Carotid chemoafferent plasticity in adult rats following developmental hyperoxia

Respiratory Physiology & Neurobiology 145 (2005) 3–11

Carotid chemoafferent plasticity in adult rats following developmental hyperoxia Gerald E. Bisgarda,∗ , E. Burt Olson Jr.b , Ryan W. Bavisc , Julie Wenningera , Erik V. Nordheimd , Gordon S. Mitchella a

Department of Comparative Biosciences, University of Wisconsin, School of Veterinary Medicine, 2015 Linden Drive, Madison, WI 53706, USA b Department of Population Health Sciences, University of Wisconsin-Madison, Madison, WI 53706, USA c Department of Biology, Bates College, Lewiston, ME 04240, USA d Department of Statistics, University of Wisconsin-Madison, Madison, WI 53706, USA Accepted 25 October 2004

Abstract Developmental hyperoxia impairs carotid chemoreceptor development and induces long-lasting reduction in carotid sinus nerve (CSN) responses to hypoxia in adult rats. Studies were carried out to determine if CSN responses to acute hypoxia would exhibit hypoxia-induced plasticity in adult 3–5-months-old rats previously treated with postnatal hyperoxia (60% O2 , PNH) of 1, 2, or 4 weeks duration. CSN responses to acute hypoxia were assessed in adult rats exposed to 1 week of sustained hypoxia (12% O2 , SH). In normal adult rats and adult rats treated with 1 week of PNH, CSN responses to acute hypoxia were significantly increased in urethane-anesthetized rats when studied 3–5 h after SH. Apparent increases in CSN responses to hypoxia were not significant in rats treated with 2 weeks of PNH and were clearly absent after 4 weeks of PNH, but exponential analysis suggests a PNH durationdependent plasticity of the CSN response to acute hypoxia after SH. In a second study rats exposed to 2 weeks of PNH were treated with SH for 1 week as adults and acute hypoxic responses were tested 4–5 months later. CSN responses in these rats were unaffected by SH suggesting a lack of persistent SH-induced functional plasticity. We conclude that rats treated with 1 week of PNH retain the capacity for hypoxia-induced plasticity of carotid chemoafferent function and some potential for plasticity may be present after 2 weeks of PNH, whereas 4 weeks of PNH impairs the capability of rats to exhibit plasticity following 1 week of SH. © 2004 Elsevier B.V. All rights reserved. Keywords: Carotid sinus nerve; Carotid body; Hypoxia; Control of breathing; Functional impairment; Developmental plasticity; Ventilatory acclimatization

∗

Corresponding author. Tel.: +1 608 263 4649; fax: +1 608 263 3926. E-mail address: [email protected] (G.E. Bisgard). 1569-9048/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.resp.2004.10.008

