|
|
||||||||
1Department of Physiology, Medical College of Wisconsin, 2Department of Physical Therapy, Marquette University, and 3Zablocki Veterans Affairs Medical Center, Milwaukee, Wisconsin 53226
Submitted 22 June 2004 ; accepted in final form 5 August 2004
| ABSTRACT |
|---|
|
|
|---|
central chemoreception; control of breathing
One technique utilized to study chemoreceptor function in the intact, unanesthetized animal is creation of focal acidosis (FA) by microdialysis (MD) in putative chemoreceptor regions of mock cerebral spinal fluid (mCSF) equilibrated with high levels of CO2 (11, 15, 16, 2425). Indeed, Nattie et al. (16, 2425) reported that FA in the NTS, RTN, or MRN in rats increase minute ventilation in a state-independent (NTS) or state-dependent (RTN, MRN) manner. In addition, our laboratory recently reported that FA in the MRN of goats also increases ventilation during wakefulness but found no effect of FA on breathing during sleep (11). However, in both rats and goats, the resulting increases in breathing ranged from 8 to 28% above control ventilation, which is small compared with a 250% increase during an equivalent systemic (brain) acidosis resulting from inhalation of 7% CO2. The relatively small increases in breathing during FA may result because no single chemoreceptor site can account for the overall CO2 sensitivity, but rather activation of multiple sites is required (23). To our knowledge, there are no published data on the ventilatory effects of FA at multiple brain sites during physiological conditions. Therefore, the major aim of this study was to test the hypothesis that FA at multiple chemoreceptor sites will have a greater effect on breathing than FA at one chemoreceptor site. Obtaining these data may provide insight into the question of why there are chemoreceptors at multiple sites in the brain.
| METHODS |
|---|
|
|
|---|
Surgical Procedures
Instrumentation surgery. An initial surgery was performed to elevate a 5-cm segment of the carotid arteries. In this and subsequent surgeries, the goats were anesthetized initially with a combination of ketamine and xylazine, intubated, and mechanically ventilated. Throughout surgery, anesthesia was maintained with 11.5% halothane in O2. Under sterile conditions, the carotid arteries were isolated from the vagi, they were elevated superficial to the muscle, and the skin was sutured. After surgery, the goats received ceftifur sodium (2 mg/kg) daily as an antibiotic for 1 wk.
Microtubule implantation surgery.
After
3 wk, a second surgery was performed to chronically implant two or three microtubules (MTs) into the medullary raphe (n = 7). An occipital craniotomy was created, and dura mater was excised to expose the posterior cerebellum and dorsal aspect of the medulla for visualization of the obex. The dorsal surfaces of the medulla, the obex, and the midline were all used as reference points for stereotaxic coordinates in the dorsoventral, rostrocaudal, and mediolateral planes. The implantation sites were along the caudal (raphe obscurus) and rostral (raphe pallidus) midline. The MTs used were 18-gauge stainless steel tubes 70 mm in length. After placement, the MTs were secured with screws in the bone and dental acrylic. Arterial blood pressure (BP) and rectal temperature were continuously monitored throughout the duration of surgery.
Laboratory personnel monitored the goats continuously for a minimum of 24 h after the MT implantation surgery. Most goats were unable to maintain normal sternal recumbent posture and/or stand for 36 h postsurgery. Two goats were unable to stand unassisted for
3 days after the implant. However, after full recovery, all animals were studied. Food and water intake were monitored closely in all goats daily after the implantation surgery. Brain edema was minimized with dexamethasone injections (0.4 mg·kg1·day1 iv for 2 days, then decreasing by 0.05 mg·kg1·day1) three times a day for 1 wk. Infection was minimized with chloramphenicol injections (20 mg/kg iv) for 3 days and with daily injections thereafter of ceftifur sodium (2 mg/kg) and gentamyacin (3 mg/kg). Buprenorphine was administered 312 h after implantation to minimize pain.
