Am J Physiol Endocrinol Metab 290: E199-E203, 2006.
First published August 16, 2005; doi:10.1152/ajpendo.00337.2005
0193-1849/06 $8.00
TRANSLATIONAL PHYSIOLOGY
Steroidogenesis in human aldosterone-secreting adenomas and adrenal hyperplasias: effects of hypoxia in vitro
Hershel Raff,
Eric D. Bruder the St. Luke's Medical Center Adrenal Tumor Study Group
Endocrine Research Laboratory, St. Luke's Medical Center, Medical College of Wisconsin, Milwaukee, Wisconsin
Submitted 25 July 2005
; accepted in final form 13 August 2005
 |
ABSTRACT
|
|---|
The synthesis of adrenal steroids requires molecular oxygen. Because arterial hypoxemia is a common clinical condition, the purpose of the present study was to examine steroidogenesis in vitro under physiological changes in O2 tension (PO2) in cells from human adrenal glands with aldosterone-secreting adenomas (ASA; n = 3) or with bilateral adrenal hyperplasia causing Cushing's syndrome (n = 4). A decrease in PO2 from 150 mmHg (mild hyperoxia) to 80 mmHg had minimal effect on steroid production. A reduction to 40 mmHg (still well within the physiological range) significantly inhibited cAMP- and ACTH-stimulated aldosterone, cortisol, and dehydroepiandrosterone (DHEA) production from ASA. Furthermore, cortisol and DHEA production in cells from histologically normal tissue, adjacent to ASA and from bilateral adrenal hyperplasias, was also inhibited under a PO2 of 40 mmHg. We conclude that physiological decreases in PO2 to levels typical for adrenal venous PO2 under mild hypoxia inhibit steroidogenesis. These studies may have implications for oxygen therapy in critically ill patients with functional adrenal insufficiency, as well as for therapeutic options in patients with adrenal neoplasms.
oxygen; cortisol; dehydroepiandrosterone; adrenal cortex
PRIMARY INCREASES IN ADRENAL FUNCTION are usually caused by a neoplasm [e.g., aldosterone-secreting adenomas (ASA)] or hyperplasia (e.g., bilateral adrenal hyperplasia) (14). The primary treatment for solitary tumors is usually surgical removal, whereas therapy for bilateral hyperplasia is more complex (14). A number of studies have characterized the expression and function of these different abnormal adrenal tissues, but none have evaluated their sensitivity to oxygen delivery (5, 8, 10, 13, 25).
Steroidogenesis is catalyzed by enzymes, most of which are cytochrome P-450 in nature and therefore require molecular oxygen to sequentially transform cholesterol into the final products [i.e., cortisol, aldosterone, dehydroepiandrosterone (DHEA) in the adrenal gland]. Our previous studies using dispersed bovine, rabbit, and rat adrenal cells have found steroidogenesis to be sensitive to changes in oxygen levels in the physiological range (3, 15, 18). Our original studies in bovine adrenal cells found that aldosterone production was much more sensitive to decreases in oxygen compared with cortisol production (18). We had proposed that a decrease in aldosterone during hypoxia would have beneficial diuretic effects, although the maintenance of cortisol production would be necessary to sustain metabolic and vascular function (16, 17, 19). Others have also demonstrated O2-sensitive steroidogenesis in a variety of in vitro steroidogenic cell models (4, 6, 9, 23).
Finally, the role of hypoxia in solid tumors and the interaction of oxygen delivery and chemotherapy have been extensively evaluated (24, 26). It has been suggested that manipulation of oxygen sensitivity with chemical agents may play a role in the treatment of solid tumors (21, 26, 27). Therefore, this initial study has evaluated the sensitivity of dispersed cells from adenomas and histologically normal adjacent tissue, as well as from bilateral adrenal hyperplasia, to a decrease in PO2 within the physiological range as well as to mild hyperoxia.
 |
METHODS
|
|---|
Tissue source and isolation of cells.
