This review presents a schematic attempt to classify the major pain pathways, based on the results of recent studies in our laboratory, with a special emphasis on the parabrachial system. Our view is based on results from experiments in the rat, using very small iontophoretic injections of anterograde tracers. As illustrated in this report, we have found a very dramatic difference between ascending projections originating from deep laminae compared with those arising from lamina I of the dorsal horn. We propose three main pain systems and discuss their functional-anatomical relationships. The first system is centred on the projection from deep laminae to three caudal reticular areas - the lateral reticular nucleus (LRN), the subnucleus reticularis dorsalis (SRD) and the gigantocellular lateral paragigantocellular reticular nuclei (NGc) - and the parabrachial internal lateral subnucleus (PBil). The second system is centred on the projection from lamina I to the ventral posterolateral nucleus (VPL), the ventral posteromedial (VPM), the posterior nuclear group (Po) and triangular posterior nucleus (PoT) of the thalamus. The third system is centred on the projection from lamina I to the lateral parabrachial area. We also present the four main projections from the latter area to the extended amygdala, the hypothalamus, the periaqueductal grey matter (PAG), and the ventrolateral medulla (VLM), and their involvement in emotional and autonomic (homeostatic) aspects of pain. Experimental Physiology (2002) 87.2, 251-258.

Mitochondria are subcellular organelles, devoted mainly to energy production in the form of ATP, that contain their own genetic system. Mitochondrial DNA codifies a small, but essential number of polypeptides of the oxidative phosphorylation system. The mammalian mitochondrial genome is an example of extreme economy showing a compact gene organization. The coding sequences for two ribosomal RNAs (rRNAs), 22 transfer RNAs (tRNAs) and 13 polypeptides are contiguous and without introns. The tRNAs are regularly interspersed between the rRNA and protein-coding genes, playing a crucial role in RNA maturation from the polycistronic transcripts. A single major non-coding region, called the D-loop region, contains the main regulatory sequences for transcription and replication initiation. This genetic organization has its precise correspondence in the mode of expression and distinctive structural features of the RNAs. The basic mechanisms of mitochondrial DNA transcription and replication and the main cis-acting elements playing a role in both processes have been determined. Many trans-acting factors involved in mitochondrial gene expression, including the RNA and DNA polymerases, have been cloned or identified. However, the regulatory mechanisms participating in mitochondrial gene expression are still poorly understood. The interest to complete this knowledge is increased by the involvement of mitochondria in human diseases, in basic processes such as heat production, Ca2+ homeostasis and apoptosis, and by their potential role in ageing and carcinogenesis. Experimental Physiology (2003) 88.1, 41-56.

In this study species-specific values for whole blood viscosity (WBV), plasma viscosity (PV) and erythrocyte aggregation (EA) were determined in a total of 360 animals. We used 40 individual adult animals of nine mammalian species: horse, pig, dog, cat, rat, cattle, sheep, rabbit and mouse. WBV measurements were carried out using a LS30 viscometer, PV was measured using OCR-D and EA was measured using a Myrenne aggregometer and the LS30 (aggregation index at low shear rate). At low shear rates (0.7 s-1 and 2.4 s-1) haematocrit (Hct)-standardized (40 % Hct) samples showed a higher value of WBV and EA in horse, pig, dog and cat. In cattle, sheep, rabbit and mouse, EA and WBV were markedly decreased and EA was almost undetectable, although the plasma fibrinogen concentration was higher in these animals. Rats showed the highest WBV at low shear rate in native blood and WBV was not different from horse in Hct-standardized blood; however, EA was very low in the rat, a result that might be explained by mechanical or geometrical properties of the red blood cell. EA correlated with the plasma protein concentration in each species except dog and mouse. In horse, cattle and pig, EA correlated with the plasma fibrinogen concentration. At high shear rate (94 s-1), WBV was higher in cattle than cat and rat, and dog had higher values than horse, suggesting specific interspecies differences depending on low shear and high shear values of WBV, as a result of mechanisms that influence RBC flexibility. PV was highest in cattle and lowest in rabbit and mouse and did not correlate with WBV. Haemorheological parameters differed between the species. Each species has its own rheological fingerprint. The physiological significance of these variations among mammalian species has not yet been established. Viscosity contributes to endothelial cell shear stress. While haemorheological parameters differ across the species it may be postulated that factors influencing flow-mediated endothelial cell signal transduction are different among the species. Experimental Physiology (2003) 88.3, 431-440.