Physiological Measurements
For all studies, a fitted mask was taped firmly to the snout, and a two-way breathing valve was attached to the mask to measure inspiratory flow (
I) with a pneumotachograph and for collection of expired air analyzed for O2 and CO2 concentrations required for O2 uptake (
O2) and CO2 production (
CO2) calculations. A chronically placed catheter in the elevated carotid artery was used to measure arterial BP and heart rate (HR) and for arterial blood sampling to obtain pH, arterial PO2, and arterial PCO2 (PaCO2) values (model 278, Ciba-Corning). Rectal temperature of the animal was measured at regular intervals.
Whole body ventilatory CO2 sensitivity was assessed on multiple days before and on days of FA studies. Room air (RA) breathing, BP, and HR were measured for 30 min before exposure to three levels of elevated inspired CO2 (2.5, 5.0, and 7.5% CO2 in RA). Arterial blood samples were drawn during the control period and during the fourth and fifth minute of each CO2 exposure level. The changes in expired ventilation (
E) and PaCO2 from RA breathing to all levels of CO2 were used determine the slope of the relationship between
E and PaCO2, and they were used as an index of CO2 sensitivity.
FA studies. Studies began at least 2 wk after implantation (30). Throughout the protocol, all goats were in good health, with stable baseline breathing and PaCO2 values. During the first experiments performed in some animals, shivering occurred. These data were excluded from the final analysis of all physiological data.
The CMA (CMA Microdialysis, Solna, Sweden) 12 MD probes (20-kDa molecular mass cutoff) had a 70-mm shaft length, a 2-mm membrane length, and a 0.5-mm membrane diameter. Identical MD probes were used in our laboratory's initial sets of experiments, when we established that MD with 50 and 80% CO2 generated an extracellular fluid acidosis slightly greater than, or three times greater than, that observed with 7.5% inspired CO2 in room air (11). The contents of the dialysate (mCSF) have been previously described (11).
I, HR, BP,
CO2, and
O2 were measured continuously or at regular intervals during a 15-min control period, during 45 min of microdialysis, and for 15 min after termination of the dialysate flow. Arterial blood was drawn during the final 5-min period of the control, microdialysis, and recovery periods. Three different dialysate pH and PCO2 conditions were tested: 1) 6.4% CO2 (pH = 7.317.36, PCO2 = 4147 Torr), 2) 50% CO2 (pH = 6.56.6, PCO2 > 250 Torr), and 3) 80% CO2 (pH = 6.36.4, PCO2 > 250 Torr). These studies were performed in individual and multiple MTs with a flow rate (50 µl/min) identical to that previously described (11).
Histological studies. After completion of these protocols, the animals were euthanized (Beuthanasia), and the brain was perfused with PBS solution (pH = 7.357.4) and 4% paraformaldehyde fixative in PBS. The medulla was then removed, postfixed in 4% paraformaldehyde solution for 2448 h, and cryoprotected in a 30% sucrose solution. The medulla was then frozen and serial sectioned (25 µm) in a transverse plane, and the sections adhered to chrom alum-coated slides. The tissue was then stained with hematoxylin and eosin, coverslipped, and examined microscopically. The MT implantation site was identifiable by visualization of an area of absent or disrupted tissue, which extended over a finite rostrocaudal distance (0.91.1 mm). The implantation site was defined as being at the tip (ventral-most aspect) and middle of the MT-induced tissue disruption. The implant locations were derived from calculation of caudorostral distance from obex, dorsoventral measurements at the midline, and the published data from the anatomic atlas of the goat medulla (6).
Data and Statistical Analyses
We calculated
I, mean arterial BP, HR, breathing frequency, tidal volume (VT), inspiratory flow rate [VT/inspiratory time (TI)],
O2,
CO2, respiratory quotient (RQ), dead space volume (VD), and the ratio of VD to VT (VD/VT). As previously described, the calculated breath-by-breath data were binned into 5-min averages for the control, MD and recovery periods, where all bins were divided by the 15-min control mean for normalization (%control). PaCO2, VD, and VD/VT values were averaged for each of the control, MD, and recovery periods. The mean data were statistically analyzed by using a one-way ANOVA with repeated measures and Bonferroni post hoc analysis. The threshold for significance was set at P < 0.05.
