Abstract
Chronic mountain sickness (CMS) results from chronic hypoxia. It is unclear why certain highlanders develop CMS. We hypothesized that modest increases in fetal hemoglobin (HbF) are associated with lower CMS severity. In this cross-sectional study, we found that HbF levels were normal (median = 0.4%) in all 153 adult Andean natives in Cerro de Pasco, Peru. Compared with healthy adults, the borderline elevated hemoglobin group frequently had symptoms (headaches, tinnitus, cyanosis, dilatation of veins) of CMS. Although the mean hemoglobin level differed between the healthy (17.1 g/dL) and CMS (22.3 g/dL) groups, mean plasma erythropoietin (EPO) levels were similar (healthy, 17.7 mIU/mL; CMS, 12.02 mIU/mL). Sanger sequencing determined that single-nucleotide polymorphisms in endothelial PAS domain 1 (EPAS1) and egl nine homolog 1 (EGLN1), associated with lower hemoglobin in Tibetans, were not identified in Andeans. Sanger sequencing of sentrin-specific protease 1 (SENP1) and acidic nuclear phosphoprotein 32 family, member D (ANP32D), in healthy and CMS individuals revealed that non-G/G genotypes were associated with higher CMS scores. No JAK2 V617F mutation was detected in CMS individuals. Thus, HbF and other classic erythropoietic parameters did not differ between healthy and CMS individuals. However, the non-G/G genotypes of SENP1 appeared to differentiate individuals with CMS from healthy Andean highlanders.
Chronic mountain sickness (CMS) was first described by Carlos Monge in 1925 in Cerro de Pasco, Peru at 4,338 m, and is a manifestation of chronic hypoxia characterized by excessive erythrocytosis (EE). Symptoms of EE include headache, dizziness, dyspnea, sleep disturbance, tinnitus, fatigue, alterations of memory, loss of appetite, and bone and muscle pain. Additionally, individuals with EE may have physical signs of cyanosis, venous dilatation of extremities, and clubbing of the fingers and toes. The severity of CMS is assessed using the Qinghai CMS score, which is based on the grading of symptom severity [1].
The prevalence of CMS increases with age and varies among the mountainous regions of the world. CMS is an adult disease, and its prevalence is the lowest in Ethiopians (<1%) and Tibetans (~ 1%), higher in South Asian Indians (6.2%) and Han Chineserelocated to highaltitudes (6%), andthe highest among South American Andeans (15%) [2–8]. Althougha familial component has been suggested for elevated hemoglobin concentration values in high-altitude populations, no differences in classic erythropoiesis-related genes, such as EPO and erythropoietin receptor (EPOR), have been reported between healthy individuals and thosewith CMS at the same altitude and within the same population [9].
Single-nucleotide polymorphisms (SNPs) in EPAS1 (also known as HIF-2α) and EGLN1 (prolyl hydroxylase domain 2, PHD2) have been reported to be associated with lower hemoglobin levels in Tibetans [10–14]. These SNPs were first discovered through whole-genome scans and confirmed by targeted or whole-exome sequencing [10– 13]. Among Andean high-altitude dwellers, detailed studies of these two genes have also been reported. Studies at high altitude have indicated that both Andeans and Tibetans exhibited positive selection in EGLN1, although with different signature patterns, but only the Tibetans have the protective EPAS1 variant [11,15]. However, we do not yet know whether Andeans with CMS have different SNPs than Andeans without CMS or whether healthy highlanders exhibit a protective variant(s).
One such protective trait could be fetal hemoglobin (HbF). A relationship between HbF levels and hypoxia has long been recognized, with increased HbF levels noted in a variety of clinical settings, including intrauterine hypoxia [16], maternal smoking [17], anemia of prematurity [18], birth at high altitude [19], and postnatal hypoxemia from congenital heart disease [20]. Most notably, adult alpacas living at high altitude in Peru were found to have HbF levels of 55% [21], and young baboons had increases in HbF after hypobaric hypoxia [22]. Although the magnitude of the HbF response appears to be genetically determined—for example, baboons have a more robust HbF response than rhesus macaques [23]dand some patients with sickle cell disease (SCD) respond better to hydroxyurea, HbF levels could be maintained long-term by continued erythropoietic stress such as repeated phlebotomy-induced anemia [24].
It has been proposed that hypoxia and HbF are linked through the hypoxia-inducible factor (HIF) system, where HIF-α is an essential component of the oxygen-sensing mechanism. Under normoxic conditions, HIF-α isoforms are hydroxylated by prolyl hydroxylases (PHDs) and subsequently degraded by the ubiquitin and proteosomes. Under hypoxic conditions, HIF-α isoforms are not hydroxylated; they heterodimerize with HIF, bind to hypoxic response element motifs, and induce transcription of hypoxiaresponsive genes to ameliorate the effects of hypoxia [25]. HIF can also be pharmacologically stabilized using competitive PHD inhibitors. We previously tested a potent oral PHD inhibitor (FG-2216) in rhesus macaques to stabilize HIF at sea level. FG-2216 induced EPO production and significant erythrocytosis and prevented anemia induced by weekly phlebotomy. Modest increases in red cells and reticulocytes containing HbF were observed by flow cytometry [26]. These data suggest that hypoxia, or its simulation pharmacologically, induces HbF, and lack of such a response could explain CMS. Increased HbF levels in high-altitude dwellers would validate this mechanism as a potential target of exploration for the treatment of hemoglobinopathies.