This study was approved by the Institutional Review Board of Aurora Health Care, and all patients gave signed informed consent. Adrenal glands were obtained during clinically indicated laparoscopic unilateral or bilateral adrenalectomy. Tissue was obtained from patients with either clinically established and biochemically verified primary aldosteronism due to ASA (n = 3) or Cushing's syndrome due to bilateral adrenal hyperplasia (BAH; n = 4). Resected tissue was immediately examined by the pathologist. Adrenals from patients with ASA were sectioned and then designated as grossly normal adjacent adrenocortical or nodular tissue. Patients with BAH were identified preoperatively, and the diagnosis was confirmed by the pathologist in the operating room (in all cases, final histopathology confirmed the tissue assignments). After sectioning, tissue was placed in ice-cold Krebs-HEPES-calcium buffer containing bovine serum albumin (1 mg/ml). Tissue samples were then digested with collagenase (4 mg/ml) in fresh BSA buffer (37°C). Cells were harvested every 30 min during a 90-min incubation. Each batch of cells was counted and assessed for viability by trypan blue exclusion, counting only cells with visible lipid droplets. The dispersed cells were then placed in fresh buffer and diluted to an experimental concentration of 100,000 cells/ml.
Assessment of steroidogenesis in vitro.
An aliquot of newly dispersed cells was placed in 5-ml polystyrene test tubes. Experimental treatments included the following: control, ACTH (0.2 or 2.0 ng/ml), or dibutyryl cAMP (1.0 mM). All treatments were performed in triplicate. Cells were incubated for 2 h under metal tents in a shaker-water bath at 37°C. These tents were vented with room air (21% O2) or low O2 (10% or 0%) in N2. Because the metal tents are a flow-through system, room air was not completely excluded. Therefore, the resultant O2 tension (PO2) levels were 150 mmHg for 21% O2, 80 mmHg for 10% O2, and 40 mmHg for 0% O2 measured by oxygen electrode (Radiometer ABL2). The experiment was stopped by placing the tubes on ice and then centrifuging the cells at 4°C. Supernatants were frozen for further analysis.
Measurement of steroids.
Cortisol was measured by radioimmunoassay (RIA; Diagnostic Products, Los Angeles, CA) as used previously (18). DHEA was measured by enzyme immunoassay (Diagnostic Systems Laboratories, Webster, TX). Aldosterone was measured by RIA as used previously (3). The accumulation of steroid in the cell medium (final concentration) was used as an assessment of steroid synthesis and release from the cell.
Chemical reagents.
HEPES, BSA, and cAMP were purchased from Sigma Chemical (St. Louis, MO). Collagenase was purchased from Worthington Biochemical (Freehold, NJ). ACTH-(139) was purchased from Peninsula Laboratories (Belmont, CA). All other chemicals were of reagent grade and were purchased from Sigma or Fisher Scientific (Fair Lawn, NJ).
Statistical analysis.
All data were analyzed by two-way ANOVA for repeated measures and Duncan's multiple range test for multiple comparisons (SigmaStat 2.03). P < 0.05 was considered significant. Data are presented as means ± SE. Basal and peak steroid production from adenomas and adjacent tissue varied quite widely between experiments. Therefore, these results are expressed as a percentage of basal steroid production at 21% O2. Basal steroid production was compared between different cell types by ANOVA on ranks (SigmaStat 2.03). Data from BAH tissue were analyzed after logarithmic transformation and are presented on a logarithmic scale.
 |
RESULTS
|
|---|
Table 1 shows basal (control) steroid concentrations in the supernatant of incubated cells at 21% O2 (PO2 = 150 mmHg). Aldosterone levels were undetectable in supernatants from normal tissue adjacent to ASAs and in three of four cell preparations from BAH. There were no significant differences in basal steroid concentrations between cell types. The high variability between cell preparations led us to normalize the data from ASA and adjacent tissue to the control steroid concentrations shown in Table 1.
Figure 1 depicts aldosterone production by cells from aldosterone-producing adenomas studied in vitro. Basal aldosterone production was not inhibited by a decrease in PO2 to 80 or 40 mmHg. cAMP (1.0 mM) and both concentrations of ACTH significantly increased aldosterone production in the control (150 mmHg) oxygen exposure. ACTH and cAMP-stimulated aldosterone production was inhibited by a decrease in PO2 to 40 mmHg. There were very low or undetectable levels of aldosterone production from cells obtained from tissue adjacent to the adenomae.