| RESULTS |
|---|
|
|
|---|
All implanted MTs were <1.2 mm lateral of midline, and <1.0 mm dorsal to midline raphe nuclei (Fig. 1). MD probes when inserted for studies extended <2 mm beyond the ventral-most aspect of the MT, indicating that FA was generated within midline raphe nuclei.
|
Ventilatory variables while breathing RA and whole body CO2 sensitivity were stable and within the normal physiological range for goats over 45 days before and during the days of FA studies (Table 1). Both BP (8.4 ± 3.3%) and HR (16.9 ± 6.0%) tended to increase with increasing inspired CO2, but these changes were not significantly different from control (P
0.083; n = 5).
|
Microdialysis of mCSF equilibrated with 6.4% CO2 (control) or 50% CO2 had no effect on all calculated variables during the dialysis or recovery period (P
0.101). However, FA with 80% CO2 at a single caudal or rostral raphe site increased
I and VT (P
0.006), but it had no effects on all other variables during the dialysis or recovery periods (P
0.209). Two-way ANOVA analyses comparing 6.4 and 80% CO2 dialysis effects indicated significant increases in
I, VT, and breathing frequency during FA at single raphe sites (P < 0.05; Fig. 2).
|
Microdialysis with 6.4% CO2 had no effect on all measured variables (P > 0.153).
I increased (P = 0.015) with FA at multiple raphe sites with 50% CO2, with no effects on all other measured variables (P > 0.05). Microdialysis with 80% CO2 increased
I, VT, VT/TI,
O2,
CO2, and HR (P < 0.05), and it led to a small but significant hyperventilation (1.6 Torr; P < 0.005) with no effect on breathing frequency, RQ, VD, VD/VT, or BP (P > 0.05). CO2 dialysis at multiple raphe sites with both 50 and 80% CO2 increased
I compared with 6.4% CO2, where the effect of 80% CO2 was greater than that of 50% CO2 at two time points (P > 0.05; Fig. 3). Dialysis with 80% CO2 at multiple sites had greater effects on
I and
O2 (P < 0.05) than dialysis with 80% CO2 at single raphe sites, but VT and HR were not different (P > 0.05; Fig. 4).
|
|
| DISCUSSION |
|---|
|
|
|---|
Limitations of Study
In our model, the capability to generate FA in the medullary raphe under physiological conditions requires chronic implantation of stainless steel MTs. Previously, implantation of two MTs into other respiratory-related nuclei was accomplished with minimal surgical and/or recovery complications, largely due to avoiding passing the microtubules through the cerebellum. Caudal raphe regions lie caudal to the posterior aspect of the cerebellum, and therefore they can be reached without damaging the cerebellum. However, to reach more rostral raphe sites required passing MTs through the medial cerebellum. In initial goats, we only implanted one MT in each of the caudal and more rostral raphe, and the goats recovered well; thus, in subsequent goats, we implanted two MTs caudally and one more rostrally, or two rostrally. The latter goats in particular had a longer than normal recovery period because it took 4 days before they were able to stand unassisted. Therefore, we limited the number of raphe sites tested in a given animal because of the limitations of recovery from the surgical approach.
Additionally, the interpretations of these data are in the context of one nucleus: the medullary raphe. We did not test the effects of focal acidification at more than one chemosensitive nucleus but rather the effects of multiple acid foci within one chemoreceptive nucleus. That is to say that these data do not specifically address multiple chemoreceptor nuclei function per se, nor the previously noted state-dependent function of the raphe nucleus.
In previous investigations, pH measurements were made within
200 µm of the site of FA to determine the degree of acidification created with various CO2/H+ levels (11). In the present study, we did not measure pH changes created with FA, and therefore we cannot directly confirm that we have indeed generated a greater degree of acidosis with FA at multiple sites or whether there were regional differences in the degree of acidification. However, there are lines of evidence that address these issues. First, we previously found that the pH change at the MD site was greater when two MD probes (CMA 11) were simultaneously used in parallel compared with the pH change using one probe (unpublished observations). Second, local pH regulatory mechanism do not appear to differ greatly between medullary regions because we found no differences in the change in the extracellular fluid pH with increasing inspired CO2 fraction (0.07) when measurements were made in the caudal or rostral raphe or in the retrotrapezoid, facial, or gigantocellularis reticularis nuclei (unpublished observations). These findings plus the finding that breathing and metabolic rate increase more with multiple vs. single sites of microdialysis lead us to conclude that indeed a FA was created at multiple sites.
