The International Hypoxia Symposium 2015 in Lake Louise: A Report

In early March 2015, the 19^th edition of the bi-annual International Hypoxia Symposia took place in the gorgeous winter settings of the Château Lake Louise in the Canadian Rockies. Some 190 scientists, doctors, students, and other interested people gathered to learn and exchange on the multiple facets of life's fascinating relationship with the element oxygen. It was the late Charlie Houston who launched the International Hypoxia Symposia several decades ago, to the great satisfaction of an increasingly international audience. The symposia still continue, now organized by Rob Roach, with help from Peter Hackett and myself. Even if visa problems unfortunately prevented some prospective attendees to make it to Lake Louise, the 2015 symposium again brought together people from all around the world. The (informal) competition on who travelled the most miles to attend was close, with several attendees from places on the opposite side of the planet. The prize went to Keith Burgess who came all the way from Australia. As customary the symposium started on Tuesday at 6 PM with a welcome reception, the perfect occasion to meet many friends and colleagues again.

On Wednesday the first session Hypoxia: From Microcirculation to Genetic Selection started off with Can Ince from Amsterdam, who reported some of his fascinating work on hypoxia in the intensive care settings. He introduced his talk with the paradoxical observation that patients with systemic circulatory problems, such as encountered in sepsis or major traumatic hemorrhage, once stabilized with regard their systematic hemodynamics, still show serious signs of tissue hypoxia. Ince underlined the important point that one cannot draw any conclusions on the state of the microcirculation looking only at systematic hemodynamics (Kanoore et al., 2015). Initially using NADH fluorescence and later porphyrin quenching techniques, Ince and colleagues have convincingly shown that hypoxic pockets in otherwise healthy tissue can exist because of changes on microcirculatory level. Ince and colleagues now have developed handheld microscopes to capture video of the microcirculation (e.g., under the tongue) to monitor the microcirculatory state in patients. Ince also reminded us that increasing Fio₂ in patients does not necessarily lead to improved oxygenation there were needed, and that an increase in Po₂ may even lead to microcirculatory vasoconstriction. Ince ended his presentation citing from his article in a Christmas issue of the British Medical Journal titled ‘Why Rudolph's Nose is Red: Observational Study' (Ince et al., 2012), in which he and Norwegian colleagues used handheld microscopes to image the microcirculation in reindeer in comparison to men, showing that reindeer have a higher microcirculatory density.

Andy Lovering followed with the troubling observation that more than a third of the general population has a patent foramen ovale (PFO). Even if Claudius Galen already reported the existence of PFOs, it was Hagen who first quantified prevalence between 19% and 36%, based on cadaver observations (Hagen et al., 1984). The modern technique of saline contrast echocardiography has now led to estimations of PFO prevalence between 25% and 38%.

This is troubling because such prevalence implies that much of what has been published to date on cardio-pulmonary gas exchange has not taken into account any effects of shunting through such a PFO. For example, PFO⁺ subjects show widening of their alveolar-arterial O₂ difference (AadO₂) with exercise intensity. The size of the PFO matters, because the larger the PFO the greater the shunt. Thus, depending on size, PFOs have relevant effects, as also suggested by recent observations during the AltitudeOmics project (Elliott et al., 2015). In PFO^- subjects, acclimatization over 16 days improved the initially widened Alveolar-arterial O₂ difference (AadO₂), while in PFO⁺ subjects the AadO₂ did not improve. PFO⁺ subjects further showed blunted ventilatory acclimatization as compared to PFO⁻ subjects for reasons that remain unknown. Lovering further presented some intriguing results on the effects of a PFO on thermoregulation. The idea is that the pulmonary circulation is important for cooling and that a right-left shunt through a PFO may impair this, leading to a measurable increase in body temperature in PFO⁺ subjects. Lovering concluded by suggesting that the presence of a PFO confers an increased risk for AMS and HAPE.