1. Introduction One to four weeks of postnatal hyperoxia (30–60% O2 ) (PNH) significantly impairs ventilatory and

4

G.E. Bisgard et al. / Respiratory Physiology & Neurobiology 145 (2005) 3–11

phrenic nerve responses to acute hypoxia in adult rats (Bavis et al., 2002, 2003; Fuller et al., 2002; Ling et al., 1997). The diminished hypoxic response does not involve CNS mechanisms, but results from attenuated chemoafferent inputs from the carotid body (CB) and chemoafferent neurons in the carotid sinus nerve (CSN) (Bisgard et al., 2003; Fuller et al., 2002; Ling et al., 1997). The adult CB is greatly reduced in size after postnatal hyperoxia (Bisgard et al., 2003; Erickson et al., 1998; Fuller et al., 2002) and there are approximately 40% fewer unmyelinated afferent axons in the CSN when measured immediately after 4 weeks of PNH (Erickson et al., 1998). Newborn rats in the first 2 weeks of life are most susceptible to hyperoxia-induced attenuation of CB/CSN development (Bavis et al., 2002; Bisgard et al., 2003). Structural changes may play an important role in the deficit of CB chemoafferent function since there are fewer CB type 1 cells and CSN afferent fibers (Erickson et al., 1998; Prieto-Lloret et al., 2003). Evidence that CB type 1 cell function is also compromised includes failure of some cells to increase intracellular [Ca2+ ] in response to hypoxia and abnormal catecholamine metabolism and release (Prieto-Lloret et al., 2003). This group found that CSN afferent responses to hypoxia were not detected from most CB/CSN preparations, but those that did respond had responses similar to normal animals. On the other hand, Carroll et al. (2003) reported impaired discharge frequency in responses to hypoxia in single afferent fibers from rat pups that had been treated with hyperoxia. We postulate that there is a general failure of CB/CSN development due to lack of sensory (hypoxic) stimulation during a critical developmental period within the first two weeks of life. In adult rats the CB undergoes significant structural and functional plasticity after exposure to sustained hypoxia (SH). For example, SH induces increased gene expression, cellular hyperplasia, vasodilation, increased CB volume, increased tyrosine hydroxylase activity and, increased catecholamine metabolism (Wang and Bisgard, 2002). These changes may play a role in increased CB hypoxic chemosensitivity after SH (Bisgard, 2000). Some of these forms of SH-induced CB plasticity were examined in adult rats subjected to PNH and evidence of similar plasticity was observed (vasodilation and increased CB volume) (Fuller et al., 2001). Furthermore, in adult rats subjected to 4 weeks of PNH, SH increased integrated phrenic nerve re-

sponses to acute hypoxia immediately after SH (Fuller et al., 2001). Though it was postulated this was due to SH-induced CB plasticity, a central neural mechanism that increases the response to acute hypoxia could not be ruled out. Because of the strong possibility that the Fuller et al. findings could be explained by SH-CB plasticity, we predicted that PNH rats would exhibit recovered function in CSN recordings immediately after SH exposure. Because of this expectation, we also explored the potential for long-term (i.e. persistent) CB functional recovery. Therefore, in the present study, our purpose was to examine the CSN sensory responses to acute hypoxia after SH in adult rats previously subjected to PNH. We hypothesized that enhanced carotid chemoafferent function could explain the increase in hypoxic phrenic responses observed by Fuller et al. (2001). Two studies were completed in adult rats exposed to PNH: (1) carotid chemoafferent responses to acute hypoxia immediately after one week of SH, and (2) carotid chemoafferent responses to acute hypoxia 4–5 months after SH exposure to determine the potential for long-lasting functional recovery of carotid body function.

2. Methods Experimental procedures were approved by the Animal Care Committee of the School of Veterinary Medicine of the University of Wisconsin-Madison. 2.1. Animal model Pregnant female Harlan Sprague Dawley (Strain 236b, Harlan Sprague Dawley, Madison, WI) rats were placed in 60% O2 + balance N2 in the last 2–4 days of pregnancy. The females were maintained in the hyperoxic chamber through parturition with their litters and for 1, 2 or 4 weeks post-partum. After hyperoxic exposure, the rats were returned to room air until adulthood (3–5 months of age). Only males were retained for study to avoid potential female hormonal effects on carotid chemoafferent responses to acute hypoxia (Tatsumi et al., 1995). Age matched control, normoxic rats were studied as were other groups of PNH rats kept under similar conditions but not exposed to 1 week of SH. The pups were selected from a minimum of 3 litters for each study group. Numbers of animals and body