We then reasoned that modest increases in the percentage of red cells containing HbF might be sufficient to prevent some highlanders from developing CMS. As the primary aim of this study, we proposed to test whether there are modest increases in HbF levels in healthy highlanders, thus distinguishing them from highlanders with CMS. The secondary aims were to determine levels of plasma factors relevant in high-altitude natives, such as EPO and pro-BNP, and perform genotyping by Sanger sequencing of genes in the HIF and related pathways and the presence of JAK2 V617F mutation, in healthy highlanders and those with CMS to investigate other causes of erythrocytosis.
Methods
Participants
One hundred fifty-three native adult high-altitude dwellers residing near the study area in Cerro de Pasco, Peru (4,338 m) were evaluated. Exclusion criteria included red blood cell (RBC) transfusion, erythropoietin or similar medications in the last 6 months, evident malnutrition, smoking (>5 cigarettes per day for >1 month), history of chronic diseases (asthma, chronic obstructive pulmonary disease, cardiovascular, and/or renal diseases), history of phlebotomy within the last year (>500 mL every 8 weeks), ongoing pregnancy, and visits to lower altitudes (<3,000 m) with total absences longer than 1 year or within the last 6 months. The protocol and consent were reviewed and approved by the institutional ethics/review boards of UPCH in Lima, Peru, and the National Institute of Diabetes, Digestive, and Kidney Diseases in Bethesda, Maryland (clinicaltrials.gov, NCT00695123). All individuals signed informed consents before enrollment.
Clinical evaluation CMS assessment and definition of subgroups
Blood pressure (BP) was measured using a conventional sphygmomanometer after a 15-min rest in the supine position. Arterial oxygen saturation (SaO2) was measured by digital pulse oximetry (Nellcor OxiMax N-560).
The presence and severity of CMS were assessed using the Qinghai CMS score (Supplementary Table E1, online only, available at www.exphem.org). The severity of each symptom was assessed according to the clinical criteria (Supplementary Table E2, online only, available at www.exphem.org). Symptoms were graded on a scale from 0 (absence) to 3 (severe). A hemoglobin value ≥21 g/dL (hematocrit ≥63%) in men and ≥19 g/dL (hematocrit ≥57%) in women was assigned 3 points, and lower hemoglobin values were assigned 0 points. CMS was calculated as the sum of the points given for symptoms and hemoglobin: 0–5 = no CMS; 6–10 = mild CMS; 11–14 = moderate CMS; ≥15 = severe CMS. Also, healthy adults were further divided into two groups to detect differences in clinical features and a possible progression to CMS: healthy adults (HAs) had hemoglobin levels ≤18 g/dL (men) and ≤16 g/dL (women); healthy borderline adults (BDLs) had hemoglobin levels of 18.1–20.9 g/ dL (men) and 16.1–18.9 g/dL (women).
Blood sampling and testing
Hematocrit was obtained by puncturing the side of a finger with a lancet, collecting the blood into heparinized capillary tubes, centrifuging at 11,500 rpm for 10 min, and reading using a microhematocrit reader. Peripheral blood was collected into EDTA-containing tubes. Hemoglobin electrophoresis was performed with a Bio-Rad high-performance liquid chromatograph for HbF. Plasma samples were used for EPO and pro-N-terminus BNP determinations with an IMMULITE 2000 XPi analyzer.
Sanger sequencing
DNA was extracted using the MagNA Pure automated instrument (Roche) according to the manufacturer’s protocol. Samples with low DNA yield were amplified using the GenomiPhi V2 DNA Amplification Kit for whole-genome amplification. SNP identification was carried out by polymerase chain reaction (PCR) of the four candidate gene regions (EPAS1, EGLN1, ANP32D, and SENP1; PCR primer sequences are listed in Supplementary Table E3, online only, available at www.exphem.org). The sequence was aligned with program Sequencher (Gene Codes, Ann Arbor, MI), and SNP was picked. Probability values were calculated with a published Excel program [27] and an online calculator (http://ihg.gsf.de/cgi-bin/hw/hwa1.pl). JAK2 V617F mutation analysis was performed using the Ipsogen JAK2 Muta-QuantTM Kit (Qiagen, Gaithersburg, MD) for quantification of the JAK2 V617F/G1849T allele in genomic DNA. The assays were run per kit protocol. Absence of the JAK2 V617F/G1849T mutation does not exclude the presence of other JAK2 mutations. Results were expressed as JAK2 V617F % = [CN V617F/(CN V617F + CN WT)] × 100.
Statistical analysis
Normally distributed variables are expressed as the mean 6 SD or median with range in states of non-normality as identified by the Kolmogorov–Smirnov test. Hemoglobin levels were used to define groups for analysis: healthy, borderline, and CMS (Tables 1 and 2). Nonparametric variables were expressed as the median with range, and group comparisons were performed using the Kruskal–Wallis test. Groups were compared on the basis of CMS scores using one-way analysis of variance (ANOVA) followed by the Tukey post hoc test (Table 3). A stepwise multivariable regression model was built to identify independent predictors of low CMS score. A p value < 0.05 was considered to indicate statistical significance in all tests. Data analysis was performed using PASW Statistics (Version 17.0, IBM, Armonk, NY).