Nociceptive dorsal root ganglion (DRG) neurones have fibres that conduct in the C, Aδ and Aα/β conduction velocity range. The properties of nociceptive compared with non-nociceptive somatic afferent dorsal root ganglion neurones appear to fall into two patterns, A and B. Pattern A properties of nociceptive neurones, the more common type, include longer action potential duration and slower maximum rate of fibre firing, as well as a greater expression of substance P and calcitonin gene-related peptide immunoreactivity. The values of pattern A properties appear to be graded according to the conduction velocity group (C, Aδ or Aα/β) of the fibres. The most pronounced forms of A-type properties are expressed by nociceptive neurones with C-fibres, and these become less pronounced in nociceptive neurones with Aδ-fibres and least pronounced in those with Aα/β fibres (C > Aδ > Aα/β). Some of these properties are also expressed in a less extreme but similarly graded manner through C, Aδ and Aα/β groups of non-nociceptive low threshold mechanoreceptive (LTM) neurone. The less common pattern B properties of nociceptive neurones have similar values in C-, Aδ- and Aα/β-fibre nociceptive neurones but these clearly differ from LTM units with C-, Aδ- and Aα/β-fibre conduction velocities. These features of nociceptive neurones include consistently larger action potential overshoots and longer after-hyperpolarisation durations in nociceptive than in LTM neurones. Experimental Physiology (2002) 87.2, 239-244.

This review explores the current concepts relating the structural and functional modifications of skeletal muscle mitochondria to the molecular mechanisms activated when organisms are exposed to a hypoxic environment. In contrast to earlier assumptions it is now established that permanent or long-term exposure to severe environmental hypoxia decreases the mitochondrial content of muscle fibres. Oxidative muscle metabolism is shifted towards a higher reliance on carbohydrates as a fuel, and intramyocellular lipid substrate stores are reduced. Moreover, in muscle cells of mountaineers returning from the Himalayas, we find accumulations of lipofuscin, believed to be a mitochondrial degradation product. Low mitochondrial contents are also observed in high-altitude natives such as Sherpas. In these subjects high-altitude performance seems to be improved by better coupling between ATP demand and supply pathways as well as better metabolite homeostasis. The hypoxia-inducible factor 1 (HIF-1) has been identified as a master regulator for the expression of genes involved in the hypoxia response, such as genes coding for glucose transporters, glycolytic enzymes and vascular endothelial growth factor (VEGF). HIF-1 achieves this by binding to hypoxia response elements in the promoter regions of these genes, whereby the increase of HIF-1 in hypoxia is the consequence of a reduced degradation of its dominant subunit HIF-1α. A further mechanism that seems implicated in the hypoxia response of muscle mitochondria is related to the formation of reactive oxygen species (ROS) in mitochondria during oxidative phosphorylation. How exactly ROS interfere with HIF-1α as well as MAP kinase and other signalling pathways is debated. The current evidence suggests that mitochondria themselves could be important players in oxygen sensing. Experimental Physiology (2003) 88.1, 109-119.

Non-shivering thermogenesis in brown adipose tissue is the main mechanism for thermoregulatory heat production in small mammals and newborns. During cold acclimation the sympathetic innervation triggers the recruitment of brown adipose tissue by hyperplasia, which involves the proliferation and differentiation of precursor cells, and by hypertrophy of mature brown adipocytes. Mitochondrial biogenesis and increased synthesis of the uncoupling protein 1 (UCP-1) are hallmarks of the thermogenic recruitment process. The severalfold increase of mitochondrial protein content during cold acclimation recruits a large capacity for oxidative phosphorylation. However, UCP-1 increases proton leakage across the inner membrane of brown adipocyte mitochondria and thereby dissipates proton motive force as heat instead of ATP synthesis. During recent years considerable progress has been achieved in the analysis of transcriptional mechanisms controlling Ucp1 gene expression. However, so far only little is known about the molecular basis of cold-induced mitochondrial biogenesis in brown adipose tissue. Experimental Physiology (2003) 88.1, 141-148.

