The Relationship of Lymphocytes in Blood and in Lymph to Sleep/Wake States in Sheep

Transcription

1 THE RELATIONSHIP OF LYMPHOCYTES IN BLOOD AND IN LYMPH TO SLEEP/WAKE STATES IN SHEEP The Relationship of Lymphocytes in Blood and in Lymph to Sleep/Wake States in Sheep J.B. Dickstein PhD, J.B. Hay PhD,* F.A. Lue M Eng, H. Moldofsky MD Centre for Sleep and Chronobiology, Toronto Western Hospital, University of Toronto, 399 Bathurst Street, Toronto, Ontario, M5T 2S8 and * University of Toronto, Departments of Immunology and Pathology, 1 Kings College Circle, Toronto, Ontario, M5S 1A8 Abstract: Based on evidence of a role for immune-associated cytokines in sleep induction, we investigated the possibility that lymphocyte distribution between blood and lymphatics could be altered as a function of sleep/wakefulness. Blood and lymph sample were obtained from 5 sheep during periods of slow-wave sleep and wake. Blood and lymph lymphocytes were phenotyped using monoclonal antibodies against CD4, CD8, gd T-cell receptors and a surface marker on ovine B cells. Lymph flow rates and efferent lymph cell output were measured. Lymph flow and prescapular efferent lymphocyte output were reduced during sleep compared to wakefulness (p<0.0005). There were no differences in lymphocyte subsets in the blood and in the lymph during sleep/wake brain states. These data indicate that migration of cells in the peripheral lymphatic system is altered during sleep compared to wakefulness. Key words: Slow-wave sleep; efferent lymph; lymph node; lymphocyte migration; sheep INTRODUCTION Accepted for publication April 1999 Correspondence: Harvey Moldofsky M.D., Centre for Sleep and Chronobiology, University of Toronto, Toronto Hospital- Western Division, MP , 399 Bathurst Street, Toronto, Ontario, M5T 2S8, Canada, Tel: (416) , Fax: (416) , h.moldofsky@utoronto.ca SLEEP, Vol. 23, No. 2, LYMPHOCYTES ARE UNIQUE AMONG BLOOD CELLS IN THEIR CAPACITY TO CONTINUALLY RECIRCULATE BETWEEN BLOOD AND LYMPH. 1 Recirculation is required for the detection and elimination of pathogens and for the dissemination of immunologic memory. 2 Lymphocyte migration from the blood to the lymph is influenced by a number of factors including the concentration of cells in the blood stream, the delivery of blood to tissues, the adhesive interactions between lymphocytes and vascular endothelium, and the retention of lymphocytes in tissues following their emigration from the blood stream. 3 The migration and distribution of lymphocytes can be influenced by the activity of the central nervous system. Sleep-related changes in cytokines, hormones, and neurotransmitters have the potential to alter migration patterns of lymphocytes. Correlations in cytokine production and changes in lymphocyte activity and distribution have been observed during normal periods of sleep and wakefulness. IL-1 activity and IL-1 beta are elevated in the cerebrospinal fluid of cats during slow-wave sleep. 4 In humans, IL-1- and IL-2- like activity is increased during sleep compared to wakefulness. 5,6 Pokeweed mitogen response is enhanced, while natural killer cell activity is decreased during sleep. 5 Furthermore, sleep causes a reduction in the number of circulating monocytes, NK cells, and lymphocytes in peripheral blood. 5,6 It has been suggested that the changes in the vasculature reflect a redistribution of lymphocytes between the blood and extravascular lymphoid and nonlymphoid tissue. Studies in humans are limited because they rely on the assessment of lymphocytes in the blood vascular system and not secondary lymphoid organs. Because only about 1% of the total number of lymphocytes in the body are found in the blood, information is required on lymphocyte distribution in the lymphatic system as a function of sleep state. Since it has been postulated that sleep is a host defence mechanism, our objective was to determine how sleep/wake behavior influences lymphocyte redistribution between blood and lymphatic compartments. We hypothesized that changes in lymphocyte subsets in the blood vasculature during sleep and wakefulness would be reflected in lymphatic compartments in sheep. METHODS Experimental Design and Protocol Five female sheep, six to eight months old weighing kg (Le-Do Sheep Farm, Ajax, ON, Canada) were used for these studies. Sheep were fed hay and pellets twice daily and had free access to water. Sheep were maintained on a 12:12 light-dark cycle 06:00-18:00 and housed