Table 1.
Characteristics of adult Andean highlanders
| Men |
Women |
|||||
|---|---|---|---|---|---|---|
| Has (Hb ≤18 g/dL) | BDLs (Hb 18–20.9 g/dL) | Those with CMS (Hb ≥21 g/dL) | Has (Hb ≤16 g/dL) | BDL (Hb 16–18.9 g/dL) | Those with CMS (Hb ≥19 g/dL) | |
| N | 21 | 26 | 49 | 26 | 22 | 9 |
| Age (years) | 35.1 ± 12.1 | 32.2 ± 0.8 | 43.3 ± 11.5a,e | 29 ± 9.9 | 38.7 ± 13.6a | 50.4 ± 17.7d |
| Height (cm) | 168.6 ± 6.1 | 162.7 ± 3.9a | 163.5 ± 5.3d | 150.7 ± 4 | 149.8 ± 5.6 | 148.6 ± 6.9 |
| Weight (kg) | 61.6 ± 8.9 | 62.7 ± 6.2 | 67 ± 11.6a | 55.2 ± 7.2 | 60 ± 8.9 | 63 ± 11.2 |
| Hb (g/dL) | 17.4 ± 0.5e,f | 19.3 ± 0.9d,f | 22 ± 1.6d,e | 15.7 ± 0.9e,f | 17.4 ± 0.7d,f | 21.3 ± 1.6d,e |
| Median CMS score | 3f | 4f | 10d,e | 2b,f | 3a,f | 10a,e |
| SBP (mm Hg) | 109.3 ± 10.9 | 110 ± 16.3 | 116.6 ± 6 19 | 103.5 ± 11.3c | 109.1 ± 15.3 | 121.7 ± 20.9a |
| DBP (mm Hg) | 69.1 ± 7.7f | 69.8 ± 13.9 | 76.6 ± 6 13d | 62.5 ± 9.6c | 65.9 ± 8.8 | 76.7 ± 16.8a |
| HR (beats/min) | 65.3 ± 1 | 70.9 ± 1 | 70 ± 8.6 | 76.9 ± 1.1b | 70.2 ± 8a,c | 79.2 ± 9.2b |
| SaO2 (%) | 88.2 ± 2.4f | 86 ± 4.7c | 83.8 ± 2.9b,d | 87.2 ± 3.1f | 86.4 ± 3.1f | 81.6 ± 3d,e |
Values are means ± SD or medians.
p < 0.05: vs. HAs.
p < 0.05: vs. BDLs.
p < 0.05: vs. those with CMS, by Kruskal–Wallis test.
p < 0.01: vs. HAs.
p < 0.01: vs. BDLs.
p < 0.01: vs. those with CMS, by Kruskall–Wallis test.
Table 2.
Distribution of symptoms and signs in high-altitude adults
| Men |
Women |
|||||
|---|---|---|---|---|---|---|
| Symptom or sign | HAs (N = 21) | BDLs (N = 26) | Those with CMS (N = 49) | HAs (N = 26) | BDLs (N = 22) | Those with CMS (N = 9) |
| Symptoms/signs used in CMS scoring | ||||||
| Breathlessness and/or palpitationsa | ||||||
| Absence | 11 (52) | 17 (65) | 18 (37) | 14 (54) | 15 (68) | 2 (22) |
| Mild | 7 (33) | 9 (35) | 23 (47) | 12 (46) | 5 (23) | 5 (56) |
| Moderate | 3 (14) | 0 (0%) | 8 (16) | 0 (0) | 2 (9) | 2 (22) |
| Severe | 0 (0) | 0 (%) | 0 (%) | 0 (0) | 0 (0) | 0 (0) |
| Sleep disturbance | ||||||
| Absence | 11 (52) | 17 (65) | 28 (57) | 21 (81) | 13 (59) | 4 (45) |
| Mild | 5 (29) | 6 (23) | 9 (18) | 4 (15) | 6 (27) | 3 (33) |
| Moderate | 5 (29) | 3 (12) | 10 (20) | 1 (4) | 3 (14) | 2 (22) |
| Severe | 0 (0) | 0 (0) | 2 (4) | 0 (0) | 0 (0) | 0 (0) |
| Cyanosisb | ||||||
| Absence | 12 (57) | 7 (27) | 9 (18) | 19 (73) | 8 (36) | 1 (11) |
| Mild | 7 (33) | 12 (46) | 8 (16) | 6 (23) | 9 (41) | 1 (11) |
| Moderate | 2 (10) | 7 (27) | 18 (37) | 1 (4) | 5 (23) | 5 (56) |
| Severe | 0 (0) | 0 (0) | 14 (29) | 0 (0) | 0 (0) | 2 (22) |
| Dilatation of veinsb | ||||||
| Absence | 14 (67) | 11 (42) | 10 (20) | 23 (88) | 14 (64) | 6 (67) |
| Mild | 6 (28) | 12 (46) | 21 (43) | 2 (8) | 8 (36) | 1 (11) |
| Moderate | 1 (5) | 2 (8) | 15 (31) | 1 (4) | 0 (0) | 2 (22) |
| Severe | 0 (0) | 1 (4) | 3 (6) | 0 (0) | 0 (0) | 0 (0) |
| Paresthesia | ||||||
| Absence | 12 (57) | 15 (57) | 24 (49) | 17 (65) | 12 (54) | 5 (56) |
| Mild | 7 (33) | 8 (31) | 17 (35) | 9 (35) | 8 (36) | 2 (22) |
| Moderate | 2 (10) | 3 (12) | 6 (12) | 0 (0) | 1 (5) | 2 (22) |