View larger version (29K):
[in this window]
[in a new window]
|
Fig. 1. Aldosterone responses as %control (no secretagogue; 150 mmHg) to cAMP or ACTH in cells from aldosterone-secreting adenomas (n = 3). Cells were incubated under a PO2 of 150 mmHg, 80 mmHg, or 40 mmHg. +Significant increase from control under the same PO2; *significant inhibition compared with the same secretagogue concentration under 150 mmHg.
|
|
Figure 2 depicts cortisol production by cells from ASA (Fig. 2A) and histologically normal tissue adjacent to these adenomas (Fig. 2B). ACTH (2.0 ng/ml) and cAMP both elicited significant increases in cortisol production at all levels of oxygen studied. The lower concentration of ACTH did not result in a significant increase in cortisol except under a PO2 of 150 mmHg. The inhibition of the cortisol response to either cAMP or ACTH (2.0 ng/ml) was related to the decrease in oxygen. Comparison of cortisol production in cells from adjacent tissue (Fig. 2B) with cells from aldosterone-producing adenomae (Fig. 2A) demonstrated that the magnitude of the cortisol response to cAMP was greater than the cortisol response to ACTH in adenomae but not in adjacent tissue. However, cells from both adjacent tissue and adenomas were relatively resistant to a decrease in PO2 to 80 mmHg, whereas adenomas showed a much larger decrease in cAMP-stimulated cortisol production under a PO2 of 40 mmHg.

View larger version (22K):
[in this window]
[in a new window]
|
Fig. 2. Cortisol responses as %control (no secretagogue; 150 mmHg) to cAMP or ACTH in cells from aldosterone-secreting adenomas (A; n = 3) and from tissue adjacent to aldosterone-secreting adenomas (B; n = 3). Cells were incubated under a PO2 of 150 mmHg, 80 mmHg, or 40 mmHg. +Significant increase from control under the same PO2; *significant inhibition compared with the same secretagogue concentration under 150 mmHg.
|
|
Figure 3 depicts DHEA production from cells from ASA (Fig. 3A) and adjacent tissue (Fig. 3B). cAMP significantly increased DHEA production from ASA at all levels of oxygen (Fig. 3A). Only the lowest PO2 (40 mmHg) led to an inhibition of DHEA production from cells from ASA. Cells from adjacent tissue (Fig. 3B) displayed similar patterns in DHEA production (under each treatment) compared with cortisol (Fig. 2). Although we had only three replications in this experiment, it is noteworthy that ACTH-stimulated DHEA production from adjacent tissue under a PO2 of 80 mmHg was less than that under a PO2 of 150 mmHg.

View larger version (23K):
[in this window]
[in a new window]
|
Fig. 3. Dehydroepiandrosterone (DHEA) responses as %control (no secretagogue; 150 mmHg) to cAMP or ACTH in cells from aldosterone-secreting adenomas (A; n = 3) and from tissue adjacent to aldosterone-secreting adenomas (B; n = 3). Cells were incubated under a PO2 of 150 mmHg, 80 mmHg, or 40 mmHg. +Significant increase from control under the same PO2; *significant inhibition compared with the same secretagogue concentration under 150 mmHg.
|
|
The graph in Fig. 4 represents cortisol and DHEA production by cells isolated from BAH tissue. Basal cortisol production (control) was significantly decreased when the PO2 was decreased to 40 mmHg. Unlike cells from the adrenal adenoma experiments, ACTH (0.2 ng/ml)-stimulated cortisol production from BAH cells was significantly inhibited by decreases in oxygen to 80 and 40 mmHg. Cells from BAH tissue were able to significantly increase cortisol production in response to cAMP and ACTH (2.0 ng/ml) under a PO2 of 40 mmHg; however, these increases were attenuated compared with incubation under a PO2 of 150 mmHg. Changes in DHEA production (Fig. 4B) in response to stimuli or oxygen level nearly mimicked those found for cortisol production. Within each treatment, decreasing PO2 to 40 mmHg significantly reduced DHEA production. DHEA production in response to ACTH was not significantly increased at a PO2 of 40 mmHg compared with basal levels at the same PO2.