2 in metabolic cages. All handling and experimental procedures were performed in accordance with the Canadian Council on Animal Care and the Animal Care Committee at Sunnybrook Health Science Centre, Toronto, Ontario, Canada. All surgical procedures were performed under sterile conditions. Each sheep was initially anaesthetized with pentobarbital (15 to 25 mg/kg intravenously; Boehringer Ingelheim, Burlington, Ontario, Canada) and maintained under a general 1 2% halothane anaesthetic (Fluothane Ayerst Laboratories, New York). Each sheep underwent two surgeries. Surgeries were performed on two different occasions in order to ensure quality sleep recordings and adaptation to the environment before performing the lymphatic cannulation surgery. In the first surgery, electrodes were placed into the cranium approximately 8 mm each side of the mid-line and over the parietal and occipital regions of the cerebral cortex so that the electrodes touched the dura matter to measure electroencephalograph (EEG) activity. Electrodes were placed in the temporal bone to record electro-oculograph (EOG) activity. Wire electrodes were sutured into the superficial dorsal muscles of the neck to record electromyograph (EMG) activity. The electrodes were connected to a multiconnector socket that was fixed to the skull with acrylic cement. The sheep was placed in a metabolic cage after the surgery for acclimatization. After a one-week recovery period, chronic indwelling catheters were implanted into a jugular vein and into the efferent prescapular lymphatics. A catheter with a three way stopcock was surgically positioned in a jugular vein. Prescapular lymphatic vessels were isolated, exposed and ligated. A small incision was made in the lymphatic vessel wall and a vinyl catheter (Dural Plastics and Engineering, Auburn, Australia) was inserted into the lymph vessel and sutured firmly in place. 7 Lymphatic catheters were exteriorized and lymph was collected in sterile plastic bottles attached to the skin in the region of the lymphatic vessel. Table 1. Blood and lymph sampling protocol All animals were given 0.005mg/kg intramuscular buprenorphine HCl analgesic (Temgesic; Reckitt and Colman, Hull, UK) immediately after waking from surgery. Two days after the lymphatic surgery, the multiconnector socket affixed to each sheep's head was connected to a swivel fixed to the metabolic cage via a flexible tether (Plastics One, Renoke, VA, USA). Cables connected the swivel to a Grass 7D Polygraph in an adjacent room. This set-up allowed the sheep freedom of movement in the metabolic cage. In addition to the EEG, EOG, and EMG activity recorded by the polygraph, sheep were visually monitored throughout the experiment using a closed circuit TV camera and monitor. Blood and lymph were sampled from an adjacent room to minimize possible disturbances from the sheeps daily routine and from sleep. Blood samples were collected from the catheter positioned in the jugular vein via a 12 foot length of PE tubing connected to a blood sampling manifold system with a heparinized saline drip in the adjacent room. The continuous flow of lymph was collected via a sterile custom-made surface tension adhesion collection bottle. The lymph collecting device was flushed with heparinized saline between lymph samples. Sleep and wake samples were obtained when sheep exhibited the appropriate behavior. Wake blood and lymph samples were collected when sheep were standing and polygraph recording exhibited high frequency, low amplitude EEG. Slow-wave sleep (SWS) blood and lymph samples were collected when sheep were in the kneeling position and polygraph recordings exhibited low frequency, high amplitude