G.E. Bisgard et al. / Respiratory Physiology & Neurobiology 145 (2005) 3–11

weights in each experimental group are presented in Section 3. 2.2. Study design 2.2.1. Study 1 At 3–5 months of age, rats exposed to PNH of 1, 2 or 4 weeks duration and control rats were exposed to 1 week of SH (12% O2 + balance N2 ) with flow through the chamber sufficient to limit the CO2 levels to below 0.5%. Similar numbers of rats were maintained in normoxia as controls. Within 2 h after removal from hypoxia, anesthesia and surgical preparation for CSN recording were initiated. Recordings of CSN responses to hypoxia were carried out 3–5 h after removal from hypoxia in SH-exposed groups. 2.2.2. Study 2 Only rats subjected to 2 weeks of PNH were studied in this protocol. We elected to study these rats prior to the time we knew the outcome from study 1. We fully expected 2-week PNH rats in study 1 to show recovered CSN function based on earlier studies of Fuller et al. (2001) who found significant recovery of phrenic nerve responses to acute hypoxia immediately after SH. Exposure to SH was as in study 1 and was carried out at 3 months of age. After the animals were removed from SH they were returned to normoxia until undergoing recording of CSN responses at 7–8 months of age. Control pups were maintained in normoxia throughout life to age 7–8 months of age and were studied within the same time frame as the PNH treated rats. 2.3. Animal preparation for CSN recording Rats were placed in a chamber containing isoflurane (approximately 4% in air) for rapid induction of anesthesia. When unconscious, 0.33 M urethane was administered intraperitoneally (1 g kg−1 ). Supplementary anesthetic (1–2 ml 0.33 M urethane) was given as needed if surgical anesthesia had not been reached after 5 min. Depth of anesthesia during surgical procedures was determined by toe pinch reflex assessment. Supplemental doses of urethane to maintain surgical anesthesia were given during the 1–2 h surgical preparation (1–3 ml/h, IV). This regime was continued after paralysis for neural recording (see below). Upon achieving surgical anesthesia, the femoral artery and

5

vein were cannulated for arterial pressure recording, arterial blood sampling, and intravenous administration of drugs and solutions. Body temperature was monitored by rectal thermister and maintained near 37 ◦ C with a warm water circulated blanket. An endotracheal cannula was inserted and the animals were artificially ventilated with a rodent ventilator (model 683, Harvard Apparatus, Holliston, MA). The end-tidal CO2 was monitored with a rapid responding flow through analyzer (Novametrix, Wallingford, CT) and maintained near 40 Torr PETco2 . Throughout surgical preparation the animals were ventilated with 25–30% O2 to maintain slightly hyperoxic arterial blood (Pao2 = 110–130 Torr). 2.4. Carotid sinus nerve recording The methods used have been previously published (Bisgard et al., 2003). In summary, the CSN was exposed via ventral cervical midline incision. It was dissected to its junction with the glossopharyngeal nerve and cut at the junction. The nerve was carefully cleaned and placed on tungsten electrodes for bipolar whole nerve recording. The animal was then paralyzed with pancuronium (2.5 mg kg−1 IV). Raw nerve activity was amplified (×10 K). A window discriminator, model WD-1000 and ratemeter, model RIC-1000 (CWE, Ardmore, PA) were used to obtain the discharge rate in 1 s bins (allowing determination of peak and average discharge rate in Hz). Data were recorded on a personal computer using commercial software (WINDAQ, Dataq, Akron, OH) and included raw nerve activity, discharge frequency, and arterial blood pressure. The investigator carrying out experiments and measuring records was blinded as to the treatment group. Baseline discharge rate at a Pao2 of approximately 100 Torr, was arbitrarily set near 300 Hz using the window discriminator as described by Ling et al. (1997). We subsequently have established that baseline activity between 200 and 300 Hz does not affect statistical outcome of the studies when using the increase in discharge frequency as the response to chemoreceptor stimuli. A baseline of 200 is slightly more sensitive (gives slightly higher increased frequency responses to acute stimuli), but 300 Hz as a normoxic baseline was used to be consistent with earlier studies (for details see Bisgard et al., 2003). Baseline activity contains chemoreceptor activity, baroreceptor activity (in

6

G.E. Bisgard et al. / Respiratory Physiology & Neurobiology 145 (2005) 3–11

some cases) and electrical noise. There was no detectable reduction in discharge frequency when Pao2 was raised by 100% O2 inhalation; thus, we could not use this maneuver to define the baseline normoxic discharge frequency. Therefore, we elected to examine the increase in CSN discharge frequency with acute hypoxia as a measure of carotid chemaoafferent function. For control baseline conditions, arterial blood gases were maintained at values near Pao2 = 100 Torr, Paco2 = 36–39 Torr and pHa = 7.4. This was achieved by adjusting inspired gases, the ventilator or administration of 0.5 ml of 8.4% NaHCO3 IV as needed. When recording whole CSN activity there are sometimes baroreceptor afferents as well as chemoreceptor afferents within the raw nerve signal. In a few instances, high amplitude action potentials from baroreceptors were found. These action potentials were eliminated from the recorded frequency with the window discriminator. The response to acute isocapnic hypoxia was initiated from normoxic conditions and carried out by altering inspired gases to reduce Pao2 from approximately 100 to 35–50 Torr over a 5–10 min period. In most cases, two or more measurements were made at different levels of Pao2 to characterize the hypoxic response. However, the statistical response was determined by the increase in discharge frequency from the baseline of about Pao2 100 Torr to near 45 Torr (values falling between 40 and 50 Torr). On achieving a steady-state, an arterial blood sample (0.2 ml) was obtained and the frequency of CSN discharge and arterial blood pressure were recorded. Arterial blood gas and acid-base variables were measured using a Radiometer analyzer (Model 500).