A final consideration is the potential trauma or tissue scarring that could conceivably occur with multiple insertions of the MD probe. Indeed, the probe is repeatedly passed into the medulla 2 mm beyond the ventral-most aspect of the MTs to test the effects of single vs. multiple sites. Thus the probes are inserted a minimum of two times at a given site, potentially creating a physical barrier for diffusion by scar tissue generation with multiple insults. Although scarring may be a factor, we believe that this is unlikely because of the dramatic increase in
I (+163%) observed after ibotenic acid injection performed after completion of these studies that indicates minimal or no diffusion barrier (unpublished observations).
Physiological Effects of FA at Single and Multiple Raphe Sites
FA with microdialysis of high CO2 has previously been utilized to test for intracranial chemoreceptor function (11, 16, 2425). The changes in breathing observed from FA in regions such as the caudal and rostral NTS, RTN/Ppy, and caudal midline raphe range from 15 to 28% in rats and from 8 to 13% in goats. The changes in breathing observed with FA at single raphe sites in our laboratory's previous report (11) and in the present study were similar in magnitude and time course. In the present study, FA with 80% CO2 at single caudal or rostral raphe sites transiently increased
I (16.3 ± 4.9%), VT (11.1 ± 3.8%), and breathing frequency (5.5 ± 1.5%), similar to our laboratory's previous report where dialysis with 80% CO2 increased
I (12.0 ± 2.0%), VT/TI (13.3 ± 1.9%), and breathing frequency (5.9 ± 1.6%). Additionally, we detected no differences in response to FA in caudal vs. more rostral raphe regions in both studies despite a greater number of total, neurokinin-1 receptor-expressing, and serotonergic (chemosensitive) neurons in the rostral midline raphe (12).
FA at multiple raphe sites had a greater effect than FA at single raphe sites with an equivalent CO2 level, where dialysis of 80% CO2 had a greater effect on average than did 50% CO2. Breathing maximally increased 30% from control, which represents a fraction of the overall ventilatory response to 7.5% inspired CO2 (
250%). Increases in ventilation with FA in the medullary raphe correlate well with the idea of a "dose response" to acidification, by increasing the CO2/H+ content or number of acidic foci (Fig. 5). In light of the evidence for widespread central chemoreceptor sites, it therefore seems likely that the ventilatory response to inspired CO2 reflects a cumulative effect of stimulation of multiple chemoreceptor sites. This postulate does not preclude the possibility that certain chemoreceptor sites may be more or less sensitive to pH or CO2 or that they may function in a state-dependent manner (2123). Additionally, the functional contribution of each chemoreceptor site may differ in terms of VT and/or breathing frequency. Breathing effects from FA with microdialysis of high CO2 in the RTN were due solely to increases in VT (rats), whereas FA in the raphe (rats and goats), NTS (rats), or pre-Bötzinger complex (cats) had both VT and breathing frequency effects (11, 16, 24, 25, 30). Finally, it is likely that there are differences in the capabilities of central chemoreceptor sites to locally regulate pH as indicated by the findings of heterogeneous distribution and pH regulation of glia in the RTN and NTS in neonatal rat brain slices (8). In other words, the role of breathing in correction for a FA may not be the same at all sites in the brain.