EEG (criteria of Ruckebusch 8 ). Lymph samples were initiated after an animal had exhibited 10 minutes of continuous SWS or wakefulness. Lymph was collected for the next 10 minutes while an animal maintained a state of SWS or of wakefulness. In four of the five sheep, blood samples were obtained at the end of the lymph sample collection, after 20 minutes of SWS or of wakefulness. Only lymph was collected in sheep # 5. Phenotypic Sheep Blood phenotype samples Lymph phenotype samples Lymph cell count and flow rate samples Sleep wake sleep wake Sleep wake Sleep blood and lymph samples from sheep # 1,2,3 were collected during the dark period (18:00-21:00h) while wake blood and lymph samples were obtained in the light cycle (13:00-16:00h). In sheep #4, both sleep and wake blood and lymph samples were obtained in the light cycle (sleep 15:00-18:00h; wake 08:00-10:00h). In sheep #5, sleep lymph samples were collected during the dark (n=2) and the light (n=2) cycle. Wake lymph samples were collected during the light (n=2) and the dark (n=2) cycle. SLEEP, Vol. 23, No. 2,

3 analysis of lymphocytes was conducted in samples from sheep 1 4 while lymph cell output and flow rates were measured in samples from sheep 2 5 (see Table 1). The lymph and blood sampling protocol was determined through observations from a preliminary study in which we examined rest-activity patterns in three sheep and sleep/wake behavior in two sheep. Actigraphy data did not reveal a circadian rest-activity pattern. Rest episodes and activity occurred in both the light and dark period. Continuous baseline EEG recordings showed that each sheep exhibited 23% SWS and 2% REMS in the experimental period. SWS and REMS were distributed throughout the day and occurred in both the light and dark period. The largest amount of continuous SWS observed in the sheep was 32 minutes. Because we did not observe diurnal sleep/wake behavior, we compared lymphocyte distribution during sleep and wake periods. Processing of Blood and Lymph Samples Erythrocytes were lysed using Tris:NH4Cl solution. Blood lymphocytes and lymph lymphocytes were washed two times in Hank's Balanced Salt Solution (HBSS). Cells were immunostained in microtitre wells using standard procedures. Lymphocytes were phenotyped using monoclonal antibodies against CD4,CD8,γδ T-cell receptors, and Figure 1. Lymphocyte subsets in (A) the blood during periods of wake (open bars, n=9) and sleep (solid bars, n=13), (B) the lymph during periods of wake (open bars, n=10) and sleep (solid bars, n=14). Error bars indicate SD. No significant differences were observed in lymphocyte subsets CD4, CD8, γδ, and B cells in the blood or in lymph between wake and sleep states. a surface marker on ovine B cells (provided by Wayne Hein, Basel Institute of Immunology, Basel, Switzerland). Phycoerythrin-conjugated goat anti-mouse IgG or IgM (Cedarlane, Hornby, Ontario, Canada) were used for the detection of primary antibodies. Controls included unstained cells, and cells incubated only with the second antibody. All samples were fixed in 1% paraformaldehyde in phosphate buffered saline. FACscan flow cytometry (Becton Dickenson, San Jose, CA) was used to determine the phenotypic profile of the cells. The phenotypic profiles were determined by gating on small lymphocytes. 1 x 10 5 cells were counted. Cell counts were performed using a Coulter model Fn particle counter (Coulter Electronics, Hialeah, FL). Lymph volumes were measured and flow rates were calculated as cells/10 minutes. Statistics Means were computed on waking and sleeping samples for each sheep. Paired Students t tests were then conducted on these mean data. A value of p<0.05 was accepted as significant. RESULTS Lymphocyte Subsets in Lymph and Blood A total of 10 wake lymph and blood samples and 14 sleep lymph and blood samples were collected for phenotypic analysis (see Table 1). Two of the blood samples, one wake sample and one sleep sample, clotted and were discarded. The phenotypic profile of lymph and blood lymphocytes collected during wake was similar to the