Embryonic stem (ES) cells are pluripotent cells isolated from the inner cell mass of blastocysts. ES cells are able to differentiate into the three primitive layers (endoderm, mesoderm and ectoderm) of the organism, including the germline. To study early stages of development, as well as to investigate the impact of a gene knock-out in vitro, ES cells are differentiated into three-dimensional structures called embryoid bodies, because of their ability to mimick post-implantation embryonic tissues. This review summarises the work on ES cell differentiation into haematopoietic and vascular cells, neuronal and glial cells, myocytes, and adipocytes, using this in vitro model of early embryogenesis. We also present the potential of this method to analyse the impact of genetic alterations in vitro

The scientifically novel, but evolutionarily ancient, so-called uncoupling proteins 2 and 3 (UCP2, UCP3) are structually similar to the archetypical uncoupling protein UCP1. A series of suggestions have been forwarded for their physiological function. We discuss systematically here the pros and cons for these suggestions. We conclude that the novel UCPs do not seem to be physiologically relevant uncoupling proteins; the uncoupling property was apparently a late introduction into the subfamily through the evolution of UCP1. Physiological functions ascribed to UCP2 and UCP3 based on their purported uncoupling property may have to be revised (i.e. any type of thermogenesis, including protection against obesity, protection against the formation of reactive oxygen species and thermogenic involvement in the fever response). The presence of a mixed genetic background in most published studies of UCP2 or UCP3 gene-ablated mice also means that data concerning marked differences in diabetes propensity, infection sensitivity and production of reactive oxygen species may require confirmation in backcrossed mice. The increased expression of UCP2 and UCP3 under conditions of increased fatty acid metabolism implies an as yet undefined role in lipid metabolism. Thus, the novel UCPs should probably be considered as mitochondrial carriers, and the challenge now is to identify the transported molecule. Experimental Physiology (2003) 88.1, 65-84.

The control of self-initiated falls from different heights was studied. The objective of the study was to investigate in a quantitative manner the modulation of EMG timing (i.e. onset from take-off and duration from onset to touch-down) and amplitude (before and after foot contact) as a function of fall height. The muscles studied were m. soleus and m. tibialis anterior. Kinematic (ankle joint angle) and kinetic (ground reaction force) variables were also measured. Six subjects took part in the experiments that consisted of ten landings from each of five heights (0·2, 0·4, 0·6, 0·8 and 1 m) onto a force platform. We found a consistent pattern of co-contraction before and after touch-down across the fall heights studied. In both muscles, the onset of pre-landing EMG activity occurred at a longer latency following take-off when landing from greater heights. The absolute EMG duration was affected to a lesser extent by increasing fall height. These findings suggest that the onset of muscle activity of the muscles studied prior to foot contact is timed relative to the expected time of foot contact. Pre- and post-landing EMG amplitude tended to increase with height. Despite a doubling in the magnitude of ground reaction force, the amplitude of ankle joint rotation caused by the impact remained constant across heights. These findings suggest that the observed pattern of co-contraction is responsible for increasing ankle joint stiffness as fall height is increased. The attainment of an appropriate level of EMG amplitude seems to be controlled by (a) timing muscle activation at a latency timed from the expected instant of foot contact and (b) varying the rate at which EMG builds up.

Thyroid hormone (T3) has a profound effect on mitochondrial biogenesis. T3-regulated gene expression is mediated by thyroid hormone receptor (TR) binding to thyroid hormone response elements (TREs). In concert with the action of various coactivators and corepressors this interaction leads to a modulation of the chromatin structure and subsequently to a modulation of gene expression of adjacent target genes. However, as numerous genes are endogenous regulated by T3 whereas a TRE appears to be absent in their regulatory elements, a TR-independent pathway of T3-mediated gene regulation is likely. In this review, we discuss the direct mechanisms of TR-dependent regulation of gene expression on the nuclear and mitochondrial genome by T3. We also summarise recent observations on an indirect mechanism of T3 action via intermediate factor(s). We discuss the regulation of nuclear respiratory factor 1 (NRF-1) and peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC-1α) by T3, suggesting NRF-1 and PGC-1α as attractive candidate factors for an intermediate factor of T3 action in vivo. Experimental Physiology (2003) 88.1, 121-128.