suggest different CSN responses to acute hypoxia, decreasing with increased duration of PNH. Furthermore, enhanced responses following SH treatment diminished in rats with longer durations of PNH. In order to statistically analyze these trends, we examined the increase in CSN frequency when reducing Pao2 from normoxia (approximately 100 Torr) to near 45 Torr as outlined in the methods section (Fig. 2). One-way ANOVA and a Tukey-Kramer post hoc test confirmed earlier findings showing that 1–4 weeks of PNH significantly attenuates the CSN response to acute hypoxia in animals maintained in normoxia (Nx) (P < 0. 01) and that 1 week of PNH is less effective than 2–4 weeks of PNH in this regard (P < 0.05) (Bisgard et al., 2003). Due to unequal variances, two way-ANOVA was not used to test for differences between SH treatment groups. CSN responses after SH in each group of control or PNH rats could be considered as a separate experiment as each duration of PNH (0, 1, 2 and 4 weeks) had its own control and SH treated groups. Therefore, each of these four groups could be subjected to a Student’s t-test. This analysis indicated significantly enhanced CSN responses to acute hypoxia after SH only in the control (0 weeks PNH, P < 0.01) and 1 week PNH groups (P < 0.02). Since inspection of acute hypoxic responses from PNH rats suggested that there was a PNH durationdependent exponential reduction in CSN response to acute hypoxia (Fig. 2), we further analyzed the data using exponential regression techniques (SAS procedure NLIN, Software Release 8.2, SAS Institute Inc., Cary, NC, USA). A three parameter model was chosen to analyze exponential decay curves relating the acute CSN hypoxic response and the duration of PNH:

2.5. Data analysis

General model y = ae−bt + c

Means ± S.E.M. were obtained for each treatment. Statistical methods are outlined in Section 3.

3. Results 3.1. 1, Study 1 Fig. 1 illustrates plots of the CSN responses to acute hypoxia in adult control rats and in rats treated with 1, 2 and 4 weeks of PNH. The curves generated

y = change in CSN discharge frequency (Pao2 45–100 Torr), Hz; a = maximal increase in CSN discharge frequency (Hz) from the plateau, c; c = mean plateau value reached by both control (Nx) and SH curves (48.3 Hz); b = rate constant; t = duration of PNH, weeks By modeling both groups (SH versus Nx) simultaneously, a and b were estimated for each treatment while holding c constant, effectively creating a five parameter model. The decision to hold c constant was based on the preliminary observation of identical mean values for c in SH and Nx groups.

G.E. Bisgard et al. / Respiratory Physiology & Neurobiology 145 (2005) 3–11

7

Fig. 1. Whole CSN discharge frequency of all animals in response to acute isocapnic reduction in Pao2 . Closed squares and solid lines indicate normoxic control studies; open circles and dashed lines indicate animal exposed to 1 week of sustained hypoxia (12% O2 ). Curves through data points were drawn by Excel 2000 program (Microsoft Corp., Seattle, WA). PNH indicates exposure to postnatal hyperoxia.

The parameter estimates from this regression analysis were: aNx = 115.8 ± 22.3 (± S.E.M.) aSH = 209.6 ± 22.7(different from aNx , P < 0.05) bNx = 1.05 ± 0.56 bSH = 0.73 ± 0.21 (nsd from bNx ) The significant increase in aSH (with constant c) indicates a progressively diminishing upward shift in the CSN response to acute hypoxia due to SH (Fig. 3), with the greatest effect in control rats and no effect after 4 weeks of PNH. The analysis does not reveal a difference in the rate of decay (b) between SH and Nx treatments, indicating a constant relative time course from a greater initial value.

