Blood flow is redistributed from the viscera to the periphery during periods of heat stress to maximize heat loss. The heat-induced redistribution of blood flow is strongly influenced by nonthermal inputs such as hydration status. At present, little is known about where thermal and nonthermal information is integrated to generate an appropriate effector response. Recently, the periventricular tissue that surrounds the anteroventral third ventricle (AV3V) has been implicated in the integration of thermal and osmotic information. The purpose of the present study was to determine the effects of electrolytic lesions of the AV3V on the cardiovascular response to a passive heat stress in unanesthetized, free-moving male Sprague-Dawley rats. Core temperature was elevated at a constant rate of ∼0.03°C/min in sham- and AV3V-lesion rats using an infrared heat lamp. Changes in mesenteric and hindquarter vascular resistance were determined using Doppler flow probes, and heat-induced salivation was estimated using the spit-print technique. The rise in mean arterial pressure (MAP), heart rate (HR), and mesenteric resistance in response to elevations in core temperature were all attenuated in AV3V-lesion rats; however, hindquarter resistance was unaffected. Heat-induced salivation was also diminished. In addition, AV3V-lesion rats were more affected by the novelty of the experimental environment, resulting in a higher basal core temperature, HR, and MAP. These results indicate that AV3V lesions disrupt the cardiovascular and salivatory response to a passive heat stress in rats and produce an exaggerated stress-induced fever triggered by a novel environment.
- mesenteric resistance
- Doppler flowmetry
- novel environment
the cardiovascular system plays a pivotal role in the regulation of core body temperature during heat stress. As core temperature rises, cardiac output is redistributed to allow increased cutaneous blood flow and heat dissipation (38). At the same time, vasoconstriction occurs throughout the visceral vasculature to maintain central blood volume (38). The heat-induced redistribution of blood flow is strongly influenced by nonthermal inputs such as hydration status. In a dehydrated state, the thermal threshold for cutaneous dilation and evaporative heat loss are delayed in both humans and experimental animals, leading to excessive heat storage, hyperthermia, and potentially heat illness (14, 29, 31, 44, 45). At present, little is known about the neural pathways regulating core temperature and where nonthermal inputs may be integrated to generate an appropriate effector response.
The periventricular tissue surrounding the anteroventral third ventricle (AV3V) has been implicated in the integration of thermal and osmotic information (15, 34, 50). The term AV3V refers to the anatomic region that encompasses the organum vasculosum of the lamina terminalis (OVLT), the ventral portion of the median preoptic nucleus (MnPO), and the preoptic periventricular nucleus (PPO) (7). This anatomically small region of the brain has been functionally implicated in multiple behavioral, neural, and hormonal controls involved in body fluid and cardiovascular homeostasis. Components of the AV3V are ideally connected with structures of the lamina terminalis and other parts of the central nervous system to act as a functional integrative site (17, 18, 40). The MnPO receives afferent input describing body fluid status from the OVLT and subfornical organ (SFO), as well as pressure/volume information via the nucleus of the solitary tract (NTS) (40). Reciprocal connections also exist between the MnPO and the preoptic anterior hypothalamus, which is generally considered to be the major thermoregulatory area within the central nervous system (40).
Neural pathways connecting the AV3V to the peripheral sympathetic nervous system have been described in a number of studies, implicating the AV3V in cardiovascular control (42, 53). Electrolytic lesions of this brain region abolish the pressor response normally seen after central or peripheral injections of ANG II or hypertonic saline and also prevent the development of several forms of experimentally induced hypertension (7). Electrical stimulation of the AV3V produces a shift in regional blood flow, similar to that seen during heat stress, increasing flow to the hindquarter, while simultaneously decreasing splanchnic and renal flows (9). Mangiapane and Brody (27) further refined the location of these vasoconstrictor and vasodilator sites when they demonstrated that stimulation of the OVLT increased resistance in the renal and splanchnic vasculature, whereas PPO or MnPO stimulation decreased resistance in the hindquarter. In addition to affecting the sympathetic response to changes in fluid balance, there is strong anatomic and physiological evidence to suggest that the AV3V is also vital for accompanying neuroendocrine responses (16, 41, 55).