| Severe | 0 (0) | 0 (0) | 2 (4) | 0 (0) | 1 (5) | 0 (0) |
| Headacheb | ||||||
| Absence | 12 (57) | 13 (50) | 8 (16) | 19 (73) | 11 (50) | 2 (22) |
| Mild | 6 (29) | 10 (38) | 28 (57) | 6 (23) | 10 (45) | 2 (22) |
| Moderate | 3 (14) | 3 (12) | 10 (20) | 1 (4) | 1 (5) | 4 (45) |
| Severe | 0 (0) | 0 (0) | 3 (6) | 0 (0) | 0 (0) | 1 (11) |
| Tinnitusa | ||||||
| Absence | 12 (57) | 9 (34) | 20 (41) | 20 (77) | 17 (77) | 1 (11) |
| Mild | 5 (24) | 15 (58) | 18 (37) | 6 (23) | 4 (18) | 4 (45) |
| Moderate | 4 (19) | 2 (8) | 9 (18) | 0 (0) | 1 (5) | 3 (33) |
| Severe | 0 (0) | 0 (0) | 2 (4) | 0 (0) | 0 (0) | 1 (11) |
| Other symptoms and signs | ||||||
| Dizziness | ||||||
| Absence | 19 (95) | 25 (97) | 44 (90) | 25 (96) | 21 (96) | 7 (78) |
| Mild | 1 (5) | 1 (3) | 5 (10) | 1 (4) | 1 (4) | 2 (22) |
| Moderate | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
| Severe | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
| Fatigue | ||||||
| Absence | 11 (53) | 12 (46) | 18 (37) | 14 (54) | 12 (54) | 5 (56) |
| Mild | 7 (33) | 11 (42) | 20 (41) | 8 (31) | 10 (46) | 2 (22) |
| Moderate | 3 (14) | 3 (12) | 9 (18) | 4 (15) | 0 (0) | 2 (22) |
| Severe | 0 (0) | 0 (0) | 2 (4) | 0 (0) | 0 (0) | 0 (0) |
| Failing memory | ||||||
| Absence | 8 (38) | 12 (46) | 19 (39) | 15 (58) | 11 (50) | 4 (44) |
| Mild | 11 (52) | 10 (39) | 24 (49) | 9 (34) | 10 (46) | 4 (44) |
| Moderate | 1 (5) | 4 (15) | 5 (10) | 2 (8) | 1 (4) | 1 (12) |
| Severe | 1 (5) | 0 (0) | 1 (2) | 0 (0) | 0 (0) | 0 (0) |
All values are percentages.
p < 0.05 in women.
p < 0.05 in men and women.
Table 3.
Analysis of hemoglobin, plasma EPO, and plasma pro-BNP levels
| Group | CMS score | Hemoglobin (g/dL) | EPO (mIU/mL) | Pro-BNP (pg/mL) |
|---|---|---|---|---|
| Healthy | ≤5 | 17.1 ± 0.2 (N = 101) | 17.7 ± 2.4 (N = 47) | 27.5 ± 3.1 (N = 47) |
| Mild CMS | 6–10 | 20.6 ± 0.4a (N = 45) | 12.6 ± 1.6 (N = 30) | 66.5 ± 14.1a (N = 30) |
| Moderate CMS | 11–14 | 22.3 ± 0.4a (N = 22) | 11.2 ± 2.3 (N = 15) | 42.0 ± 9.1 (N = 15) |
| Severe CMS | ≥15 | 22.5 ± 1.3a (N = 4) | 10.6 ± 2.7 (N = 2) | 47.5 ± 23.5 (N = 2) |
The normal range for EPO at our institution is 3.7–31.5 and 0–124 for pro-BNP. Blood samples were drawn throughout the day. Men and women were included in each group. The distribution of male and female individuals is 1:1 for healthy, 2:1 for mild CMS, 4:1 for moderate CMS, and for severe CMS, 1 man and 1 woman. There were no statistical differences between men and women.
p < 0.001 when compared with healthy group by one-way ANOVA, Tukey–Kramer method.
p = 0.005 when compared with healthy group by one-way ANOVA, Tukey–Kramer method.
Results
Demographics
One hundred fifty-three adults, 96 men (63%) and 57 women (37%), participated in the study. The mean age for all adults was 37.6 ± 13.1 years (range: 18–86). HAs included 21 men and 26 women; BDLs, 26 men and 22 women; and the CMS group, 49 men and 9 women (4 were premenopausal and 5 postmenopausal). Ten BDLs had mild CMS scores. HAs were taller and weighed less, but these differences were not statistically significantly between groups or sexes (Table 1).