View larger version (26K):
[in this window]
[in a new window]
|
Fig. 4. Cortisol (A) and DHEA (B) concentrations without (control) or with cAMP or ACTH in cells from bilateral adrenal hyperplasias (n = 4) under a PO2 of 150 mmHg, 80 mmHg, or 40 mmHg. +Significant increase from control under the same PO2; *significant inhibition compared with the same secretagogue concentration under 150 mmHg. Note that the y-axis is a logarithmic scale.
|
|
 |
DISCUSSION
|
|---|
This study evaluated steroidogenesis in cells from human adrenal glands under a physiological range of PO2. Decreases in PO2 to 40 mmHg resulted in a decrease in the production of all steroids measured in cells from aldosterone-secreting adenomas from histologically normal tissue adjacent to aldosterone-secreting adenomas and bilateral adrenal hyperplasias. Hyperoxia (150 mmHg) did not consistently increase steroid production compared with 80 mmHg. Of note was that aldosterone-secreting adenomas released large amounts of cortisol and DHEA in addition to aldosterone and that all steroids produced from adenomas were dramatically stimulated by cAMP. This cAMP-stimulated steroid production was partially inhibited when cells were incubated under a PO2 of 40 mmHg. Also of note, but not unexpected, was that histologically normal tissue adjacent to aldosterone-secreting adenomas did not release measurable levels of aldosterone, probably because of long-term suppression of renin, commonly found in patients with primary hyperaldosteronism (12).
First, a comment about the levels of PO2 chosen for these experiments. We were only interested in changes in PO2 within the pathophysiological range as well as under slightly hyperoxic conditions. The adrenal gland has very high blood flow per gram of tissue (1, 2). Even though it uses molecular oxygen for steroidogenesis, as well as for the maintenance of cellular function, like most endocrine organs, adrenal blood flow is considerably higher than required to maintain oxygen consumption. This is obviously to maximize hormone delivery to the systemic circulation. Measurement of adrenal venous PO2 (an index of tissue oxygen levels) in the dog has demonstrated levels ranging from 70 mmHg under normoxic conditions to 20 mmHg under severe, acute hypoxia (1, 2). This suggests that the range of adrenal PO2 in our study is right at critical patho-physiological levels of oxygen, from mildy hyperoxic to roughly normoxic (80 mmHg) to a level that one would expect with moderate, clinical hypoxia (40 mmHg).
We have previously demonstrated that bovine, rabbit, and rat adrenocortical cells are sensitive to changes in PO2 within the physiological range (3, 15, 18). This is the first study, to our knowledge, to evaluate this phenomenon in cells from human adrenal tissue. A major difference from these previous studies to the human cells in the present study was the high sensitivity of cortisol production to modest decreases in oxygen, because our previous study in bovine cells showed that aldosterone production was more sensitive to inhibition under low oxygen compared with cortisol (18).
It is also interesting to point out that aldosterone-secreting adenomas produced large amounts of cortisol and DHEA in the basal state and dramatically increased production of these steroids with stimulation with cAMP. Although patients with aldosterone-secreting adenomas do not usually have clinical hypercortisolism or hyperandrogenism (14), this in vitro phenomenon has been described previously (13, 25) and has the potential to be helpful diagnostically (8).
Physiological implications.
We were unable to obtain tissue from truly normal adrenal glands because they are no longer routinely removed during unilateral "nephron-sparing" nephrectomy (22). Therefore, we must extrapolate from histologically normal tissue adjacent to adrenal adenomas. In all cases, steroidogenesis was very sensitive to modest decreases in PO2, whereas hyperoxia tended to, but did not consistently, increase steroid production. We previously hypothesized that decreases in aldosterone during hypoxia might allow a beneficial natriuresis and diuresis to prevent edema (16, 17, 19). It is remarkable, therefore, that cortisol and DHEA production were also inhibited by a decrease in PO2 to 40 mmHg, and we do not yet have a teleological explanation for this. Whether this would occur during hypoxia in vivo in patients with primary increases in adrenal steroidogenesis is not known.
Translational physiology.
One aspect of this study relevant to clinical medicine is critical illness. It is well known that critically ill patients (e.g., with sepsis) studied in an intensive care unit can have diminished adrenal function (20). The diminished adrenal function in septic, critically ill patients is more evident when they are stimulated with physiological doses of ACTH (11). One potential explanation for this that is relevant to the present study is the possibility that these critically ill patients have diminished oxygen delivery to the adrenal glands. Therapy focused on improving oxygen delivery could therefore potentially improve steroidogenesis in critically ill patients.