The recent hypothesis suggesting thermal and nonthermal information may be integrated within the AV3V is based on a convergence of evidence from anatomical, electrophysiological, and functional studies. Levels of the protein Fos are elevated in the MnPO of heat-stressed or dehydrated rats (34). Furthermore, in vitro studies have shown that neurons within the MnPO are sensitive to changes in osmolality, temperature, and ANG II concentration (50). In a recent study, we used electrolytic lesions to demonstrate the importance of the AV3V in thermoregulation (54). Heat-induced salivation was greatly reduced in AV3V-lesion rats, and they were consequently unable to maintain their core temperature during heat stress.
Given the prominent role of the AV3V in both cardiovascular and thermal regulation, the purpose of this study was to assess the regional and systemic hemodynamic effects of AV3V lesions during a passive heat stress. Here, we demonstrate that, in addition to reducing salivation, AV3V lesions disrupt specific aspects of the cardiovascular response to elevations in core temperature.
MATERIALS AND METHODS
Thirteen male Sprague-Dawley rats (Harlan, Indianapolis, IN) weighing 270–370 g at the time of lesion were used for the experimental procedures. Animals were housed individually in hanging wire cages, at 22 ± 2°C, in a 12:12-h light-dark cycle (lights on 0600) with ad libitum access to food and water, except during the experiment. Rats were handled and accustomed to the experimental environment for 20 min each day for at least 10 days before the experiment. All experimental procedures were approved by the University of Iowa Animal Care and Use Committee.
AV3V lesion surgery.
Rats were anesthetized with Equithesin (0.33 ml/100 g body wt), secured in a stereotaxic apparatus, and the skull leveled between bregma and lambda. An electrode (24-gauge nichrome wire insulated except at the tip) was lowered on the midline, 0.2 mm posterior to bregma, 7.5 mm below dura, and anodal current (2.5 mA) passed for 20 s. In sham-lesion rats (n = 6), the electrode was lowered 6.5 mm, but no current was passed. To determine the effectiveness of the lesions, overnight water intake was measured 24 h postsurgery. Because lesions of the AV3V produce adipsia (7), we substituted a 10% sucrose solution for tap water and then gradually weaned the rats back to water over a 2- to 3-wk period. Fluid intake was determined daily, and any animal that had consumed ≤10 ml was supplemented with 10 ml of saline (37°C) subcutaneously. Rats were allowed 2–3 wk of recovery before Doppler flow probes were implanted.
Doppler flow probe surgery.
Rats were again anesthetized using Equithesin. A heparin-saline-filled catheter (PE-10 heat welded to PE-50) was implanted in the femoral artery, tunneled under the skin, and exteriorized at the back of the neck. Next, a midline laparotomy was performed, and 5-mm sections of the superior mesenteric artery and the lower abdominal aorta, immediately rostral to the common iliac bifurcation, were isolated. Doppler flow probes (Iowa Doppler Products, Iowa City, IA) were positioned around each of the arteries, and the wire leads were tunneled under the skin and exteriorized with the femoral catheter. Before closing the wound, a radiotelemeter (Barrows, Magalia, CA) was implanted for the measurement of peritoneal temperature (Tper). Rats were allowed 4 days of recovery before the experiment.
To minimize circadian variations in Tper, all experiments were initiated between 0845 and 0945. An initial Tper was determined before moving the rat from the animal room. The rat was then weighed to the nearest 0.1 g and placed in a circular Plexiglas cage (diameter 30 cm × height 30 cm); the catheter and flow probes were connected to the recording equipment. Once the animal was in the experimental cage, a 2-ml blood sample was withdrawn remotely from the arterial line for the measurement of hematocrit, plasma osmolality, sodium, and protein concentrations. A 0.5-ml bolus of warm (37°C), sterile saline was immediately returned to the rat while the remaining blood volume was replaced after resuspension of the red blood cells. The experiment began 1 h after the blood draw. The experimental protocol was divided into two parts, an initial 30-min control period (ambient temperature = 24°C) followed by a heat stress during which the animals were heated at a constant rate (0.033°C/min) until they reached a Tper = 41.5°C. Heating was accomplished using an infrared heat lamp positioned ∼70 cm above the animal. At the completion of the heat stress, the rat was reweighed, and the weight lost was corrected for fecal matter. A second 2-ml blood sample was obtained before the rat was deeply anesthetized (Nembutal, 150 mg/kg; Abbott Laboratories, Chicago, IL) and perfused transcardially with 0.01 M PBS followed by 10% Formalin. The brain was removed and stored in fixative until frozen sections (40 μm) were cut and stained for Nissl substance using cresyl violet.