Description of individuals with CMS
Adults with CMS had higher BP (Table 1). Seventeen percent of those with CMS had systolic BPs (SBPs) >140 mm Hg versus 10% of BDLs and 0% of HAs. Nineteen percent with CMS had diastolic BPs (DBPs) >90 mm Hg compared with 8% of BDLs and 4% of HAs. These increases may have resulted from the higher hemoglobin levels. CMS participants also had lower SaO2 and higher heart rate (HR). SaO2 levels <80% were measured in 6 participants with CMS, 2 BDLs, and 1 HA.
The prevalence of symptoms among HAs, BDLs, and individuals with CMS are detailed in Table 2. Approximately one-half of men and two-thirds of women had no symptoms. Thirty-one adults rated their symptoms as mild (26 men, 5 women), 22 as moderate (20 men, 2 women), and 5 as severe (3 men, 2 women). The most frequent symptoms were cyanosis, dilatation of vessels, and headaches in men, and cyanosis, headaches, and tinnitus in women. Dizziness was the least common symptom. Adults with CMS had a mean score of 10. Hemoglobin levels >25 g/dL were observed in seven participants; all were in men with CMS. Ten participants had previous phlebotomies.
Fetal hemoglobin
In our first analysis, the entire group of 153 individuals had normal HbF levels: median of 0.4% (range: 0.2%–1.9%, normal level: <1%). We then compared men and women as well as HAs and BDLs with individuals with CMS; there were no statistical differences in both analyses. Thus, HbF did not sufficiently differ and was not included in the subsequent statistical analyses.
Sanger sequencing of EPAS1, EGLN1, ANP32D, SENP1, and JAK2
Previous studies had indicated that high-altitude dwellers in Tibet harbored protective SNPs in EPAS1 (HIF-2α) and EGLN1 (PHD2). Thus, we selected a larger cohort, 26 men with the highest hemoglobin and 13 individuals with normal hemoglobin levels, and performed Sanger sequencing in these two genes. There were no differences in the EPAS1 gene between those with high hemoglobin levels and those with normal levels; all had the C/C genotype. The T allele in EGLN1 (rs480902) was in 69% versus 85% (p 5 0.14).
Genotyping by Sanger sequencing of ANP32D (rs72644851) revealed that the G allele in those with high versus normal hemoglobin levels was 64% versus 96% (p = 0.002). The G allele in SENP1 (rs7963934) was in 84% versus 100% (p = 0.045). Additionally, 20 samples with high hemoglobin levels (range: 21–26 g/dL, and one sample with hemoglobin of 18 g/dL) were subjected to JAK2 V167F mutation analysis to rule out another cause of erythrocytosis: 13 had a borderline number of wildtype DNA copies (between 1,092 and 7,142, because of low DNA concentrations), and 7 had a sufficient number of wild-type DNA copies (O7,142). No mutation for JAK2 V617F was detected.
Plasma levels of EPO and pro-BNP and multivariate analysis
Among the participants, 45 had mild CMS (score: 6–10), 22 had moderate CMS (score: 11–14), and 4 had severe CMS (score: ≥15). Approximately one-half to two-thirds of each group had a sufficient volume of plasma for EPO and pro-BNP determinations (Table 3). Because the hemoglobin level is a large portion of the CMS score, higher mean hemoglobin correlated with higher CMS score: 17.1 ± 0.2 g/ dL in the healthy group and 22.5 ± 1.3 in the severe CMS group (p < 0.001). There were no differences in mean plasma EPO levels between groups (p = 0.24) or between men and women. In contrast, the healthy group had the lowest pro-BNP levels (27.5 ± 3.1 pg/mL), and the mild CMS group had the highest levels (66.5 ± 14.1 pg/mL), but these were still within the normal range (0–124 pg/mL).
Univariate analysis of SENP1 showed the G/G genotypes correlated with lower hemoglobin levels (p = 0.005) and lower CMS scores (p = 0.035), but not with any plasma factors or SaO2. We then performed a logistic stepwise regression multivariable analysis of CMS scores using age, sex, EPO, pro-BNP, SaO2, SENP1 genotypes, and ANP32D genotypes. The SENP1 C/G or C/C SNPs were significantly associated with higher CMS score, compared with G/G (p = 0.014). When the hemoglobin level was included in the multivariate model, hemoglobin level was significantly associated with CMS scores (p = 0.004), and SENP1 remained significant (p = 0.024).
Discussion
In this large cross-sectional study of high-altitude dwellers in Cerro de Pasco, Peru, our cohort had a higher prevalence of CMS (38%) than in previous reports, likely because of our efforts to recruit individuals with CMS. This rate was reflected in a larger proportion with mild and moderate disease (20.2% mild, 14.4% moderate, and 3.3% severe CMS). We first tested our primary hypothesis that healthy high-altitude Andeans have higher HbF levels. HbF testing in 153 individuals revealed normal values (median of 0.4%), which were similar to results at sea level. Therefore, mechanisms other than HbF are responsible for protection against CMS.