The sensitivity of steroidogenesis in human adrenals to modest decreases in oxygen may have implications in potential treatment options, particularly since it appears that the relatively hyperactive adenomas were inhibited by a PO2 of 40 mmHg. Manipulation of oxygen delivery to solid tumors as an adjunct to chemotherapy has been of interest for many years (21, 26, 27). In fact, there are now adjuvant therapies that are targeted to alter the sensitivity to oxygen delivery. The high endogenous sensitivity to hypoxia that we have demonstrated may suggest that adrenal tumors might be amenable to this chemotherapeutic approach. Although we did not study adrenal carcinoma, this difficult-to-treat, and often fatal, cancer may also be an appropriate target for manipulation of oxygen sensitivity during chemotherapy.
In conclusion, steroidogenesis in aldosterone-secreting adenomas, bilateral adrenal hyperplasias, and normal adrenocortical tissues appears to be sensitive to modest decreases in PO2. It remains to be seen whether this can be exploited in the therapy of adrenal neoplasms.
 |
ACKNOWLEDGMENTS
|
|---|
We acknowledge the members of the St. Luke's Medical Center Adrenal Tumor Study Group: D. E. Klinger (Surgery); J. W. Findling, B. M. Lalande, and S. B. Magill (Endocrine-Diabetes Center); G. F. Neitzel, P. J. Rykwalder, M. I. Malik, J. G. Pelligrini, T. M. Wallace, W. G. Doos, J. E. Williams, and M. M. Anderson (Pathology).
 |
FOOTNOTES
|
|---|
Address for reprint requests and other correspondence: H. Raff, St. Luke's Physician's Office Bldg., 2801 W KK River Pkwy, Suite 245, Milwaukee, WI 53215 (e-mail: hraff{at}mcw.edu)
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
|
|---|
- Breslow MJ, Ball TD, Miller CF, Raff H, and Traystman RJ. Adrenal blood flow and secretory relationships during hypoxia in anesthetized dogs. Am J Physiol Heart Circ Physiol 257: H1458H1465, 1989.[Abstract/Free Full Text]
- Breslow MJ, Tobin JR, Mandrell TD, Racusen LC, Raff H, and Traystman RJ. Changes in adrenal oxygen consumption during catecholamine secretion in anesthetized dogs. Am J Physiol Heart Circ Physiol 259: H681H688, 1990.[Abstract/Free Full Text]
- Bruder ED, Nagler AK, and Raff H. Oxygen-dependence of ACTH-stimulated aldosterone and corticosterone synthesis in the rat adrenal cortex: developmental aspects. J Endocrinol 172: 595604, 2002.[Abstract]
- Chabre O, Defaye G, and Chambaz EM. Oxygen availability as a regulatory factor of androgen synthesis by adrenocortical cells. Endocrinology 132: 255260, 1993.[Abstract]
- Fujita H, Shibata H, Ogata E, and Kojima I. In vitro responsiveness of aldosterone-producing adenoma to angiotensin II, K and ACTH. Endocrinol Jpn 37: 563570, 1990.[Medline]
- Gabbe SG and Villee CA. The effect of hypoxia on progesterone synthesis by placental villi in organ culture. Am J Obstet Gynecol 111: 3137, 1971.[Medline]
- Gandara DR, Lara PN Jr, Goldberg Z, Le QT, Mack PC, Lau DH, and Gumerlock PH. Tirapazamine: prototype for a novel class of therapeutic agents targeting tumor hypoxia. Semin Oncol 29: 102109, 2002.[CrossRef][ISI][Medline]
- Grondal S, Grimelius L, Thoren M, and Hamberger B. Basal and ACTH-stimulated cortisol and aldosterone release from adrenocortical adenomas in vitro. Eur J Surg 157: 179183, 1991.[Medline]
- Hanke CJ and Campbell WB. Endothelial cell nitric oxide inhibits aldosterone synthesis in zona glomerulosa cells: modulation by oxygen. Am J Physiol Endocrinol Metab 279: E846E854, 2000.[Abstract/Free Full Text]
- Hirata Y, Tomita M, Yoshimi H, Kuramochi M, Ito K, and Ikeda M. Effect of synthetic human atrial natriuretic peptide on aldosterone secretion by dispersed aldosterone-producing adenoma cells in vitro. J Clin Endocrinol Metab 61: 677680, 1985.[Abstract]
- Marik PE and Zaloga GP. Adrenal insufficiency during septic shock. Crit Care Med 31: 141145, 2003.[CrossRef][ISI][Medline]
- Mulatero P, Stowasser M, Loh KC, Fardella CE, Gordon RD, Mosso L, Gomez-Sanchez CE, Veglio F, and Young WF Jr. Increased diagnosis of primary aldosteronism, including surgically correctable forms, in centers from five continents. J Clin Endocrinol Metab 89: 10451050, 2004.[Abstract/Free Full Text]
- Naruse M, Obana K, Naruse K, Yamaguchi H, Demurea H, Inagama T, and Shizume K. Atrial natriuretic polypeptide inhibits cortisol secretion as well as aldosterone secretion in vitro from human adrenal tissue. J Clin Endocrinol Metab 64: 1016, 1987.[Abstract]
- Orth DN, Kovacs WJ, and DeBold DR. The adrenal cortex. In: Williams Textbook of Endocrinology, edited by Wilson JD and Foster DW. Philadelphia: WB Saunders, 1992.