Tper was determined every 2 min, and the lamp intensity altered accordingly to maintain the desired rate of heating. Relative levels of salivation were estimated by lining the floor of the Plexiglas cage with cresol red-stained filter paper (11). These “spit-prints” act as a pH indicator with saliva (alkaline) leaving a pink stain, while urine (acidic) produces a yellow-brown stain. The spit-prints were scanned (HP Officejet Pro), and the percent area stained pink was determined (Adobe Photoshop). The arterial catheter was connected to a pressure transducer, and mean arterial pressure (MAP) and heart rate (HR) were obtained from the pulsatile signal. Relative changes in regional blood flow velocities were determined by connecting the Doppler flow probes to a directional pulsed Doppler flowmeter (545C-4, Bioengineering, University of Iowa). Regional vascular resistances were expressed as a percent change from time 0 using the following formula: where Rt is the MAP divided by the mean velocity signal in the test period and Rc is the MAP divided by the mean velocity signal of the control period (22).
Plasma osmolality was determined using a freezing-point osmometer (model 5004, Precision Systems, Natick, MA). Plasma sodium concentrations were measured in a Na+/K+ analyzer with ion-specific electrodes (Nova 1, Nova Biomedical, Waltham, MA). Plasma protein was determined using a refractometer (National Instrument, Baltimore, MD). Hematocrit was determined using the microhematocrit method. All measurements were performed in triplicate.
Data have been normalized as 1) the percent time taken to reach a Tper of 41.5°C, to allow comparisons between animals that were heated for different durations, and 2) as the change in Tper. Data are expressed as means ± SE and were analyzed using a general linear regression model procedure for ANOVA with repeated measures, followed by a multiple-comparison test with Tukey's honestly significant difference test. Comparisons between sham- and AV3V-lesion rats at either initial or final time points were tested for significance using a Student's t-test. The level of significance for all tests was set at P < 0.05.
All animals with AV3V lesions exhibited damage to the OVLT, MnPO, and rostral periventricular tissue, as well as the various nuclei contained within the medial portion of the medial preoptic area (Fig. 1). As is typical of rats with AV3V lesions, the experimental group displayed a significant adipsia in the 24 h immediately following surgery (water intake, AV3V 2.0 ± 1.2 ml to sham 26.3 ± 11.3 ml, P < 0.05) and also gained less weight throughout the recovery period.
Tper. The initial Tper, taken before the animals were placed in the experimental cage, were no different between the two groups. However, after 1 h in the experimental environment, the Tper of AV3V-lesion rats had increased significantly, whereas that of the sham-lesion animals remained unchanged (Fig. 2). As a consequence of the higher Tper, the AV3V-lesion rats required a shorter duration of heating to reach a Tper of 41.5°C (Table 1). In addition, the thermal input required to elevate the Tper of AV3V-lesion animals was significantly less than that of sham-lesion rats (Table 1).
As already noted, rats with AV3V lesions initially weighed less than those with sham lesions (Table 1). By the end of the experiment, both groups of rats had lost a significant amount of weight through a combination of saliva, urine, and feces. When total water loss was expressed as a percentage of the initial body weight, sham-lesion animals were found to have lost twice as much as those with AV3V lesions (Table 1). This was reflected by the greater proportion of the sham-lesion rats' spit-prints that were stained with saliva (Fig. 3). Even when corrected for time, the sham-lesion rats still lost weight at a faster rate than those with AV3V lesions (sham 0.04 ± 0.002% loss/min to AV3V 0.03 ± 0.004% loss/min, P < 0.05).
AV3V-lesion rats had a significantly higher basal osmolality and sodium level, although there were no significant differences in plasma protein concentration or hematocrit between the groups (Fig. 4). The heating protocol produced a thermal dehydration in both groups of rats, as can be seen by the significant elevations in plasma osmolality and sodium level, whereas hematocrit and plasma protein values remained unchanged. When normalized for the duration of heating, the change in osmolality was significantly greater in sham-lesion rats compared with those with AV3V lesions (sham 0.15 ± 0.02 mosmol·kgH2O−1·min−1 to AV3V 0.08 ± 0.02 mosmol·kgH2O−1·min−1, P < 0.05). This reflects the significantly greater total water loss seen in sham-lesion animals.