Thyroid hormone (T3) has a profound effect on mitochondrial biogenesis. T3-regulated gene expression is mediated by thyroid hormone receptor (TR) binding to thyroid hormone response elements (TREs). In concert with the action of various coactivators and corepressors this interaction leads to a modulation of the chromatin structure and subsequently to a modulation of gene expression of adjacent target genes. However, as numerous genes are endogenously regulated by T3, and a TRE appears to be absent in their regulatory elements, a TR-independent pathway of T3-mediated gene regulation is likely. In this review, we discuss the direct mechanisms of TR-dependent regulation of gene expression on the nuclear and mitochondrial genome by T3. We also summarise recent observations on an indirect mechanism of T3 action via intermediate factor(s). We discuss the regulation of nuclear respiratory factor 1 (NRF-1) and peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC-1α) by T3, suggesting NRF-1 and PGC-1α as attractive candidates for an intermediate factor of T3 action in vivo.

We established a mouse model of cardiac dysfunction due to myocardial infarction (MI). For this we ligated the left anterior descending coronary artery in male C57BL/6J mice and assessed healing and left ventricular (LV) remodelling at 1, 2 and 4 days and 1, 2 and 4 weeks after MI. Echocardiography was performed at 1 and 2 weeks and 1, 2, 4 and 6 months after MI. We found that neutrophil infiltration of the infarct border was noticeable at 1-2 days. Marked macrophage infiltration occurred at day 4, while lymphocyte infiltration was apparent at 7-14 days. Massive proliferation of fibroblasts and collagen accumulation began by day 7-14, and scar formation was completed by day 21. LV diastolic dimension increased markedly at 2 weeks and remained at the same level thereafter. LV shortening fraction decreased significantly at 2 weeks and then slowly decreased. In non-infarcted areas of the LV, myocyte cross-sectional area and interstitial collagen fraction increased progressively, reaching a maximum at 4 months. This study provides important qualitative and quantitative information about the natural history of cardiac remodelling after MI in mice. Experimental Physiology (2002) 87.5, 547-555.

Biogenesis of mitochondria is happening constantly due to the physiological and developmental situation of a cell. As mitochondrial biogenesis is a complex process producing about 20 % of cellular protein, the expression of the 1000 genes involved is expected to be coordinated and regulated tightly. The variety of physiological stimuli and differentiation states lead to the idea of a complex network connecting many different regulatory pathways. By analysing nuclear encoded mitochondrial genes some of the factors involved in the regulation and coordination of mitochondrial gene expression were identified. These factors include general transcription factors such as Sp1 or YY1, as well as transcription factors specific for mitochondrial genes like the nuclear respiratory factors NRF1 and 2. An important control function linked to the physiological situation of a cell is triggered by hormones such as steroid and thyroid hormones. Even cell type-specific regulatory proteins like the myogenin transcription factor family have a strong influence on some mitochondrial genes in the specific cellular background. The regulatory function of most of these proteins can be modulated and enhanced by the coactivators PGC-1α and β and PRC. Although regulatory pathways have been characterized in more detail in recent years, no regulation mechanism has been shown to work on all analysed mitochondrial genes, and the general concept of mitochondrial regulation still remains unclear. Experimental Physiology (2003) 88.1, 33-40.

The dorsal horn of the spinal cord plays an important role in transmitting information from nociceptive primary afferent neurones to the brain; however, our knowledge of its neuronal and synaptic organisation is still limited. Nociceptive afferents terminate mainly in laminae I and II and some of these contain substance P. Many projection neurones are located in lamina I and these send axons to various parts of the brain, including the caudal ventrolateral medulla (CVLM), parabrachial area, periaqueductal grey matter and thalamus. The neurokinin 1 (NK1) receptor on which substance P acts is expressed by certain neurones in the dorsal horn, including approximately 80 % of lamina I projection neurones. There is also a population of large NK1 receptor-immunoreactive neurones with cell bodies in laminae III and IV which project to the CVLM and parabrachial area. It has been shown that the lamina III/IV NK1 receptor-immunoreactive projection neurones are densely and selectively innervated by substance P-containing primary afferent neurones, and there is evidence that these afferents also target lamina I projection neurones with the receptor. Both types of neurone are innervated by descending serotoninergic axons from the medullary raphe nuclei. The lamina III/IV neurones also receive numerous synapses from axons of local inhibitory interneurones which contain GABA and neuropeptide Y, and again this input shows some specificity since post-synaptic dorsal column neurones which also have cell bodies in laminae III and IV receive few contacts from neuropeptide Y-containing axons. These observations indicate that there are specific patterns of synaptic connectivity within the spinal dorsal horn. Experimental Physiology (2002) 87.2, 245-249.