Mean arterial blood pressure decreased during acute hypoxia in CSN studies in all groups (Table 1). These decreases were attenuated to the same degree after SH, but normoxic BP tended to be higher in all groups. Because arterial blood pressure changes were comparable among treatment groups, it is unlikely blood pressure changes influenced differences in CSN responses between groups. Body weights did not differ between groups within this study (Table 1). 3.2. Study 2 In this study, 2 week PNH rats were exposed to 1 week of SH (12% O2 ) at 3 months of age and were then studied at 7–8 months of age. These animals maintained depressed CSN responses to acute hypoxia

8

G.E. Bisgard et al. / Respiratory Physiology & Neurobiology 145 (2005) 3–11

Fig. 2. Mean (± S.E.M.) whole CSN responses to acute hypoxia expressed as the increase in discharge frequency when Pao2 was decreased from near 100 Torr to approximately 45 Torr (ranges defined in the methods section). Abbreviations are as in the legend to Fig. 1.Control animals (Con.) were not treated with PNH. + indicates that PNH significantly reduced the acute response to hypoxia as compared to control, Nx animals (P < 0.01). * Indicates a significant increase in the mean response to acute hypoxia following SH compared to the mean Nx value within each group (P < 0.02).

comparable to those not treated with SH (Fig. 4, Table 2). Thus, there was no evidence of any long-term recovery of the CSN response to acute hypoxia. Age matched control animals maintained their normal CSN responses to acute hypoxia (Fig. 4, Table 2). Mean arterial blood pressures were similar among groups during normoxia and decreased significantly, but to similar extents, during acute hypoxia in all studies (Table 2). Body weights did not differ between groups within this study (Table 2).

Fig. 3. Individual data points used to create Fig. 2 were plotted as exponential decay curves as defined and characterized in Section 3. Arrowhead indicates control values (animals not treated with PNH). The analysis suggests a diminishing effect of SH to increase the acute CSN response to hypoxia that is dependent on the duration of PNH.

Fig. 4. CSN responses in control (Con.) rats and rats treated with 2 weeks of PNH (PNH). Responses were obtained 4–5 months after exposure of adult rats to 1 week of sustained hypoxia, carried out at 3 months of age (open bars) or in age matched rats maintained in normoxia (solid bars). The data were derived as presented for Fig. 2.There was a significant reduction of responses after PNH (P < 0.001, +) as compared to its respective control group, but 1 week of sustained hypoxia (SH) did not significantly improve the response to acute hypoxia 4–5 months later. Blood gas data are presented in Table 2.

4. Discussion Our data confirm that the chemoafferent response to acute hypoxia in adult rats is significantly attenuated after PNH of 1–4 weeks duration, and that 1 week of PNH does not produce the same degree of impairment as does 2 or 4 weeks of PNH (Bisgard et al., 2003). Immediately after 1 week of SH, the CSN responses to acute hypoxia were elevated in nearly all normal adult rats and in rats treated with 1 week of PNH. Based on the results of phrenic nerve recordings in an earlier study (Fuller et al., 2001, see below), we expected SHinduced recovery of CSN responses to acute hypoxia in rats treated with 2 and 4 weeks of PNH. By exponential regression analysis we found a trend to increased CSN responses to hypoxia in the 2-week SH group, but there were only 4 rats in this group. We could detect no evidence of recovery immediately after SH in 4-week PNH rats. There was no evidence for recovery in 2-week PNH rats 4–5 months after SH. Thus, functional chemoafferent plasticity induced by 1 week of SH occurred in normal rats and was found in rats treated with 1 week of PNH, but was limited after 2 weeks of PNH and absent after 4 weeks of PNH. 4.1. Critique of methods Ideally, one would choose to carry out single fiber studies to assess CB afferent responses to acute hypoxia