Because of the differences in starting temperature, AV3V-lesion rats were heated for a shorter duration; therefore, changes in cardiovascular variables have been expressed in terms of both the percent time taken to raise Tper to 41.5°C (Fig. 5A) and the relative change in Tper (Fig. 5B). Irrespective of the method of presentation, the conclusions that can be drawn from the data are identical. Therefore, in the interest of clarity, only statistical data relating to the change in cardiovascular variables with respect to the change in Tper are presented.
Both groups increased mesenteric resistance as Tper rose (F5,45 = 32.10, P < 0.05, Fig. 5, A and B), although the increase was significantly attenuated in AV3V-lesion animals (group × temperature interaction F5,45 = 3.33, P < 0.05). Mesenteric resistance increased immediately in sham-lesion rats, even though their Tper had not changed (Fig. 5A). This reaction was absent in animals with AV3V lesions; however, after the initial elevation, mesenteric resistance increased to a similar degree in either group (F4,36 = 0.58, P = 0.68, Fig. 5, A and B).
AV3V- and sham-lesion rats exhibited an identical decline in hindquarter resistance (group × temperature interaction: F5,30 = 0.29, P = 0.91, Fig. 5, A and B). After the initial sharp drop, which occurred immediately after the heat lamp was switched on, hindquarter resistance remained unaltered by any further increases in Tper.
HR was significantly elevated in AV3V-lesion rats before heating (Table 1). Although elevations in Tper increased HR in both groups (F5,50 = 332.86, P < 0.05, Fig. 5, A and B), the change in HR was larger in sham-lesion rats (Fig. 5, A and B). At the completion of heating, HR tended to be higher in sham-lesion animals, although there was no significant difference in the final absolute value (Table 1). When data were expressed in relation to the change in Tper, no significant difference was found between the groups using an ANOVA; however, post hoc tests suggested that the change in HR was significantly greater in sham-lesion rats after the Tper had been raised 2.5°C (Fig. 5B).
MAP was also higher in AV3V-lesion rats before heating (Table 1). MAP showed a similar change, in response to heating, as that seen in mesenteric resistance. Animals with sham lesions exhibited a sharp rise in MAP at the onset of heating, which did not occur in those with AV3V lesions (Fig. 5A). This was then followed by a comparable rise in MAP in both groups as Tper increased (Fig. 5, A and B). There was a significantly greater change in the MAP of sham-lesion rats when the data were expressed either in relation to time (F1,11 = 14.72, P < 0.05, Fig. 5A) or the change in Tper (F1,11 = 5.05, P < 0.05, Fig. 5B); however, there was no difference in the final absolute MAP between the two groups (Table 1).
Cardiac output is redistributed during periods of heat stress to maximize heat loss. The redistribution of blood flow from the visceral to the cutaneous circulations is strongly influenced by nonthermal inputs such as hydration status (14, 29, 31, 36). Anatomic and in vitro evidence suggest that thermal and nonthermal information may be integrated within the AV3V (15, 34, 50). The aim of the present study, therefore, was to determine whether lesions of the AV3V have an impact on thermoregulatory responses to a passive heat stress. The principle findings were that AV3V lesions 1) disrupt selected aspects of the cardiovascular response to a passive heat stress, 2) attenuate heat-induced salivation, and 3) exaggerate and/or prolong the thermogenic response to a novel environment.
Previous studies have demonstrated that electrical stimulation of the AV3V (9), and, in particular, the OVLT (27), produces an increase in mesenteric resistance; therefore, the attenuated rise found in the AV3V-lesion rats was not unexpected. The elevation in mesenteric resistance, witnessed in the present study, can be divided into two distinct phases: an initial increase, apparently unrelated to Tper, followed by a linear, Tper-driven climb. The initial rise, which is absent in rats with AV3V lesions, occurs very rapidly, before any change in Tper, leading us to believe that it is driven primarily by thermal input from the skin. Skin temperature was not quantified in the current study; however, this hypothesis is supported by the apparent sawtooth pattern seen in the mesenteric resistance of only the sham-lesion rats (Fig. 5A). Peaks and valleys on the tracing correlated to the fluctuations in heat lamp intensity used to drive Tper up at a constant rate.