The HIF pathway has been studied at high altitude in healthy highlanders and individuals with CMS. Recent reports confirmed that the protective gene signatures in EPAS1 and EGLN1 have been identified in Tibetans only, and these alleles were acquired many years ago [11,15,28]. However, the results in these genetic analyses in Andeans have not been definitive. Appenzeller et al. reported correlations between the higher protein expression levels of HIF-1α, CMS-related symptoms at high altitude, and CMS score [29]. However, Mejia et al. sequenced the SNPs in similar genes that regulate the stabilization and HIF-1α degradation and found no genetic association between HIF-related genes and individuals with CMS or high-altitude controls [30]. This is not surprising because these EPAS1 genotype results were detected in Himalayan highlanders only and were likely transferred from the Denisovan genome [28,31].
In whole-genome sequencing recently performed only in Andeans, there were no differences in EPAS1 or EGLN1 between 10 participants with severe CMS (score: ≥15) and 10 healthy high-altitude dwellers, but there was a large frequency differential in SNPs in SENP1 and ANP32D and increased hypoxia-induced expression of both genes in cultured skin fibroblasts from the same individuals with CMS [32]. The higher frequency in SENP1 SNP in individuals with CMS has recently been confirmed in a larger study [33]. SENP1 belongs to a family of proteases that are present in the nucleus and cytoplasm, and it reverses the posttranslational modification of small ubiquitin-like modifiers (SUMOs). Higher SENP1 expression has been hypothesized to be associated with higher androgen activity in Andean highlanders with EE [34]. Higher expression of SENP1 has been documented to induce HIF-1α and reduce apoptosis, thus promoting the development or progression of prostate [35] and other cancer [36,37] cells. Recently, absence of SENP1 has been reported to produce inflammatory cytokines, leading to type 1 diabetes in a murine model [38], and reduced expression led to impaired glucosedependent insulin exocytosis in an in vitro human islet model [39]. Unfortunately, we were not able to test the protease activities of SENP1 and ANP32D to correlate the genotypes and expression levels. Certainly, our understanding of SENP1 in hypoxia, cancer development, and other disease models is growing.
Our genotyping by Sanger sequencing of EPAS1 and EGLN1 revealed that the SNPs associated with low hemoglobin in Tibetans were not found at high frequency among Andean highlanders. We genotyped two other genes (SENP1 and ANP32D) and found that only SENP1 was associated with CMS. This finding is similar to previous reports that defined CMS cases as individuals with a CMS score O15 [32] or O11 [33]. Our analyses differed from those reports, with the majority of our cohort having mild (score: 6–10) and moderate (score: 11–14) CMS. Therefore, we were able to further correlate the genotypes of SENP1 and ANP32D with the severity of CMS. Specifically, the G/G genotype in SENP1 was associated with a milder phenotype and could differentiate healthy Andeans from Andeans with CMS.
We then explored the possibility of blood markers that predict the severity of CMS. EPO has been reported to be present at higher levels in Andean and Oromo highlanders than in those living at sea level [40,41], but no differences have been detected between healthy high-altitude dwellers and those with EE. Other investigations have determined the diurnal patterns of EPO levels among high-altitude dwellers and found that those with EE and CMS had persistently higher EPO levels, without the usual circadian variation of peaks in the evening and low levels in the morning [42]. When we separated the CMS group into mild, moderate, and severe categories to correlate the differences in EPO levels with CMS severity, we found normal levels in all participants (Table 3). We also performed JAK2 V617F mutation analysis and found only the normal (wild-type) allele, indicating that no other cause of EE, such as a myeloproliferative disorder, was present.
Previous studies in Han Chinese and Tibetans have reported high levels of BNP, vascular endothelial growth factor, and endothelin-1 in individuals with CMS compared with healthy highlanders [43,44]. We measured pro-BNP levels in HAs and those with CMS, but obtained different results. The highest levels were measured in the mild CMS group, and higher CMS scores did not correlate with higher pro-BNP levels, likely because of the small number of participants with severe CMS (N = 2). All these levels were still within the normal range; thus, levels of pro-BNP or EPO were not able to differentiate healthy individuals from those with CMS.
Conclusions
In this cross-sectional cohort of Andean highlanders, we achieved our primary aim and concluded that modest increases in levels of HbF were not present in well-adapted individuals. The severity of CMS was associated with higher hemoglobin levels, and not with EPO or pro-BNP levels. Protective SNPs in EPAS1 and EGLN1 reported in Tibetans were not found in our cohort. SENP1, and not the classic HIF-regulated genes, may differentiate healthy Andean highlanders from those with CMS.
Supplementary Material
Acknowledgments
We wish to thank Maxine Weissman (DLM/CC), Jodie Keary (DLM/CC), Cheryl Johnson (DLM/CC), Jordan Perlman (MCHB/NHLBI), and Oswald Phang (MCHB/NHLBI) for their technical assistance. This work is supported in part by the intramural research program at the Clinical Center, the National Heart, Lung, and Blood Institute, and the National Institute of Diabetes, Digestive, and Kidney Diseases at the NIH.
Footnotes
Conflict of interest disclosure
No financial interest/relationships with financial interest relating to the topic of this article have been declared.