The first session was concluded by a fascinating talk given by Emilia Huerta on the presence of genes in the Tibetan genome coming from admixture with the Denisovans' genome (Huerta-Sánchez et al., 2014). She started her talk by reminding us that there are three regions on earth in which humans have been exposed to lowered Pio₂ for enough generations to have undergone selection pressure for adaptation: East-Africa, the Andes, and Tibet, even though the adaptations described for the populations living in these regions are not necessarily the same. The Tibetans do not show the typical increased hemoglobin levels seen in lowlanders when exposed to high altitude, and this is explained by a difference in the EPAS1 gene.

Apart from this gene, several other distinct genetic differences between Tibetans and their closest lowland cousins the Han Chinese have been described. Given the particular haplotype of the Tibetans in comparison to the Han Chinese, it appeared plausible that the Tibetans would have received specific alleles from another population, prompting research into this question. Huerta first jokingly mentioned the Yeti, but then presented rather compelling data suggesting that there may have been admixture with another now extinct hominid, the Denisovan, so-called after the name of the cave in the Altai Mountains in which some bones of this hominid were found, with sufficient DNA for extraction and sequencing (Huerta-Sánchez et al., 2014).

The next session Advances in Ischemic and Hypoxic Preconditioning opened with a presentation by Robert Meller. He reminded us of the well-known observation that a (short) ischemic preconditioning episode in an organ like the heart or the brain is protective for a subsequent longer and normally pathological episode. The biology of preconditioning is complex and only partly unraveled (Meller and Simon, 2013). It may, for the brain, involve downregulation of the KLF superfamily of genes, leading to a less differentiated phenotype of neurons more resilient to ischemia. Meller then presented the effects of so-called remote ischemic preconditioning (RIPC) in which an ischemic episode in, for example, a limb protects another organ like the brain or the heart from the effects of ischemia. Such effects could theoretically stem from a humoral and/or a neuronal effect. Animal experiments clearly showed a brain hemispheric bias upon single hindlimb RIPC, indicating an important neuronal effect. Interestingly enough, the protective effect was related to an increase in NREM sleep following RIPC and when this increase in NREM sleep was prevented the protective effect disappeared.

Marc Berger followed with his talk Preconditioning for Prevention of High Altitude Illness: Fact or Fiction. He reported the results of a study in which unacclimatized subjects were exposed to 18 h of normobaric hypoxia equivalent to an altitude of 4500 m with or without RIPC composed of four cycles of 5 minutes lower limb ischemia just prior to hypoxic exposure (Berger et al., 2015). RICP led to lower acute mountain sickness (AMS) scores at 5 h but not at 18 h. Berger tentatively explained this as resulting from the typical bi-phasic response of RIPC with a window of no effect between the initial protection shortly after RIPC and a delayed protection generally observed after 12–72 h. Lower ROS levels after RIPC accompanied the reduction in AMS scores. While this observation supported the concept that ROS may play a role in the pathophysiology of AMS, the findings after 18 h spoke against this hypothesis because ROS plasma levels remained decreased in both RIPC and control groups, although AMS scores had increased by then. This dissociation between ROS and severity of AMS further questions the role of increased oxidative stress in the pathophysiology of AMS.

Anthony Bain was selected from the submitted abstracts to complete this session with his talk Cerebral Metabolic Rate During Prolonged Apnea in Humans. Bain presented the results of observations in elite Croatian divers in whom the arteriovenous differences for oxygen and Doppler-estimated cerebral blood flow were used to calculate cerebral metabolic rate of oxygen (CMRO₂) (Willie et al., 2014). They observed that CMRO₂ progressively dropped towards the breaking point. Bain and colleagues then used end-tidal forcing to mimic the hypoxia and hypercapnia observed at the breaking point and again found a drop in CMRO₂. Bain suggested that the hypercapnic acidosis would perhaps slow aerobic glycolysis protecting the brain against hypoxia.

David Dubowitz ended the session with his selected abstract on The Impact of Hypoxic Acclimatization on CMRO₂ Responses to CO₂ and Visual Stimulation, reporting the effect of different lengths of exposure to hypoxia on CMRO₂ of the visual cortex in response to CO₂ and visual stimuli using calibrated BOLD MRI techniques. Similarly to Bain, Dubowitz and colleagues observed a drop in CMRO₂ with hypercapnia in hypoxia.