phenotypic profile collected during sleep. The percentage of CD4, CD8, γδ, and B cells in both blood and lymph (Figure 1) remained consistent between sleep and wake behavioral states. Changes in Lymphocyte Output and Lymph Flow Rates The data represent nine lymph samples collected during wake and 11 lymph samples collected during continuous SWS from four sheep (see Table 1). Lymph flow rates during sleep were significantly lower than lymph flow rates during wakefulness (p<0.0001) (Figure 2). Similarly, total cell output in 10 minutes was significantly reduced during SWS compared to wakefulness (p<0.0004) (Figure 3). There were no differences observed between sleep and wakefulness in the number of lymphocytes recovered in one milliliter of lymph (p>0.05). DISCUSSION These results demonstrate that during sleep, lymph flow and efferent cell output decreases significantly from the SLEEP, Vol. 23, No. 2,

4 wake state. The decrease in total cell output is a result of reduced lymph flow during sleep and not a decrease in cell concentration in the lymph. Similar decreases in lymphocyte output have been observed in humans during nighttime rest. 9 However, our hypothesis of state-dependent changes in immune cell type was not confirmed. Unlike human studies where changes in lymphocyte subsets were observed in the blood between sleep and wake states, 6 no significant differences in lymphocyte subset distribution were observed in the blood or in the lymph during sleep/wake states in the sheep. The distribution of lymphocytes in the blood and in the lymph in this study is consistent with the values reported in other studies using sheep. 10,11 The decrease in lymph flow observed during sleep can be explained by a decrease in autonomic nervous activity which occurs during sleep. During sleep, sympathetic nerve activity, blood pressure, heart rate, cardiac output, total peripheral resistance and blood flow to skeletal muscle groups are reduced. 12,13,14 Lymphatic vessels are also regulated by the autonomic nervous system. The propulsion of lymph is primarily controlled by intrinsic contractions of the smooth muscle of the lymphatic vessel. 15,16 Lymphatic vessels receive adrenergic innervations 17 and are influenced by neurogenic and humoral factors. Studies have shown that lymph flow is enhanced in response to adrenaline, noradrenaline, and electrical stimulation of regional lymphatics. These stimuli increase the frequency and the force of contraction of the lymphatic vessel. 18,19 Our observations of reduced lymph flow would be consistent with reduced sympathetic activity during SWS. It is also possible that the decrease in lymph flow rate can be reflected by a decrease in lymph production. Cote and Haddad 13 showed a reduction in blood flow to muscle cell output/10 minutes 6.E+07 4.E+07 2.E+07 0.E+00 wake behavioral state sleep Figure 2. Flow rate of efferent prescapular lymph during wake (open bars, n=9) and sleep (solid bars, n=11). Error bars indicate SD. Lymph flow during sleep was statistically different than that observed during wakefulness (p<0.0001). and other peripheral organs during slow-wave sleep. The decrease in regional blood flow and in vascular pressure would lessen the amount of fluid leaking out from the capillaries into the interstitial fluid. This reduction in capillary filtration would limit lymph production. The reduced delivery of fluid to the lymphatic vessels would decrease lymphatic contractile activity thereby decreasing lymph flow. Changes in regional blood flow can also affect lymphocyte migration from the blood to the lymph. Studies have shown that 90% of lymphocytes leaving the lymph node in efferent lymph come from post capillary blood venules within the lymph node. Furthermore, one out of every four lymphocytes from the blood migrates through the lymph node and is collected in efferent lymph. 