|
I, VT, and breathing frequency effects, we noted significant increases in metabolic rate, BP, and HR with dialysis of high CO2 in the raphe obscurus and pallidus in these and previous investigations (11). The increase in metabolic rate with FA at multiple raphe sites likely contributes to the ventilatory response. However, coincident with the substantial increased metabolic rate was a small but significant 2-Torr decrease in PaCO2, indicating that there is an effect on ventilation independent of the increased metabolic rate. This finding is consistent with FA in the RTN of awake rats, which increased
E 24% and led to a 5-Torr reduction in PaCO2 (15). Additionally, the resulting systemic hypocapnia may also inhibit other chemoreceptors (both central and peripheral), which may contribute to the relatively small increases in breathing observed with FA. The increased metabolic rate and the cardiovascular changes with FA at multiple raphe sites likely reflect the postulated influence of the medullary raphe on sympathetic drive, affecting thermogenesis [brown adipose tissue (BAT) metabolism], peripheral vascular tone (3), BP, and HR. Retrograde tracer injections into interscapular BAT indicate that, among others, the raphe pallidus may play a role in premotor regulation of sympathetic nerve activity (1, 4). Activation (via disinhibition) of raphe pallidus neurons with bicuculline or intravenous leptin administration both increased BAT sympathetic nerve activity, BAT temperature, expired CO2 (19, 20), as well as HR and BP (19). Microinjection of the 5-HT type 1A receptor agonist 8-hydroxy-2(di-n-propylamino)tetralin into the raphe pallidus eliminated the leptin-evoked response, further implicating a role for the raphe in thermogenesis (19).
Although some previous reports have shown that the increase in
O2 observed during CO2 inhalation is due to the increased O2 cost (work) of breathing (9, 17, 29, 32), others have shown
O2 increased to a greater degree than predicted by O2 cost of breathing alone with increasing inspiratory PCO2 (14). Additionally, Pappenheimer (28) found that the hypercapnia-induced hyperpnea increased
O2 by 31%, whereas a similar hypoxia-induced hyperpnea decreased
O2 by 18% in rats. Similarly, in ponies and oxen,
O2/
E was greater during a CO2-induced hyperpnea compared with a thermal-induced hyperpnea (10, 13). In addition, in ponies, the increased
O2 during CO2 inhalation was accentuated or attenuated during exposures to cold or hot environmental temperatures, respectively (13). These data led to the conclusion that the increase in
O2 during hypercapnia represents in part a response to offset the respiratory heat loss, thereby contributing to the maintenance of homeothermy.
The observation that a large fraction of the ventilatory response may be related to increased metabolic rate indicates that the traditional concept of intracranial chemoreception is too narrow. Indeed, one of the postulates for why chemoreceptors are located at several brain sites is that they do not all affect the same physiological functions or serve the same purpose, i.e., cardiovascular function, thermogenesis, and arousal. Intracranial chemoreceptors at some sites might be primarily respiratory (RTN), whereas other sites might be primarily thermogenic or cardiovascular (medullary raphe) or may serve to increase arousal (pontine raphe, locus coeruleus). In other words, hypercapnia places multiple demands on the body, which are met by activation of multiple systems and/or functions, each under control by chemoreceptors at different sites in the brain.
Finally, other studies in piglets, goats, and rats have previously shown that medullary raphe lesions attenuate whole body CO2 sensitivity (7, 12, 26). Because the present study further establishes that chemoreceptors in the raphe stimulate breathing during physiological conditions, the lesion-induced attenuation of CO2 sensitivity may reflect a specific effect on chemoreceptors rather than an effect on a nonchemoreceptor neuromodulation of breathing by the raphe.