Thursday opened with a session Intermittent Hypoxemia: Friend or Foe. The session was introduced by Jerry Dempsey, who reminded us of the dangerous consequences of obstructive sleep apnea (OSA) (Dempsey et al., 2010). Because of the repeated nocturnal episodes of hypoxia, the carotid bodies undergo structural changes leading to increased responsiveness to hypoxia and an exaggerated sympathetic arousal.

Steve Mifflin then presented results of an animal model of intermittent hypoxia (IH) mimicking OSA (Yamamoto et al., 2015). In this rat model, IH leads to persistent arterial hypertension, also during waking hours when IH is discontinued. Mifflin and colleagues found structural changes in the brain stem and observed an increased responsiveness to N₂ of NTS neurons. The paraventricular nucleus was also implicated through changes in glutaminergic drive.

David Fuller followed with his presentation Intermittent Hypoxia as a Neurorehabilitation Tool. He started his talk by answering the session's question 'Friend or Foe?' by the exclamation 'Frenemy!'. The point he made was that even though chronic IH such as encountered in OSA is pathogenic, some forms of acute IH may in fact have salutary effects (Lee et al., 2015). In a rat model of spinal cord injury, Fuller and colleagues showed that IH could improve connectivity through an embryonic graft put into the cavity left after spinal cord concussion in the adult rat. Fuller foresees that IH will become a useful additional tool in neurorehabilitation, provided the dose is optimized.

The next session titled Iron and Hypoxia: The Good and the Bad was opened by Martina Muckenthaler. She did a wonderful job in presenting the important physiological roles of iron. Not only is iron involved in oxygen carrying (hemoglobin, myoglobin, cytochromes), it is also involved in oxygen sensing and oxygen-dependent gene expression. Given its particular properties, too much iron is bad, as is too little. The uptake of iron is therefore tightly regulated by a complex system that is still not totally elucidated (Steinbicker and Muckenthaler, 2013). Hypoxia has important effects on iron status, not the least because of the increased erythropoiesis after the upregulation of EPO. Iron requirements rise and hepcidin levels decrease to allow for elevated iron absorption and release from stores. The transcription factors HIF-1 and HIF-2 are stabilized under hypoxic conditions and regulate transcription of several iron related genes (TfR1, Tf, ceruloplasmin, DMT-1, FPN1). On the last evening, Martina Muckenthaler was deservedly awarded the Jack Reeves prize for the best oral presentation.

Peter Robbins followed with a presentation linking iron and oxygen to the pulmonary circulation. Robbins started out by presenting his lab's work on the hypoxic pulmonary vasoconstrictive response (HPV). After having documented that the HPV is bi-phasic and takes several hours to attain a plateau, Robbins and colleagues use 8 h of normobaric hypoxia, followed by 3 h of normoxia, and again 1 h of acute hypoxia for phenotyping HPV. Using this model they showed that Tibetans living at low altitude have a blunted HPV compared to lowland Han Chinese (Frise and Robbins, 2014). Interestingly, in lowlanders an intravenous injection of iron blunts HPV and leads to a decreased AMS score, an effect lasting at least 6 weeks. Robbins and colleagues are now looking into the question if iron status is of importance in illnesses such as COPD in which hypoxemia plays an important role.

From the same research group Matthew Frise's abstract was selected for an oral presentation and he completed Robbins talk presenting data suggesting that iron deficiency may lead to an exaggerated HPV and that intravenous iron mitigates the second phase of HPV. Frise ended by asking the question whether iron status potentially plays a role in the susceptibility to HAPE, promising an answer for the next symposium. The second selected abstract was also from the same group by David Holdsworth. Using MR spectroscopy they looked into the role of iron status in high energy phosphate homeostasis during exercise, inspired by the finding that Chuvash patients develop an exaggerated decrease in PCr and increased lactate concentrations during exercise.

The first Hot Topics in Mountain Medicine session, based on selected abstracts, started out with Daniel Martin presenting results from sublingual microcirculation video in lowlanders and Sherpas during the Xtreme Everest 2 expedition. In lowlanders, altitude exposure increased vessel density and blood flow index, but did not reach the levels observed in the Sherpas. Martin and colleagues hypothesized that the increased vessel density and flow index may form part of what distinguishes Sherpas from lowlanders when exposed to high altitude (Gilbert-Kawai et al., 2014).