G.E. Bisgard et al. / Respiratory Physiology & Neurobiology 145 (2005) 3–11

9

Table 1 Arterial blood gases (Pao2 and Paco2 , Torr), pH and mean arterial blood pressure (MBP, Torr) during normoxia and acute hypoxia CSN responses presented in Figs. 2 and 3, before and immediately after 1 week of sustained hypoxia in adult rats Normoxia pHa

Acute hypoxia MBP

pHa

Paco2

Pao2

MBP

Normal rats, no PNH, normoxic control, BW = 409 ± 22 g, n = 13 Mean 7.39 38.0 109 126 S.E. 0.01 0.6 4.2 4.2

7.39 0.01

38.9 0.7

44.5 0.6

76* 7.4

Normal rats, no PNH, after 1 week 12% O2 , BW = 404 ± 18 g, n = 11 Mean 7.39 37.2 101 135 S.E. 0.01 0.7 2.5 3.7

7.39 0.01

37.8 37.8

44.0 1.2

92* 8.9

1 Week PNH rats, normoxic control, BW = 420 ± 34, n = 8 Mean 7.40 39.1 97.4 S.E. 0.01 0.8 2.4

108 4.9

7.41 0.01

38.8 1.4

45.1 1.3

76* 7.4

1 Week PNH rats, after 1 week 12% O2 , BW = 378 ± 13 g, n = 8 Mean 7.37 36.4 105 126+ S.E. 0.01 1.3 3.6 3.4

7.37 0.01

37.1 1.2

45.8 1.1

100* 10.8

2 Week PNH rats, normoxic control, BW = 399 ± 12 g, n = 7 Mean 7.41 40.5 109 118 S.E. 0.01 0.7 4.1 9.2

7.42 0.01

39.5 1.1

44.6 1.1

91* 12.5

2 Week PNH rats, after 1 week 12% O2 , BW = 387 ± 26 g, n = 4 Mean 7.38 37.5 92.5 119 S.E. 0.01 1.6 4.6 10.3

7.38 0.01

37.7 1.9

45.8 1.3

79 20.2

4 Week PNH rats, normoxic control, BW = 367 ± 10 g, n = 8 Mean 7.38 39.1 103 118 S.E. 0.01 1.5 4.0 2.3

7.39 0.01

38.8 1.1

45.9 1.0

82* 9.2

4 Week PNH rats, after 1 week 12% O2 , BW = 377 ± 12 g, n = 8 Mean 7.39 36.4 103.7 127 S.E. 0.01 1.6 3.3 5.2

7.40 0.01

36.0 1.5

44.5 1.2

101* 5.5

∗ +

Paco2

Pao2

Significantly different from normoxic value (P < 0.03). Significantly different from corresponding normoxic control group (P < 0.01).

after PNH. However, this is a difficult approach and the high failure rate would preclude the study of large numbers of animals required to complete such studies. We have found whole nerve recording to be both successful and practical, but with disadvantages such as the inability to precisely know resting normoxic discharge rates and the inability to completely separate chemoreceptor and baroreceptor afferent fibers. These limitations were previously addressed in detail (Bisgard et al., 2003). Fortunately, rat chemoreceptor responses to hypoxia are robust and overwhelm baroreceptor discharge to a large extent and arterial blood pressure invariably falls during acute hypoxia. It has been concluded that there is both reduced numbers of

carotid chemoafferent fibers and reduced o2 sensitivity of individual fibers after PNH (Carroll et al., 2003; Erickson et al., 1998; Prieto-Lloret et al., 2003). If there is a change in either numbers of functional fibers or hypoxic sensitivity after SH, our studies should detect a reduced CSN response. However, it is possible that small changes in hypoxic sensitivity could be missed in whole CSN recordings. The decrease in mean arterial blood pressure would reduce baroreceptor discharge frequency, slightly reducing whole nerve responses to acute hypoxia. Since PNH has no effect on baroreceptor discharge and blood pressure responses are similarly affected during acute hypoxia in all groups of animals, there should be no systematic effect on our

10

G.E. Bisgard et al. / Respiratory Physiology & Neurobiology 145 (2005) 3–11

Table 2 Arterial blood gases (Paco2 and Pao2, Torr), pH and mean arterial blood pressure (MBP, Torr) during normoxia and acute hypoxia CSN responses reported in Fig. 4, in control rats and in rats 3–4 months after 1 week of sustained hypoxia Normoxia pHa