To our knowledge, the relationship between all four endogenous female sex hormones and resting cardiac autonomic function has not been studied. The aim of the current study was to examine the association between the normal endogenous levels of oestrogen (17β-oestradiol), progesterone, luteinising hormone and follicle-stimulating hormone and heart rate variability (HRV) during the menstrual cycle in young eumenorrheic women. Ten healthy, young, female subjects volunteered for this study. HRV and endogenous hormone levels were recorded at three phases of the menstrual cycle: menses (day 3.8 ± 0.5), ovulation (day 15.8 ± 0.7) and luteal (day 22.1 ± 0.4) to ensure HRV recordings at times of low (menses) and high (ovulation and luteal) hormonal influence. Heart rate recordings were obtained from supine resting subjects and analysed on a Holter analysis system. Total power (TP, 0-1.0 Hz), low frequency (LF, 0.041-0.15 Hz), high frequency (HF, 0.15-0.80 Hz) and LF/HF components of HRV were examined. Despite a significantly greater HR at ovulation and normal cyclic variations in all endogenous sex hormone levels, no measure of HRV was significantly different between menstrual cycle phases. Significant correlations between oestrogen levels and absolute measures of HRV at ovulation were identified. The results of the current study demonstrated that the normal cyclic variations in endogenous sex hormone levels during the menstrual cycle were not significantly associated with changes in cardiac autonomic control as measured by HRV. Significant correlation between peak oestrogen levels and HRV measures at ovulation provided further support for the reported cardioprotective effects of oestrogen in healthy females. Experimental Physiology (2003) 88.3, 441-446.

Phasic changes in pressure have been reported to occur in the bladder which are not associated with micturition. Spontaneous intravesical pressure changes can be recorded from bladders in vitro or bladders in vivo isolated from the central nervous system suggesting that the bladder itself is capable of autonomous activity. Experiments using isolated cells and muscle strips indicate that the smooth muscle can generate spontaneous activity. Whether this is the origin of phasic changes in the intact organ remains unknown. The present study set out to establish the presence and characteristics of autonomous activity in the isolated guinea pig bladder. Multiple-point motion analysis and concurrent intravesical pressure recording were used to identify and quantify spontaneous and evoked activity. Highly complex autonomous activity was observed in unstimulated bladders. This activity comprised localised micro-contractions in single or multiple discrete regions, waves of activity and micro-stretches. Low-amplitude phasic 'micro-transients' were seen in the intravesical pressure trace in association with micro-contractions. Incremental increases in the intravesical volume recruited additional areas of activity. Atropine and tetrodotoxin had no effect on the micro-transients or micro-contractions. Exposure to the muscarinic agonist arecaidine (10-300 nM) initially increased the incidence of micro-contractions which subsequently became co-ordinated into phasic pressure rises and contraction waves, interspersed with periods of total quiescence. The findings describe the generation and co-ordination of autonomous activity in the bladder wall and also demonstrate complex phasic activity. This approach has shown the importance of assessing the integrative properties of the entire organ in studies of the physiology and patho-physiology of the bladder. Experimental Physiology (2003) 88.1, 19-30.

Apoptosis is an evolutionary-conserved physiological mechanism to remove cells from an organism. Cellular apoptosis is mediated via an intracellular signalling programme that involves a variety of signalling molecules and cellular organelles including caspases, sphingomyelinases, Bcl-2-like proteins and proteins to cleave the DNA and mitochondria. Mitochondria contain several pro-apoptotic molecules that activate cytosolic proteins to execute apoptosis, block anti-apoptotic proteins in the cytosol and directly cleave nuclear DNA. Mitochondria trap these pro-apoptotic proteins and physically separate pro-apoptotic proteins from their cytoplasmic targets. Apoptosis is then initiated by the release of mitochondrial pro-apoptotic proteins into the cytosol. This process seems to be regulated by Bcl-2-like proteins and several ion channels, in particular the permeability transition pore (PTP) that is activated by almost all pro-apoptotic stimuli. Experimental Physiology (2003) 88.1, 85-90.