Linda Keyes followed with interesting data on arterial blood pressure (ABP) in hypertensive and nonhypertensive trekkers. Keyes and colleagues recruited 670 trekkers and measured pressure at several altitudes on the way from Lukla (2800 m) up the Khumbu valley (Pheriche or Dingboche, 4400 m). On group level there was no effect of altitude; some had increased and some decreased ABP. However, some trekkers with pre-existing hypertension developed severe hypertension despite continued medication, suggesting that the management of hypertension in trekkers may need to be better investigated.

Fernando Moraga then presented data from Chile. There are 50,000 miners and 15,000 support personnel regularly exposed to high altitude and the numbers are expected to increase by more than half between now and 2020. All these subjects are exposed to CIH, traveling up for work and down again for off-time. Mining companies are looking into ways to decrease the health risks of this type of CIH and to improve working performance at altitude. Office workers at 5050 m were exposed to a working environment with an increased Fio₂ equivalent to 2800 m. This improved Sao₂ and, importantly, cognitive performance. Miners were given supplementary oxygen during sleep which improved perceived sleep quality and decreased periodic breathing (Moraga et al., 2014).

Keith Burgess then presented the results of an elegant study following up on earlier work (Burgess et al., 2014). He and colleagues looked, in acclimatized lowlanders, at the effect of increasing cerebral blood flow without changing arterial O₂ and CO₂ tensions (by infusing dobutamine together with acetazolamide) and found that this reduced the occurrence and severity of central sleep apneas, possibly because of a reduced hypercapnic vascular response.

Tana Uran closed the session presenting novel data on Tibetan placentas. The well-known effect of high altitude lowering birth weight is less prominent in Tibetans, and Uran set out to look for differences at placenta level potentially explaining this finding. A total of 1382 genes were found to be differently expressed comparing Tibetan with Han Chinese placentas, with more genes downregulated than upregulated. HIF1 was downregulated in Tibetans compared to the Han, whereas there was no difference for HIF2. HIF3 was upregulated in Tibetans. ER stress was less in Tibetans. Overall it seemed that the Tibetan placenta is better protected from the effects of hypoxia, also during labor, but the exact mechanisms remain to be elucidated. Tana Uran deservedly won the 2015 prize for the best trainee oral presentation.

The first Hot Topics in Hypoxia session started out with Michael Furian presenting results obtained in COPD patients during exercise at moderate altitude (2590 m). Exercise at altitude was reduced by 60% compared to that at 490 m and accompanied by hyperventilation and cerebral desaturation. Jonas Saugy then presented new data that add to the evidence base suggesting that hypobaric and normobaric hypoxia do not have the exact same effects (Millet et al., 2013). Barbara Morgan followed with results obtained in a rat model subjected to 21 days of CIH and found that this led to an increased chemoreflex control in conscious rats. Shailendra Sharma reported interesting observations of cobalt poisoning in miners in Cerro de Pasco (4200 m) suffering from CMS. Treatment with n-acetylcysteine (to chelate Co) or acetazolamide (to improve Pao₂) or their combination all improved CMS score, but only acetazolamide improved hematocrit. The last talk in this session was given by Heimo Mairbäurl presenting results from experiments in myocytes exposed to hypoxia. Mairbäurl and colleagues tried to unravel the mechanisms of the hypoxia-induced decrease in oxidative metabolism and suggest this is due to decreased expression of electron transfer chain enzymes through blocking signaling downstream of AMPK.