References
- 1.Leon-Velarde F, Maggiorini M, Reeves JT, et al. Consensus statement on chronic and subacute high altitude diseases. High Alt Med Biol 2005;6:147–157. [DOI] [PubMed] [Google Scholar]
- 2.Pasha MA, Newman JH. High-altitude disorders: Pulmonary hypertension, pulmonary vascular disease, the global perspective. Chest 2010; 137:13S–19S. [DOI] [PubMed] [Google Scholar]
- 3.Arregui A, Leon-Velarde F, Cabrera J, Paredes S, Vizcarra D, Umeres H. Migraine, polycythemia and chronic mountain sickness. Cephalalgia 1994;14:339–341. [DOI] [PubMed] [Google Scholar]
- 4.Monge C, Leon-Velarde F, Arregui A. Increasing prevalence of excessive erythrocytosis with age among healthy high-altitude miners. N Engl J Med 1989;321:1271. [DOI] [PubMed] [Google Scholar]
- 5.Ge RL, Helun G. Current concept of chronic mountain sickness: Pulmonary hypertension-related high-altitude heart disease. Wilderness Environ Med 2001;12:190–194. [DOI] [PubMed] [Google Scholar]
- 6.Appenzeller O, Claydon VE, Gulli G, et al. Cerebral vasodilatation to exogenous NO is a measure of fitness for life at altitude. Stroke 2006; 37:1754–1758. [DOI] [PubMed] [Google Scholar]
- 7.Beall CM. High-altitude adaptations. Lancet 2003;362(Suppl):s14–s15. [DOI] [PubMed] [Google Scholar]
- 8.Hainsworth R, Drinkhill MJ. Cardiovascular adjustments for life at high altitude. Respir Physiol Neurobiol 2007;158:204–211. [DOI] [PubMed] [Google Scholar]
- 9.Leon-Velarde F, Mejia O. Gene expression in chronic high altitude diseases. High Alt Med Biol 2008;9:130–139. [DOI] [PubMed] [Google Scholar]
- 10.Aggarwal S, Negi S, Jha P, et al. EGLN1 involvement in high-altitude adaptation revealed through genetic analysis of extreme constitution types defined in Ayurveda. Proc Natl Acad Sci U S A 2010;107: 18961–18966. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bigham A, Bauchet M, Pinto D, et al. Identifying signatures of natural selection in Tibetan and Andean populations using dense genome scan data. PLoS Genet 2010;6:e1001116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Beall CM, Cavalleri GL, Deng L, et al. Natural selection on EPAS1 (HIF2alpha) associated with low hemoglobin concentration in Tibetan highlanders. Proc Natl Acad Sci U S A 2010;107:11459–11464. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Yi X, Liang Y, Huerta-Sanchez E, et al. Sequencing of 50 human exomes reveals adaptation to high altitude. Science 2010;329:75–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Buroker NE, Ning XH, Zhou ZN, et al. EPAS1 and EGLN1 associations with high altitude sickness in Han and Tibetan Chinese at the Qinghai-Tibetan Plateau. Blood Cells Mol Dis 2012;49:67–73. [DOI] [PubMed] [Google Scholar]
- 15.Bigham AW, Wilson MJ, Julian CG, et al. Andean and Tibetan patterns of adaptation to high altitude. Am J Hum Biol 2013;25:190–197. [DOI] [PubMed] [Google Scholar]
- 16.Cook CD, Brodie HR, Allen DW. Measurement of fetal hemoglobin in newborn infants; Correlation with gestational age and intrauterine hypoxia. Pediatr 1957;20:272–278. [PubMed] [Google Scholar]
- 17.Bureau MA, Shapcott D, Berthiaume Y, et al. Maternal cigarette smoking and fetal oxygen transport: A study of P50, 2,3-diphosphoglycerate, total hemoglobin, hematocrit, and type F hemoglobin in fetal blood. Pediatr 1983;72:22–26. [PubMed] [Google Scholar]
- 18.Bard H, Lachance C, Widness JA, Gagnon C. The reactivation of fetal hemoglobin synthesis during anemia of prematurity. Pediatr Res 1994;36:253–256. [DOI] [PubMed] [Google Scholar]
- 19.Ballew C, Haas JD. Hematologic evidence of fetal hypoxia among newborn infants at high altitude in Bolivia. Am J Obstet Gynecol 1986;155:166–169. [DOI] [PubMed] [Google Scholar]
- 20.Versmold HT, Linderkamp C, Dohlemann C, Riegel KP. Oxygen transport in congenital heart disease: Influence of fetal hemoglobin, red cell pH, and 2,3-diphosphoglycerate. Pediatr Res 1976;10:566–570. [DOI] [PubMed] [Google Scholar]
- 21.Reynafarje C, Faura J, Villavicencio D, et al. Oxygen transport of hemoglobin in high-altitude animals (Camelidae). J Appl Physiol 1975; 38:806–810. [DOI] [PubMed] [Google Scholar]
- 22.DeSimone J, Biel SI, Heller P. Stimulation of fetal hemoglobin synthesis in baboons by hemolysis and hypoxia. Proc Natl Acad Sci U S A 1978;75:2937–2940. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.DeSimone J, Heller P, Biel M, Zwiers D. Genetic relationship between fetal Hb levels in normal and erythropoietically stressed baboons. Br J Haematol 1981;49:175–183. [DOI] [PubMed] [Google Scholar]