20 The decrease in blood flow during slow-wave sleep could reduce lymphocyte delivery to the post capillary venules where lymphocytes migrate from the blood to the lymphatic system. These changes in the autonomic nervous system may explain the differences in lymph volumes and total lymphocyte output observed during sleep and wakefulness. Lymph cell output and lymph flow rates may also be modulated by cytokines that influence sleep/wake behavior. IL-1 and TNF-α administered intracerebroventricularly or intravenously induce excess SWS in rabbits and in sheep. 21,22,23 Furthermore, plasma levels of IL-1, TNF-α, and IL-2 are elevated in humans during sleep, 5,6,24,25 and Il- 1β is elevated in cat CSF during SWS. 4 These cytokines are potent immunomodulators that have the capacity to influence the migratory properties of lymphocytes. 26,27,28 Cytokines can also affect lymph flow. IL-1 29 and endotoxin 30 depress lymphatic pumping. Decreased lymph flow rates following endotoxin have been correlated to elevated TNF-α levels in vivo. 31 Furthermore, other endogenous mediators stimulated by TNF injection may be responsible for this effect. The procedure adopted in this particular study considered wake in the upright posture and sleep in the kneeling posture. We adopted this procedure based on our observations and those of Toutain 32 and Ruckebusch 8 that sheep only sleep in a kneeling position. Sheep spend most of there waking hours in an upright position, however, they can also be awake in the kneeling position. Sheep enter the drowsy state shortly after changing from the upright to the kneeling posture. The sheep may drift from a wake state to a drowsy state a number of times before entering a period of SWS. In this study, we were interested in comparing lymphocyte migration patterns during periods of SWS and wakefulness, but not the drowsy state. There is the possibility that changes in lymph flow may be posture dependent and independent of sleep cytokines, neuroendocrines, or neurotransmitters. In these experiments, we cannulated the prescapular lymphatics that predominantly drain the skin and the subcutaneous tissues of the neck and the torso. It is possible that the prescapular SLEEP, Vol. 23, No. 2,

5 flow may be impaired during sleep because the position of the foreleg in the kneeling position during sleep may kink the vessel. In our sheep model, the hydrostatic pressure of the lymph was expected to remain constant in the standing and kneeling positions since the lymph collecting device was positioned at the level of the lymph node. However, no measurements of pressure were carried out. Human studies have shown decreases in cell output in the horizontal position. 9 Future studies are important in determining whether posture or the physiological sleep state contributes to the decreased lymph flow observed in these experiments. While human studies have shown differences in lymphocyte subsets in the blood between sleep and wake states, we did not observe any changes in lymphocyte subset (CD4, CD8, γδ, and B cells) distribution in the blood or in the lymph during sleep and wake episodes in the sheep. We conducted this work on sheep because it is possible to cannulate individual lymphatic vessels and to simultaneously collect blood and lymph. This technique provides direct access to lymph lymphocytes with a minimum of mechanical manipulation. 33 However, sleep behavior in sheep differs from sleep behavior in humans. Unlike humans, sheep do not have eight hours of consolidated sleep. Ruckebusch showed that sheep confined to a metabolic cage sleep approximately four hours in each 24- hour period. 8 Sheep predominantly slept during the night; however, sleep was also exhibited during the day. The sleep patterns of sheep are episodic and of short duration. In this study, the short sleep/wake states do not influence lymphocyte subset distribution. It is possible that changes in lymphocyte distribution may occur following longer periods of uninterrupted sleep. There is the possibility that lymphocyte subset distribution is dependent on circadian rhythms, which was not addressed in this study. lymph flow (ml/10 minutes) wake behavioral state sleep Figure 3. Efferent prescapular lymph output during wake (open bars, n=9) and sleep (solid bars, n=11). Error bars indicate SD. Lymph output during sleep was statistically different than that observed during wakefulness (p<0.0004). Another possibility is that timing of sampling within the sleep and wake intervals may influence the detection of change in lymphocyte redistribution. In human studies of sleep and wakefulness, different sampling times of blood may be responsible for the inconsistent findings of changes in immune functions. Moldofsky et al. 34 and Irwin et al. 35 reported a decrease in NK activity in response to sleep deprivation while Dinges et al. 36 saw an increase in NK activity after 64 hours of wakefulness. Similarly, Moldofsky et al. 5 demonstrated elevated levels of IL-1 activity during SWS; however, this was not observed by Born et al. 6 and Entzian et al.. 37 The blood sampling protocols were different in each of the mentioned studies. Since lymphocyte migration and cytokine production are transient activities that are always in flux, it is possible that the subset redistribution was not detected by our sampling protocol. It is also possible that sleep might have influenced lymphocyte migration into non-lymphoidal tissue. This was not measured in our study but could be measured in future experiments using methods for chronic collection of afferent lymph. 10 The reduced lymph flow and cell output observed during sleep, regardless of mechanism, suggest that there is a decrease of lymphocytes trafficking through the node. The reduced number of cells passing through the node would be expected to compromise the recirculation of lymphocytes. This reduced flow rate may also delay antigen removal from the interstitial space and its delivery to lymph nodes. Since lymphocyte recirculation is required for immune surveillance, our results could imply that the immune system may be compromised during sleep. Alternatively, the decreased lymph flow and cell output observed during sleep could represent lymphocyte retention in the lymph nodes. Retention of lymphocytes in the nodes would in turn allow for greater interaction between the lymphocyte and antigen. This concept supports the notion that sleep plays a positive role in host defences. Future studies should be devised to test these opposing hypotheses. ACKNOWLEDGMENTS We thank Dr. D.A. Homonko for thoughtful discussion and critical review of the manuscript, and T. Seabrook, Dr. M. Boulton and W. P. Luk for technical assistance. This study was supported by the Toronto Psychiatric Research Foundation and the Barrie and District Myalgic Encephalomyelitis Support Group Inc. REFERENCES 1. Gowan JL. The recirculation of lymphocytes from blood to lymph in the rat. J Physiol 1959;146: Smith JB, Cunningham AJ, Laffery KJ, and Morris B. The role of the lymphatic system and lymphoid cells in the establishment of immunological memory. Aust J Exp Biol Med Sci 1970;48: Ottaway CA, Husband AJ. Central nervous system influences on lym- SLEEP, Vol. 23, No. 2,

6 phocyte migration. Brain Behav Immun 1992;6: Lue FA, Bail M, Jephthah-Ochola J, Carayanniotis K, Gorczynski R, and Moldofsky H. Sleep and cerebrospinal fluid interleukin-1-like activity in the cat. Int J Neurosci 1988;42: Moldofsky H, Lue FA, Eisen J, Keystone E, and Gorczynski R. The relationship of interleukin-1 and immune function to sleep in humans. Psychosom Med 1986;48: Born J, Lange T, Hansen K, Molle M, and Fehm HL. Effects of sleep and circadian rhythm on human circulating immune cells. J Immunol 1997;158: Lascelles AK, Morris B. Surgical techniques for the collection of lymph from unanaesthetized sheep. Q J Exp Physiol 1961;46: Ruckebusch Y. The relevance of drowsiness in the circadian cycle of farm animals. Animal Behav 1972;20: Engeset A, Sokolowski J, Olszewski WL. Variations in output of leukocytes and erythrocytes in human peripheral lymph during rest and activity. Lymphology 1977;10: Andrade WN, Johnston MG, Hay JB. The relationship of blood lymphocytes to the recirculating lymphocyte pool. Blood 1998;91: Mackay CR, Kimpton WG, Brandon MR, and Cahill NRP. Lymphocyte subsets show marked