Friday's first session on Science of Mountain Medicine: Cerebral Hemodynamics was opened by Mark Wilson who presented his group's work on the cerebral venous circulation at high altitude. Wilson first pointed out the analogy between the Space Adaptation Syndrome and the effects of altitude on the cerebral circulation, since they both lead to venous distension (Wilson et al., 2011). Wilson and colleagues suggest that venous distension plays an important role in high altitude headache (Wilson et al., 2013). Gabriel Willmann followed with a talk on the posterior part of the eye as a window on the brain in conditions of hypoxia (Willmann et al., 2014). Observations made in the Capanna Regina Margherita (4554 m) in the Alps, where 50% of subjects had AMS upon waking after their first night up there, showed no correlation between optic disk swelling and AMS. Willmann then made the point that papilledema is not the same as optic disk swelling. Only the former is related to increased intracranial pressure (ICP), the latter not necessarily. Willmann and colleagues found a relationship between nerve sheet diameter increase and high altitude headache. Interesting was the observation that 50% of subjects developed peripheral capillary leakage while at altitude, but again this was not related to AMS scores.

Justin Lawley then presented data on spinal fluid dynamics at altitude (Lawley et al., 2014). Lawley first showed that ICP is not static but dynamic. ICP is pulsatile and further shows wave-like fluctuations below heart rate. Lawley finished by presenting some interesting preliminary data of the effects of hypoxia on ICP made on subjects with an Ommaya reservoir, which connects a small pouch under the scalp with a lateral ventricle for the injection of drugs.

The following session Latest Advances in the Science of Mountain Medicine was opened by Konrad Bloch presenting new data on the effects of altitude exposure (1860 and 2590 m) in patients with OSA or COPD (for the COPD data see the presentation by Michael Furian mentioned above). OSA patients, when exposed to altitude, developed fewer obstructive sleep apneas, but more central sleep apneas, which could be slightly inhibited with acetazolamide. The combination of CPAP and acetazolamide was best for controlling sleep apneas in this patient population at altitude (Ulrich et al., 2014).

Erik Swenson, a well-known expert on acetazolamide (Swenson, 2013), then presented a overview of the present state and prospects of the pharmacology of high altitude illness. Swenson mentioned the burgeoning understanding of the effects of acetazolamide on aquaporins and that of its cousin methazolamide on NrF-2 mediated protection from ROS, both mechanisms potentially playing a role for the prevention and treatment of AMS. Swenson ended by citing a surprising study from Korea that reported a prophylactic effect of EPO for AMS (Heo et al., 2014).

Francisco Villafuerte then gave an update on what is known about CMS in Andean highlanders. CMS, considered present when hemoglobin concentration is greater than 21 g/dL, is generally managed by bloodletting. There is a genetic basis for the risk of developing CMS conferred by the genes SENP1 and ANP32 on chromosome 12. These genes are respectively involved in HIF and GATA1 stabilization, leading to increased EPO and EPO receptor expression. Of note is that the ratio between circulating EPO and the soluble EPO receptor is a predictor of the development of excessive erythrocytosis (Villafuerte et al., 2014).

On Saturday, at the opening session Counteracting Hypoxia: Gasotransmitters, Adenosine and Vasodilation, Frank Dinenno presented data obtained with his forearm model. Dinenno uses this model to investigate the mechanisms of hypoxia-induced vasodilation. Nitric oxide (NO) and prostaglandins are involved. Using selective blocking agents, Dinenno and colleagues found that neither NO nor prostaglandins are obligatory to observe the vasodilatory response to hypoxia. NO regulates vascular tone during hypoxia independent of the prostaglandin pathway, while prostaglandins only regulate vascular tone during hypoxia when production of NO is inhibited. The complexity of this interaction was further illustrated by the finding that only the combined inhibition of NO and prostaglandins abolished the local hypoxic vasodilatation (Markwald et al., 2011).

Yang Xia then first presented a historical overview of the discovery and understanding of sickle cell disease. It is an autosomal recessive disease of hemoglobin leading to polymerization of deoxygenated hemoglobin. Because it confers a survival advantage for malaria the disease is quite prevalent in people from African descent. Since sickling is exacerbated by hypoxia, Xia's lab researches the mechanisms behind this effect of hypoxia (Sun and Xia, 2013). Preliminary data obtained on the AltitudeOmics project suggest that an increased activity of soluble CD73 drives the release of adenosine, the latter activating the erythrocyte A2B receptor in turn leading to 2,3 DPG production, which shifts the oxygen dissociation curve to the right. This normally improves oxygen unloading in the tissues and confers an advantage, but in sickle cell disease the exacerbated desaturation increases the risk for sickling.