- 24.DeSimone J, Biel M, Heller P. Maintenance of fetal hemoglobin (HbF) elevations in the baboon by prolonged erythropoietic stress. Blood 1982;60:519–523. [PubMed] [Google Scholar]
- 25.Semenza GL. HIF-1, O2, and the 3 PHDs: How animal cells signal hypoxia to the nucleus. Cell 2001;107:1–3. [DOI] [PubMed] [Google Scholar]
- 26.Hsieh MM, Linde NS, Wynter A, et al. HIF-prolyl hydroxylase inhibition results in endogenous erythropoietin induction, erythrocytosis, and fetal hemoglobin expression in rhesus macaques. Blood 2007; 110:2140–2147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Chen B, Wilkening S, Drechsel M, Hemminki K. SNP_tools: A compact tool package for analysis and conversion of genotype data for MS-Excel. BMC Res Notes 2009;2:214. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Huerta-Sanchez E, Asan Jin X, et al. Altitude adaptation in Tibetans caused by introgression of Denisovan-like DNA. Nature 2014;512: 194–197. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Appenzeller O, Minko T, Qualls C, et al. Gene expression, autonomic function and chronic hypoxia: Lessons from the Andes. Clin Auton Res 2006;16:217–222. [DOI] [PubMed] [Google Scholar]
- 30.Mejia OM, Prchal JT, Leon-Velarde F, Hurtado A, Stockton DW. Genetic association analysis of chronic mountain sickness in an Andean high-altitude population. Haematologica 2005;90:13–19. [PubMed] [Google Scholar]
- 31.Lou H, Lu Y, Lu D, et al. A 3.4-kb copy-number deletion near EPAS1 is significantly enriched in high-altitude Tibetans but absent from the Denisovan sequence. Am J Hum Genet 2015;97:54–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Zhou D, Udpa N, Ronen R, et al. Whole-genome sequencing uncovers the genetic basis of chronic mountain sickness in Andean highlanders. Am J Hum Genet 2013;93:452–462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Cole AM, Petousi N, Cavalleri GL, Robbins PA. Genetic variation in SENP1 and ANP32D as predictors of chronic mountain sickness. High Alt Med Biol 2014;15:497–499. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Gonzales GF, Chaupis D. Higher androgen bioactivity is associated with excessive erythrocytosis and chronic mountain sickness in Andean highlanders: A review. Andrologia 2015;47:729–743. [DOI] [PubMed] [Google Scholar]
- 35.Burdelski C, Menan D, Tsourlakis MC, et al. The prognostic value of SUMO1/Sentrin specific peptidase 1 (SENP1) in prostate cancer is limited to ERG-fusion positive tumors lacking PTEN deletion. BMC Cancer 2015;15:538. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Zhang W, Sun H, Shi X, et al. SENP1 regulates hepatocyte growth factor-induced migration and epithelial–mesenchymal transition of hepatocellular carcinoma. Tumour Biol 2015. December 22; [Epub ahead of print]. [DOI] [PubMed]
- 37.Ao Q, Su W, Guo S, Cai L, Huang L. SENP1 desensitizes hypoxic ovarian cancer cells to cisplatin by up-regulating HIF-1alpha. Sci Rep 2015;5:16396. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Shao L, Zhou HJ, Zhang H, et al. SENP1-mediated NEMO deSUMOylation in adipocytes limits inflammatory responses and type-1 diabetes progression. Nat Commun 2015;6:8917. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Ferdaoussi M, Dai X, Jensen MV, et al. Isocitrate-to-SENP1 signaling amplifies insulin secretion and rescues dysfunctional beta cells. J Clin Invest 2015;125:3847–3860. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Leon-Velarde F, Monge CC, Vidal A, Carcagno M, Criscuolo M, Bozzini CE. Serum immunoreactive erythropoietin in high altitude natives with and without excessive erythrocytosis. Exp Hematol 1991;19: 257–260. [PubMed] [Google Scholar]
- 41.Schmidt W, Spielvogel H, Eckardt KU, Quintela A, Penaloza R. Effects of chronic hypoxia and exercise on plasma erythropoietin in high-altitude residents. J Appl Physiol 1985;1993(74):1874–1878. [DOI] [PubMed] [Google Scholar]
- 42.Bernardi L, Roach RC, Keyl C, et al. Ventilation, autonomic function, sleep and erythropoietin. Chronic mountain sickness of Andean natives. Adv Exp Med Biol 2003;543:161–175. [PubMed] [Google Scholar]
- 43.Ge RL, Mo VY, Januzzi JL, et al. B-Type natriuretic peptide, vascular endothelial growth factor, endothelin-1, and nitric oxide synthase in chronic mountain sickness. Am J Physiol Heart Circ Physiol 2011; 300:H1427–H1433. [DOI] [PubMed] [Google Scholar]
- 44.Ge RL, Shai HR, Takeoka M, et al. Atrial natriuretic peptide and red cell 2,3-diphosphoglycerate in patients with chronic mountain sickness. Wilderness Environ Med 2001;12:2–7. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.