Regularly performed exercise in the form of endurance training produces a well-established adaptation in skeletal muscle termed mitochondrial biogenesis. The physiological benefit of this is an enhanced performance of muscle when subject to endurance exercise. This is not only of great advantage for athletic endeavours, but it also clearly improves the quality of life of previously sedentary individuals and those involved in injury rehabilitation. Here we review the molecular basis for mitochondrial biogenesis in muscle, from the initial signals arising in contracting muscle, to the transcription factors involved in mitochondrial and nuclear DNA transcription, as well as the post-translational import mechanisms required for the synthesis of the organelle. We discuss specific protein components associated with reactive oxygen species production, and suggest some questions which remain unanswered with respect to the role of exercise-induced mitochondrial biogenesis in ageing, apoptosis and disease. Experimental Physiology (2003) 88.1, 99-107.

The origin of significant differences between the apparent affinities of heart mitochondrial respiration for exogenous ADP in isolated mitochondria in vitro and in permeabilized cardiomyocytes or skinned fibres in situ is critically analysed. All experimental data demonstrate the importance of structural factors of intracellular arrangement of mitochondria into functional complexes with myofibrils and sarcoplasmic reticulum in oxidative muscle cells and the control of outer mitochondrial membrane permeability. It has been shown that the high apparent Km for exogenous ADP (250-350 µM) in permeabilized cells and in ghost cells (without myosin) and fibres (diameter 15-20 µm) is independent of intrinsic MgATPase activity. However, the Km may be decreased significantly by a selective proteolytic treatment, which also destroys the regular arrangement of mitochondria between sarcomeres and increases the accessibility of endogenous ADP to the exogenous pyruvate kinase-phosphoenolpyruvate system. The confocal microscopy was used to study the changes in intracellular distribution of mitochondria and localization of cytoskeletal proteins, such as desmin, tubulin and plectin in permeabilized cardiac cells during short proteolytic treatment. The results show the rapid collapse of microtubular and plectin networks but not of desmin localization under these conditions. These results point to the participation of cytoskeletal proteins in the intracellular organization and control of mitochondrial function in the cells in vivo, where mitochondria are incorporated into functional complexes with sarcomeres and sarcoplasmic reticulum.Experimental Physiology (2003) 88.1, 175-190.

A novel isolated Langendorff perfused rabbit heart preparation with intact dual autonomic innervation is described. This preparation allows the study of the effects of direct sympathetic and vagus nerve stimulation on the physiology of the whole heart. These hearts (n = 10) had baseline heart rates of 146 ± 2 beats min-1 which could be increased to 240 ±11 beats min-1 by sympathetic stimulation (15 Hz) and decreased to 74 ± 11 beats min-1 by stimulation of the vagus nerve (right vagus, 7 Hz). This model has the advantage of isolated preparations, with the absence of influence from circulating hormones and haemodynamic reflexes, and also that of in vivo preparations where direct nerve stimulation is possible without the need to use pharmacological agents. Data are presented characterising the preparation with respect to the effects of autonomic nerve stimulation on intrinsic heart rate and atrioventricular conduction at different stimulation frequencies. We show that stimulation of the right and left vagus nerve have differential effects on heart rate and atrioventricular conduction. Experimental Physiology (2001) 86.3, 319-329.

The purpose of the present study was to investigate the effect of temperature on the rates of isometric force development and relaxation in electrically activated fresh and fatigued human adductor pollicis muscle. Following immersion of the lower arm for 20 min in water baths of four different temperatures, muscle temperatures were approximately 37, 31, 25 and 22°C. Maximal isometric force was reduced by 16·8 ± 1·5 % at 22°C. The stimulation frequency-force and -rate of force development relationships were shifted to the left at lower temperatures. Q10 values for the maximal rates of force development and relaxation, and the times for 100 to 50 % and 50 to 25 % force relaxation, were about 2·0 between 37 and 25°C and about 3·8 between 25 and 22°C. However, the time for 50 to 25 % force relaxation had a relatively high Q10 value between 25 and 22°C (6·9) and this parameter also appeared to be more sensitive to fatigue compared to the other indices of relaxation. Nevertheless, the effect of fatigue on all parameters decreased with cooling over the entire (37-22°C) temperature range.