Acute hypoxia Paco2

Pao2

MBP

pHa

Paco2

Pao2

MBP

Normal rats, normoxic control, BW = 423 ± 27 g, n = 9 Mean 7.39 38.8 106 S.E. 0.01 0.7 4.7

122 6.7

7.39 0.01

39.4 1.0

46.0 1.0

81* 10.1

Normal rats, 1 week adult hypoxia, BW = 468 ± 21 g, n = 4 Mean 7.40 37.6 103 S.E. 0.01 1.57 1.1

136 9.5

7.41 0.01

38.1 0.9

46.0 1.2

78* 7.3

2 weeks PNH, normoxic control, BW = 473 ± 13 g, n = 8 Mean 7.39 37.4 107 S.E. 0.01 1.0 2.2

126 2.8

7.39 0.01

37.0 0.9

43.7 0.9

70* 8.2

2 weeks PNH, 1 week adult hypoxia, BW = 475 ± 21 g, n = 6 Mean 7.39 37.4 103 S.E. 0.01 1.4 1.1

123 5.7

7.39 0.01

38.3 1.2

44.2 1.1

70* 11.7

∗

Significantly different from normoxia value (P < 0.05).

results due to variable baroreceptor responses during hypoxia. 4.2. Previous studies Fuller et al. (2001) found an increase in the integrated phrenic nerve response to acute hypoxia after 1 week of SH in adult animals subjected to 4 weeks of PNH. The acute increase in minute phrenic activity was equivalent to that found in normal rats suggesting a strong recovery of the response. However, they did not do studies of a normal group of rats after SH. Our data in 4-week PNH rats after SH, which should be most similar to the animals studied by Fuller et al., indicated a nearly identical range of depressed responses to acute hypoxia with and without SH treatment (Figs. 1–3). Thus, our data for the 4 week PNH rats suggests little potential for CB/CSN plasticity after SH in contrast to the integrated phrenic responses to acute hypoxia after SH reported by Fuller et al. (2001). Because they found some histologic evidence for vasodilation and increased CB volume similar to that seen in normal animals associated with SH, they speculated that it was likely that functional CB chemoafferent plasticity had occurred, making the CB/CSN input to the CNS greater during acute hypoxia. In normal animals SH causes cell proliferation and vascular dilation resulting in CB enlargement, increased catecholamine metabolism and

increased CB/CSN sensitivity to acute hypoxia after SH (Wang and Bisgard, 2002). Our data do not support the speculation of Fuller et al. (2001) that CB/CSN plasticity explained enhanced phrenic nerve responses to acute hypoxia after SH in PNH rats. We suggest that their results could be associated with greater CNS response to CB chemoafferent input after SH resulting in a significantly greater integrated phrenic response as has been previously shown in normal rats (Dwinell and Powell, 1999; Powell et al., 2000). Recently Bavis et al. (2003) showed that there was an age-dependent increase in the phrenic response to acute hypoxia in rats treated with PNH in the first week of life. The 1-week PNH rats had been maintained in normoxia (no exposure to hypoxia) for their entire life (up to 1 year). Rats maintained for longer periods (4 weeks) of PNH show no such spontaneous recovery of function (Fuller et al., 2002). Other studies have suggested that the first two weeks of life in the rat are the most critical for developmental susceptibility to hyperoxic attenuation of CSN or phrenic responses to hypoxia (Bisgard et al., 2003; Bavis et al., 2002). Our present results are compatible with these findings. We postulate that SH-induced plasticity in CSN hypoxic responses is observed in 1-week PNH rats because they have more functioning CB and CSN tissue than rats treated with 4 weeks of PNH. We also suggest that the

G.E. Bisgard et al. / Respiratory Physiology & Neurobiology 145 (2005) 3–11

residual functioning tissue is sufficient to contribute to the slow spontaneous recovery of phrenic responses to acute hypoxia found by Bavis et al. (2003). 4.3. Conclusions Our studies suggest that longer durations of PNH (>2 weeks) limit the ability of rats to increase CSN responses to acute hypoxia after SH. Rats treated with one week of PNH, with more responsive CBs, retain the ability to mount hypoxia-induced plasticity revealed as an increased CSN response to acute hypoxia.