A second Hot Topics in Hypoxia session followed with Chris Willie presenting results from experiments at sea-level and the Pyramid in Nepal (5050 m) on the effects of hypoxia and hypercapnia on brain metabolism as estimated with cross-brain arteriovenous differences for oxygen and glucose. Overall, the results showed remarkable autoregulation and similar fuel utilization between conditions (Willie et al., 2015). Sara Hartmann then presented results of research on the effects of aging and vitamin C supplementation on the cerebrovascular and ventilatory responses to acute hypoxia. Interestingly, vitamin C paradoxically had pro-oxidant properties, while peripheral chemosensitivity was not altered, suggesting that acute oxidative stress does not alter the ventilatory response.

The following session Oxygen Sensing: From Bench to Bedside was opened by Kurt Stenmark presenting work on pulmonary vascular remodeling in hypoxic forms of pulmonary hypertension (Pugliese et al., 2015). Pulmonary artery remodeling is presumably driven by fibroblasts expressing HIF1 and 2 and developing an anaerobic glycolytic phenotype, as observed in cancer. These fibroblasts recruit, retain, and activate macrophages inducing inflammation in the adventitia, which subsequently leads to remodeling. Gregg Semenza followed with a nice overview how CIH in OSA causes hypertension and how this involves HIF in the carotid body (Semenza and Prabhakar, 2012). Holger Eltzchig closed the session with an overview of his group's work on the relationship between hypoxia and inflammation, especially with regard to the lung in clinical conditions such as ARDS (Eltzschig and Carmeliet, 2011).

The last session on Hot Topics in Hypoxia started with Sabine Lague presenting data on the hypoxic ventilatory response in the bar-headed geese (Scott et al., 2015) and the Andean geese. Interestingly, these two geese employ very different methods for maintaining oxygen supply in hypoxia. Lan Zhao then presented an animal model obtained through crossbreeding that confers protection to HPV and found that this is probably linked to the ZIP12 gene coding for a Zinc transporter. Zhao's group now focuses on identifying further pharmacological targets using this animal model.

Mike Grocott then presented data on the effects of altitude exposure on mitochondrial function comparing Sherpas to lowlanders. Using respirometry on skinned muscle fibers, altitude exposure leads to greater oxygen efficiency and lower fatty acid oxidation. The lowlanders' mitochondria, with acclimatization, became similar to those of the Sherpas. Michael Tymko presented results of experiments investigating the error made when using end-tidal CO₂ tensions instead of arterial ones when quantifying cerebral vasoreactivity to CO₂ (Tymko et al., 2015). Melissa Bouwsema presented data on the effect of sex on pulmonary capillary blood volume and its role as a limiting factor for gas exchange. The results showed that only size was of importance, there was no independent effect of sex. Andrew Beaudin closed the session presenting results of experiments in a human model of IH looking into the effect of indomethacin or celecoxib on the hypoxic and hypercapnic ventilatory response after 6 hours of IH. Celecoxib augmented the hypercapnic ventilatory response.

Apart from the great science, there again was sufficient time for play and informal contact between the attendees. While some preferred the ski slopes to talk about science while sitting on the chair lifts, others participated in the traditional US–Canada hockey tournament organized by Jerry Dempsey. On Saturday evening the attendees gathered in the Brewster Barn to conclude this high quality edition of the International Hypoxia Symposium over a typical trapper's meal before heading back home again the next day.

A new feature of the International Hypoxia Symposia, in collaboration with the Journal of Applied Physiology and its Chief Editor Peter Wagner, is that the speakers are invited to submit Mini-Reviews on the topic or their presentation. The majority of the speakers have accepted the invitation and after the usual peer-review process their articles will be published so that the readers of this report can look forward reading those Mini-Reviews in a forthcoming issue of the Journal of Applied Physiology. Finally, don't forget to mark your agendas: The next edition of the International Hypoxia Symposia will take place at the Chateau Lake Louise 7–12 February 2017. It will be the 20^th anniversary of the International Hypoxia Symposia and certainly the best edition ever.