- Papanek PE, Jankowski BM, and Raff H. Aldosterone release from adrenal cells is inhibited by hypoxia in vitro during maturation in rabbits. Reprod Fertil Dev 8: 11311136, 1996.[CrossRef][Medline]
- Raff H. The renin-angiotensin-aldosterone system during hypoxia. In: Response and Adaptation to HypoxiaOrgan to Organelle, edited by Lahiri S, Cherniak N, and Fitzgerald RS. New York: Oxford Univ. Press, 1991, p. 211222.
- Raff H. Endocrine adaptation to hypoxia. In: Handbook of Physiology. Environmental Physiology. Bethesda, MD: Am. Physiol. Soc., 1996, sect. 4, vol. II, chapt. 54, p. 12591275.
- Raff H, Ball DL, and Goodfriend TL. Low oxygen selectively inhibits aldosterone secretion from bovine adrenocortical cells in vitro. Am J Physiol Endocrinol Metab 256: E640E644, 1989.[Abstract/Free Full Text]
- Raff H, Brickner RC, and Jankowski B. The renin-angiotensin-aldosterone system during hypoxia: Is the adrenal an oxygen sensor? In: Hypoxia and Mountain Medicine, edited by Sutton JR, Coates G, and Houston CS. Oxford and New York: Advances in the Biosciences, Pergamon, 1992, vol. 84, p. 4249.
- Raff H and Findling JW. Aldosterone control in critically ill patients: ACTH, metoclopramide, and atrial natriuretic peptide. Crit Care Med 18: 915920, 1990.[Medline]
- Rockwell S. Oxygen delivery: implications for the biology and therapy of solid tumors. Oncol Res 9: 383390, 1997.[Medline]
- Steinberg AP, Kilciler M, Abreu SC, Ramani AP, Ng C, Desai MM, Kaouk JH, and Gill IS. Laparoscopic nephron-sparing surgery for two or more ipsilateral renal tumors. Urology 64: 255258, 2004.[Medline]
- Stevens VL, Aw TY, Jones DP, and Lambeth JD. Oxygen dependence of adrenal cortex cholesterol side chain cleavage. J Biol Chem 259: 11741179, 1984.[Abstract/Free Full Text]
- Stone HB, Brown JM, Phillips TL, and Sutherland RM. Oxygen in human tumors: correlations between methods of measurement and response to therapy. Radiat Res 136: 422434, 1993.[ISI][Medline]
- Stowasser M, Tunny TJ, Klemm SA, and Gordon RD. Cortisol production by aldosterone-producing adenomas in vitro. Clin Exp Pharmacol Physiol 20: 292295, 1993.[Medline]
- Vaupel P, Thews O, and Hoeckel M. Treatment resistance of solid tumors. Med Oncol 18: 243259, 2001.[CrossRef][ISI][Medline]
- Wouters BG, Weppler SA, Koritzinsky M, Landuyt W, Nuyts S, Theys J, Chiu RK, and Lambin P. Hypoxia as a target for combined modality treatments. Eur J Cancer 38: 240257, 2002.[CrossRef][ISI][Medline]
Copyright © 2006 by the American Physiological Society.