ 0:0832x 1 119:02, r ˆ 0:0387).

82

K.-T. Yang et al. / Neuroscience Letters 269 (1999) 79±82

Unlike depressor neurons, Rf-regulating neurons do not have cholinergic receptors. In contrast, both Glu and ACh stimulation elicit similar responses in reducing SAP and PNA amplitude, suggesting that the effects are exerted by the same neuronal population. Alternatively, the effects may be due to an integration of distinct neuronal populations, which are linked by either direct or indirect neural connections. This argument could be further supported by our observation that there indeed exist a signi®cant positive correlation between the responses of SAP and PNA amplitude (Fig. 3). Cardiorespiratory integration is manifested by the respiratory-related ¯uctuation of SNA. Blocking nonNMDA receptors in the CVLM leads to an elevation of postinspiratory SNA [12]. Thus, glutaminergic inputs to CVLM neurons could modulate the coupling between SNA and central respiratory drive. In combination with our ®ndings that depressor and PNA amplitude-regulating neurons could be the same or correlated neuronal populations. It is reasonable to speculate that some CVLM neurons with cholinergic receptors may serve as a bridge for the cross-talk between vasomotor- and respiratory-controlling neural mechanisms. We are grateful to Dr. M. Seah for comments on this manuscript. This work was supported by grants from National Science Council of ROC (NSC 88-2314-B-001002) and Shih-Chun Wang Research Memorial Fund. [1] Agarwal, S.K., Gelsema, A.J. and Calaresu, F.R., Inhibition of rostral VLM by baroreceptor activation is relayed through caudal VLM. Am. J. Physiol., 258 (1990) R1271± R1278. [2] Blessing, W.W., Depressor neurons in rabbit caudal medulla act via GABA receptors in rostral medulla. Am. J. Physiol., 254 (1988) H686±H692. [3] Connelly, C.A., Dobbins, E.G. and Feldman, J.L., Pre-BoÈtzinger complex in cats: respiratory neuronal discharge patterns. Brain Res., 590 (1992) 337±340. [4] Cravo, S.L. and Morrison, S.F., The caudal ventrolateral medulla is a source of tonic sympathoinhibition. Brain Res., 621 (1993) 133±136. [5] Cravo, S.L., Morrison, S.F. and Reis, D.J., Differentiation of two cardiovascular regions within caudal ventrolateral medulla. Am. J. Physiol., 261 (1991) R985±R994. [6] Funk, G.D., Smith, J.C. and Feldman, J.L., Generation and transmission of respiratory oscillations in medullary slices:

[7] [8] [9]

[10]

[11] [12]

[13] [14]

[15]

[16] [17]

[18]

[19]

[20]

role of excitatory amino acids. J. Neurophysiol., 70 (1993) 1497±1515. Gordon, F.J., Aortic baroreceptor re¯exes are mediated by NMDA receptors in caudal ventrolateral medulla. Am. J. Physiol., 252 (1987) R628±R633. Hardy, S.G., Horecky, J.G. and Presley, K.G., Projections of the caudal ventrolateral medulla to the thoracic spinal cord in the rat. Anat. Rec., 250 (1998) 95±102. Hsieh, J.H., Chang, Y.C., Su, C.K., Hwang, J.C., Yen, C.T. and Chai, C.Y., A single minute lesion around the ventral respiratory group in medulla produces fatal apnea in cats. J. Auton. Nerv. Sys., 73 (1998) 7±18. Jung, R., Bruce, E.N. and Katona, P.G., Cardiorespiratory responses to glutamatergic antagonists in the caudal ventrolateral medulla of rats. Brain Res., 564 (1991) 286± 295. Merrill, E.G. and Fedorko, L., Monosynaptic inhibition of phrenic motoneurons: a long descending projection from BoÈtzinger neurons. J. Neurosci., 4 (1984) 2350±2353. Miyawaki, T., Minson, J., Arnolda, L., Chalmers, J., Llewellyn-Smith, I. and Pilowsky, P., Role of excitatory amino acid receptors in cardiorespiratory coupling in ventrolateral medulla. Am. J. Physiol., 271 (1996) R1221±R1230. Monteau, R., Morin, D. and Hilaire, G., Acetylcholine and central chemosensitivity: in vitro study in the newborn rat. Respir. Physiol., 81 (1990) 241±253. Murakoshi, T., Suzue, T. and Tamai, S., A pharmacological study on respiratory rhythm in the isolated brainstemspinal cord preparation of the newborn rat. Br. J. Pharmacol., 86 (1985) 95±104. Rekling, J.C. and Feldman, J.L., PreBoÈtzinger complex and pacemaker neurons: hypothesized site and kernel for respiratory rhythm generation. Ann. Rev. Physiol., 60 (1998) 385±405. Schwarzacher, S.W., Smith, J.C. and Richter, D.W., PreBoÈtzinger complex in the cat. J. Neurophysiol., 73 (1995) 1452±1461. Smith, J.C., Ellenberger, H.H., Ballanyi, K., Richter, D.W. and Feldman, J.L., Pre-BoÈtzinger complex: a brainstem region that may generate respiratory rhythm in mammals. Science, 254 (1991) 726±729. Stocker, S.D., Steinbacher, B.C.J., Balaban, C.D. and Yates, B.J., Connections of the caudal ventrolateral medullary reticular formation in the cat brainstem. Exp. Brain Res., 116 (1997) 270±282. Willette, R., Punnen, S., Krieger, A.J. and Sapru, H.N., Interdependence of rostral and caudal ventrolateral medullary areas in the control of blood pressure. Brain Res., 321 (1984) 169±174. Yu, D. and Gordon, F.J., Anatomical evidence for a bineuronal pathway connecting the nucleus tractus solitarius to caudal ventrolateral medulla to rostral ventrolateral medulla in the rat. Neurosci. Lett., 205 (1996) 21±24.