After the initial rise, mesenteric resistance increased with Tper in either group. This linear increase in mesenteric resistance is similar to that seen previously in both anesthetized and unanesthetized rats that were passively heated (21, 22). It is unclear why we saw an initial jump in the mesenteric resistance of our sham-lesion animals, whereas other studies have only seen the linear increase (21, 22). In the present study, animals were handled for at least 10 days before the experiment. Getting the rats accustomed to both the experimenter and the experimental environment appears to have reduced effects normally associated with a novel environment because the HR, MAP, and Tper of sham-lesion rats were all lower than those reported elsewhere (22). It may be that the initial increase in mesenteric resistance is related to the novel experience of the onset of heating. Thus it may not have been seen in previous studies of unanesthetized rats because the animals were in a novel environment and had already redistributed their blood flow as part of their defense response before the onset of heating. This would also explain why the initial increase is not seen in studies using an anesthetized preparation. As discussed below, rats with AV3V lesions appeared more affected by the experimental environment than sham-lesion animals. As such, the resistance in their mesenteric vasculature may have already been increased, preempting the additional stressor of the heat lamp.
Hindquarter resistance appears to be unaffected by AV3V lesions. This was surprising because stimulation of the MnPO produces hindquarter dilation (27). Neuronal pathways innervating the tail of the rat have been traced back to nuclei within the anterior hypothalamus, including the MnPO (43). Although there are hemodynamic changes in the tail in response to heating, it is possible that measuring changes in blood flow at the level of the abdominal aorta is not sensitive enough to detect changes in tail dilation. Tail dilation occurs in the rat as a result of the withdrawal of sympathetic nerve activity (33). This is, in turn, brought about through the inhibition of sympathetic premotor groups in the medullary raphe by input from the preoptic area (49). Stimulation of the rostral ventrolateral medulla also promotes tail constriction, yet this pressor output is not inhibited by the preoptic area (49). It is plausible then that although gross stimulation of the AV3V may promote hindquarter dilation, it is not necessarily involved during thermoregulation. The fact that rats with AV3V lesions are still capable of hindquarter vasodilation speaks to the specificity of the lesions generated in this study. Previously published work, in which the preoptic area has been ablated, has demonstrated the vital role of this brain region in thermoregulation (12). Furthermore, warming the preoptic anterior hypothalamus using thermodes stimulates robust tail dilation (37, 48).
Elevations in HR during passive heating appear to be due to the heat itself, rather than a cardiopulmonary baroreflex, as altering plasma volume does not alter the elevation in HR (47). Increasing core temperature may also directly affect the intrinsic pacemakers of the heart (19); however, the attenuated HR response, seen in the AV3V-lesion rats, demonstrates a central component to the heat-induced tachycardia. Presumably, this is due to disruption of fibers that pass through the periventricular tissue and medial portion of the medial preoptic area (1) because electrical stimulation of the OVLT and MnPO does not affect HR (27).
Despite being handled and exposed to the experimental setup on a daily basis, HR, MAP, and Tper were all elevated in AV3V-lesion rats after 1 h in the experimental cage. Although stress-induced fevers are routinely seen in animals upon exposure to a novel environment (32), they are normally attenuated after repeated exposure to the stressor (23). The elevation in Tper was unexpected, as a previous study, in which rats were handled for a similar duration, revealed no difference between the Tper of sham- and AV3V-lesion animals (54). The fact that the Tper of rats with sham lesions did not change suggests that the animals were adequately accustomed to handling and the experimental environment. In addition to Tper, novel situations are also known to elevate HR and MAP (28). Although we did not record the initial levels of these variables in the current study, previous work has consistently shown normotension or even slight hypotension in AV3V-lesion rats at rest (8, 13). This leads us to believe that the elevated baseline seen in AV3V-lesion animals was stress related. Interestingly, a similar novelty-induced elevation in HR and MAP has been observed in borderline hypertensive rats with AV3V lesions (39).
Immunocytochemical studies have demonstrated a prostaglandin-dependent increase in Fos expression within the AV3V, after exposure to a psychological stressor (6, 52). Current data suggest that these activated neurons are, in fact, part of an inhibitory feedback mechanism that limits the extent of the rise in core temperature. This is further supported by data describing an exaggerated activation of the hypothalamo-pituitary-adrenocortical axis in response to psychological (Whyte and Johnson, unpublished observations) and physiological stressors (4) in AV3V-lesion rats. Where the AV3V fits into the purported inhibitory neural network, however, remains open to speculation.