| CONCLUSIONS |
|---|
|
|
|---|
| GRANTS |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
|
|
|---|
| FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
P. G. Guyenet The 2008 Carl Ludwig Lecture: retrotrapezoid nucleus, CO2 homeostasis, and breathing automaticity J Appl Physiol, August 1, 2008; 105(2): 404 - 416. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. B. Dias, A. Li, and E. Nattie Focal CO2 dialysis in raphe obscurus does not stimulate ventilation but enhances the response to focal CO2 dialysis in the retrotrapezoid nucleus J Appl Physiol, July 1, 2008; 105(1): 83 - 90. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Li and E. Nattie Serotonin transporter knockout mice have a reduced ventilatory response to hypercapnia (predominantly in males) but not to hypoxia J. Physiol., May 1, 2008; 586(9): 2321 - 2329. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. G. Guyenet, R. L. Stornetta, and D. A. Bayliss Retrotrapezoid nucleus and central chemoreception J. Physiol., April 15, 2008; 586(8): 2043 - 2048. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. R. Hodges, G. J. Tattersall, M. B. Harris, S. D. McEvoy, D. N. Richerson, E. S. Deneris, R. L. Johnson, Z.-F. Chen, and G. B. Richerson Defects in Breathing and Thermoregulation in Mice with Near-Complete Absence of Central Serotonin Neurons J. Neurosci., March 5, 2008; 28(10): 2495 - 2505. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. K. Mulkey, D. L. Rosin, G. West, A. C. Takakura, T. S. Moreira, D. A. Bayliss, and P. G. Guyenet Serotonergic Neurons Activate Chemosensitive Retrotrapezoid Nucleus Neurons by a pH-Independent Mechanism J. Neurosci., December 19, 2007; 27(51): 14128 - 14138. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. B. Dias, T. B. Nucci, L. O. Margatho, J. Antunes-Rodrigues, L. H. Gargaglioni, and L. G. S. Branco Raphe magnus nucleus is involved in ventilatory but not hypothermic response to CO2 J Appl Physiol, November 1, 2007; 103(5): 1780 - 1788. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Nattie and A. Li Neurokinin-1 receptor-expressing neurons in the ventral medulla are essential for normal central and peripheral chemoreception in the conscious rat J Appl Physiol, December 1, 2006; 101(6): 1596 - 1606. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Li, S. Zhou, and E. Nattie Simultaneous inhibition of caudal medullary raphe and retrotrapezoid nucleus decreases breathing and the CO2 response in conscious rats J. Physiol., November 15, 2006; 577(1): 307 - 318. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. S. Moreira, A. C. Takakura, E. Colombari, and P. G. Guyenet Central chemoreceptors and sympathetic vasomotor outflow J. Physiol., November 15, 2006; 577(1): 369 - 386. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. F. Martino, M. R. Hodges, S. Davis, C. Opansky, L. G. Pan, K. Krause, B. Qian, and H. V. Forster CO2/H+ chemoreceptors in the cerebellar fastigial nucleus do not uniformly affect breathing of awake goats J Appl Physiol, July 1, 2006; 101(1): 241 - 248. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Sood, E. Raddatz, X. Liu, H. Liu, and R. L. Horner Inhibition of serotonergic medullary raphe obscurus neurons suppresses genioglossus and diaphragm activities in anesthetized but not conscious rats J Appl Physiol, June 1, 2006; 100(6): 1807 - 1821. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. C. T. Takakura, T. S. Moreira, E. Colombari, G. H. West, R. L. Stornetta, and P. G. Guyenet Peripheral chemoreceptor inputs to retrotrapezoid nucleus (RTN) CO2-sensitive neurons in rats J. Physiol., April 15, 2006; 572(2): 503 - 523. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. G. Guyenet, D. K. Mulkey, R. L. Stornetta, and D. A. Bayliss Regulation of Ventral Surface Chemoreceptors by the Central Respiratory Pattern Generator J. Neurosci., September 28, 2005; 25(39): 8938 - 8947. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. B. Richerson Re: Retrotrapezoid nucleus: a litmus test for the identification of central chemoreceptors Exp Physiol, May 1, 2005; 90(3): 253 - 257. [Full Text] [PDF] |
||||
![]() |
G. B. Richerson, W. Wang, M. R. Hodges, C. I. Dohle, and A. Diez-Sampedro Homing in on the specific phenotype(s) of central respiratory chemoreceptors Exp Physiol, May 1, 2005; 90(3): 259 - 266. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. R. Hodges, C. Opansky, B. Qian, S. Davis, J. M. Bonis, K. Krause, L. G. Pan, and H. V. Forster Carotid body denervation alters ventilatory responses to ibotenic acid injections or focal acidosis in the medullary raphe J Appl Physiol, April 1, 2005; 98(4): 1234 - 1242. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||