Acknowledgements The authors thank Gordon Johnson for excellent technical assistance and Peter Crump and Julia Wilkerson for assistance with statistical analysis. This research was supported by National Heart Lung and Blood Institute Grants HL68255, HL07654 and HL70506.

References Bavis, R.W., Olson Jr., E.B., Mitchell, G.S., 2002. Critical developmental period for hyperoxia-induced blunting of hypoxic phrenic responses in rats. J. Appl. Physiol. 92, 1013–1018. Bavis, R.W., Olson Jr., E.B., Vidruk, E.H., Bisgard, G.E., Mitchell, G.S., 2003. Level and duration of developmental hyperoxia influence impairment of hypoxic phrenic responses in rats. J. Appl. Physiol. 95, 1550–1559. Bisgard, G.E., 2000. Carotid body mechanisms in acclimatization to hypoxia. Respir. Physiol. 121, 237–246.

11

Bisgard, G.E., Olson Jr., E.B., Wang, Z.-Y., Bavis, R.W., Fuller, D.D., Mitchell, G.S., 2003. Adult carotid chemoafferent responses to hypoxia after 1, 2 and 4 weeks of postnatal hyperoxia. J. Appl. Physiol. 95, 946–952. Carroll, J.L., Kim, I., Boyle, K.M., Carle, C.M., Donnelly, D.F., 2003. Perinatal hyperoxia impairs responsiveness of single unit rat carotid chemoreceptor activity. FASEB J. 17, 127. Dwinell, M.R., Powell, F.L., 1999. Chronic hypoxia enhances the phrenic nerve response to arterial chemoreceptor stimulation in anesthetized rats. J. Appl. Physiol. 87, 817–823. Erickson, J.T., Mayer, C., Jawa, A., Ling, L., Olson Jr., E.R., Vidruk, E.H., Mitchell, G.S., Katz, D.M., 1998. Chemoafferent degeneration and carotid body hypoplasia following chronic hyperoxia in newborn rats. J. Physiol. (Lond.) 509, 519–526. Fuller, D.D., Bavis, R.W., Vidruk, E.H., Wang, Z.-Y., Olson Jr., E.B., Bisgard, G.E., Mitchell, G.S., 2002. Life-long impairment of hypoxic phrenic responses in rats following 1 month of developmental hyperoxia. J. Physiol. (Lond.) 538.3, 947–955. Fuller, D.D., Wang, Z.-Y., Ling, L., Olson Jr., E.B., Bisgard, G.E., Mitchell, G.S., 2001. Induced recovery of hypoxic phrenic responses in adult rats exposed to hyperoxia for the first month of life. J. Physiol. (Lond.) 536.3, 917–926. Ling, L., Olson Jr., E.B., Vidruk, E.H., Mitchell, G.S., 1997. Developmental plasticity of the hypoxic ventilatory response. Respir. Physiol. 110, 261–268. Pieto-Lloret, J., Caceres, A.I., Obeso, A., Rocher, A., Rigual Agapito, M.T., Bustamante, R., Casta˜neda, J., Perez-Garcia, M.T., L´opezL´opez, J.R., Gonz´alez, C., 2003. Ventilatory responses and carotid body body function in adult rats perinatally exposed to hyperoxia. J. Physiol. (Lond) 554.1, 126–144. Powell, F.L., Huey, K.A., Dwinell, M.R., 2000. Central nervous system mechanisms of ventilatory acclimatization to hypoxia. Respir. Physiol. 121, 223–236. Tatsumi, K., Hanhart, B., Moore, L.G., 1995. Hormonal influences on ventilatory control. In: Dempsey, J.A., Pack, A.I. (Eds.), Regulation of Breathing, second ed. Dekker, New York, pp. 829– 864. Wang, Z.-Y., Bisgard, G.E., 2002. Chronic hypoxia-induced morphological and neurochemical changes in the carotid body. Micr. Res. Tech. 59, 168–177.