As a consequence of the elevated preheating Tper in the AV3V-lesion rats, data were described in two manners. In Fig. 5A, data are depicted as the percent time taken to reach a final Tper of 41.5°C, whereas in Fig. 5B data are normalized for the change in Tper. The latter approach assumes that the stress-induced rise in Tper is, in fact, a true fever, as opposed to a hyperthermia, and that the set point around which Tper is now being regulated has risen. This assumption appears to be valid, as stress-induced rises in Tper are similar in warm or cold environments (5, 25), are attenuated by antipyretic drugs (20), and are poorly correlated with changes in physical activity (26).
The evaporation of saliva groomed over the body surface constitutes a major component of the rat's heat defense, accounting for ∼85% of all heat loss at an ambient temperature of 40°C (10). Hubschle et al. (15) demonstrated an anatomic link between the AV3V and the salivary glands of the rat, and previous work from our laboratory has demonstrated that AV3V lesions disrupt heat-induced salivation (54). The data presented in the current study confirm our previous findings (54). A significantly larger area of the spit-prints of control rats were stained for saliva, and they lost a greater percentage of their starting body weight when compared with those with AV3V lesions. When the length of the heat stress was taken into account, total water loss was still larger in the sham-lesion rats, indicating that the greater total water loss was due to a higher level of salivation rather than a prolonged heat exposure.
As expected, AV3V-lesion rats were found to be hyperosmotic and hypernatremic compared with sham-lesion animals (7). Although total body water is not altered by AV3V lesions, plasma volume may be slightly reduced (3), although the reduction (0.3% of body weight) is probably too small to have a noticeable impact on thermoregulation. Elevations in the osmolality of plasma and cerebrospinal fluid are also known to inhibit evaporative heat loss and cutaneous vasodilation in response to a heat stress (45, 46, 51). However, given that AV3V lesions eliminate the vasopressin, thirst, and pressor responses to an elevation in plasma osmolality, it would seem logical to assume that the osmotic modulation of thermoregulatory effector mechanisms would also be disrupted in AV3V-lesion rats (7). Furthermore, dehydration generally results in an elevation in basal core temperature (2), which we did not see in the present study.
As with any electrolytic lesion study, data obtained from AV3V-lesion rats must be interpreted with caution, because it is unclear whether the observations are due to destruction of cell bodies within the lesion site itself or of fibers of passage. For example, fibers from the SFO course through the AV3V en route to the supraoptic and paraventricular nuclei (17, 18, 24). In all probability the AV3V-lesion interrupts these pathways, thereby altering magnocellular function. A second criticism of the AV3V-lesion protocol is that it destroys or damages several anatomically distinct nuclei. Although we feel this criticism is somewhat less valid in that it assumes functional specificity follows anatomic specificity, it does raise an important point. If the results observed in the present study were due to the destruction of a single population of neurons (e.g., MnPO), this would indicate a high degree of integration; however, the same results can just as easily be attributed to the simultaneous destruction of several independent neuronal populations within the AV3V. This latter interpretation is in line with the hypothesis that thermoregulatory effector mechanisms are independent of each other (30). Which of these two possibilities proves to be correct remains to be determined and will require the use of a more discrete lesion approach. That being said, the current study is the first to demonstrate a functional role for the AV3V in the cardiovascular response to a heat stress and forms a basis from which future studies can examine the impact that specific neuronal structures have on individual thermoregulatory effectors.
In summary, the present study provides functional evidence supporting the hypothesis that neural structures within the AV3V, possibly the MnPO, are involved in thermoregulation. Ablating the AV3V disrupts thermally induced salivation and elevations in mesenteric resistance, HR, and MAP, but not hindquarter flow. In addition, we have demonstrated that physiological responses to a mild stressor are exaggerated in AV3V-lesion animals.
D. G. Whyte is the recipient of an American Heart Association Heartland Affiliate Predoctoral Fellowship (0315230Z). This work was supported by ACSM Foundation Research Grant FRG19 and National Institutes of Health Grants HL-14388 and HL-57472.
We would like to thank Dr. Ralph Johnson for helping with the statistical analysis and Boyd Knosp for developing the protocol for the quantification of the spit-prints.
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.
- Copyright © 2005 the American Physiological Society