differences in their distribution between blood and the afferent and efferent lymph of peripheral lymph nodes. J Ex Med 1988;167: Mancia G, Bacelli G, Adams DB, Zanchetti A. Vasomotor regulation during sleep in the cat. Am J Physiol 1971;220: Cote A, Haddad GG. Effect of sleep on regional blood flow distribution in piglets. Pediatr Res 1990;28: Somers VK, Dyken ME, Mark AE, Abboud FM. Sympathetic-nerve activity during sleep in normal subjects. N Engl J Med 1993;328: Roddie IC, Mawhinney HJD, McHale NG, Kirkpartrick CT, Thornbury K. Lymphatic motility. Lymphol 1980;13: Zawieja DC, Davis KL, Schuster R, Hinds WM, Granger HJ. Distribution, propogation and coordination of contractile activity in lymphatics. Am J Physiol 1993; 264:H Alessandrini C, Gerli R, Sacchi G, Ibba L, Pucci AM, Fruschelli C. Cholinergic and adrenergic innervation of messenteric lymph vessels in guinea pig. Lymphol 1981;14: McHale NG, Roddie IC. The effect of intravenous adrenaline and noradrenaline infusion on peripheral lymph flow in the sheep. J Physiol 1983;341: McGeown JG, McHale NG, Thornbury KD. The effect of electrical stimulation of the sympathetic chain on peripheral lymph flow in the anaesthetized sheep. J Physiol 1987;393: Hay JB, Hobbs BB. The flow of blood to lymph nodes and its relation to lymphocyte traffic and the immune response. J Ex Med 1977; 145: Shoham S, Davenne D, Cady AB, Dinarello CA, Krueger JM. Recombinant tumor necrosis factor and interleukin 1 enhance slow wave sleep. Am J Physiol 1987;253:R Kapas L, Hong L, Cady AB, et al. Sonmogenic, pyrogenic, and anorectic activities of tumor necrosis factor-a and TNF-a fragments. Am J Physiol 1992;263:R Dickstein JB, Moldofsky H, Lue FA, Hay. Intracerebroventricular injection of TNF-a promotes sleep and is recovered in cervical lymph. Am J Physiol 1999;276:R Moldofsky H, Lue FA, Shahal B, Jiang CG, Gorczynski RM. Diurnal sleep/wake-related immune functions during the menstral cycle of healthy young women. J Sleep Res 1995;4: Darko DF, Miller JC, Gallen C, et al. Sleep electroencephalogram delta frequency amplitude, night plasma levels of tumor necrosis factor a, and human immunodeficiency virus infection. Proc Natl Acad Sci 1995;92: Kalaaji AN, McCullough K, Hay JB. The enhancement of lymphocyte localization in skin sites of sheep by tumor necrosis factor alpha. Immunol. Letters 1989;23: Issekutz TB, Stoltz JM. Stimulation of lymphocyte migration by endotoxin, tumor necrosis factor and interferon. Cell Immunol 1989; 120: Borron P, Hay JB. Characterization of ovine lymphatic endothelial cells and their interactions with lymphocytes. Lymphol 1994;27: Hanley CA, Elias RM, Movat HZ, Johnston MG. Suppression of fluid pumping in isolated bovine mesenteric lymphatics by interleukin- 1: interaction with prostaglandin E2. Microvasuclar Res 1989;37: Elias RM, Johnston MG, Hayashi A, Nelson W. Decreased lymphatic pumping after intravenous endotoxin administration in sheep. Am J Physiol 1987;253:H Sloan PJ, Elsasser TH, Spath Jr. JA, Albertine KH, Gee MH. Plasma tumour necrosis factor alpha during long term endotoxemia in awake sheep. J Appl Physiol 1992;73: Toutain PL, Toutain C, Webster AJF, McDonald JD. Sleep and activity, age and fatness and the energy expenditure of confined sheep. Br J Nutr. 1976;38: Abernethy NJ, Hay JB. The recirculation of lymphocytes from blood to lymph: physiological considerations and molecular mechanisms. Lymphol 1992;25: Moldofsky H, Lue FA, Davidson JR, Gorczynski R. Effects of sleep deprivation on human immune functions. FASEB 1989;3: Irwin M, Mascovich A, Gillin JC, Willoughby R, Pike J, Smith TL. Partial sleep deprivaiton reduces natural killer cell activity in humans. Psychosom Med 1994;56: Dinges DF, Douglas SD, Zaugg L, et al. Leukocytosis and natural killer cell function parallel neurobehavioural fatigue induced by 64 hours of sleep deprivation. J Clin Invest 1994; 93: Entzian P, Linnemann K, Schlaak M, Zabel P. Obstructive sleep apnea syndrome and circadian rhythms of hormones and cytokines. Am J Respir Crit Care Med 1996;153: SLEEP, Vol. 23, No. 2,