Index Introduction Methods DNA analysis Study population Measurements Statistical analysis Results Discussion References

Introduction

Individual differences in human phenotypes exist. According to Baker (1997), these individual differences are formed from influences of genotype, environment, and culture. Modern humans (Homo sapiens) spread from Africa to all parts of the world, and it is thought that, in the process, they adapted to various environments, particularly to cold climates. For that reason, cold tolerance (adaptation), a phenotype, has long been studied from the perspective of group and individual differences. General cold adaptation is known to include metabolic adaptation (Andersen et al., 1963), isolative adaptation (Hammel et al., 1959), and hypothermic adaptation (Carlson et al., 1953). Energy metabolism accompanying thermoregulation during cold exposure is an important physiological response for the evaluation of cold tolerance (Van Ooijen et al., 2001). Individual differences in energy metabolism accompanying thermoregulation are affected by age, sex, physique, diurnal variations, season, and lifestyle (Castellani et al., 1999; Van Someren et al., 2002; Mäkinen et al., 2004; Van Ooijen et al., 2004; Maeda et al., 2005). The genetic factors related to individual differences in thermoregulation and accompanying energy metabolism are almost entirely unknown.

In cold environments, the first response is for blood vessels to constrict and regulate heat loss. The human response to even stronger cold exposure depends on two types of thermogenesis: shivering thermogenesis and non-shivering thermogenesis. Shivering thermogenesis consumes 3–5 times the energy of the basal metabolic rate due to the heat energy produced by muscle activity as skeletal muscle consumes adenosine triphosphate (ATP). Non-shivering thermogenesis depends on innate heat capacity. It also depends on adaptive non-shivering thermogenesis that is upregulated when exposed to further cold. It has been suggested that skeletal muscle and brown adipose tissue are the heat sources in adaptive non-shivering thermogenesis (Astrup, 1986). Mitochondria containing UCP1 in BAT generate heat by uncoupling the respiratory chain of oxidative phosphorylation within mitochondria. Brown adipose tissue was previously thought to disappear in adult humans, but the presence of the uncoupling protein (UCP1) biochemical marker has been confirmed in adults (Cannon and Nedergaard, 2004). It has also been reported from measurements with positron emission tomography (PET) that some brown adipose tissue remains throughout adulthood (Nedergaard et al., 2007; Van Marken Lichtenbelt et al., 2009).

In recent years, there have been reports that the basal metabolic rate, which is closely related to energy metabolism, has a genetic component. As one such example, Leonard et al. (2005) reported that Inuit people and other groups living in Arctic regions have a higher basal metabolic rate than the value estimated from their body composition. A higher basal metabolic rate means higher non-shivering thermogenesis, suggesting that Inuit people have adapted genetically to the cold.

One of the genetic background factors for this may be mitochondria, which serve an important role in energy metabolism, and their genomes. Because of its evolutionary neutrality, mitochondrial DNA (mtDNA) is an important means of understanding human migrations with accompanying age estimates (Cann et al., 1987). The groups that have been formed in these migrations are called mitochondrial haplogroups. Some researchers have also claimed that, since ATP and heat are generated during oxidative phosphorylation of mitochondria, adaptations to cold have been made by mtDNA regulating the balance of ATP generation and thermogenesis in oxidative phosphorylation (Mishmar et al., 2003; Wallace, 2005; Torroni et al., 2006). From examinations of haplogroups worldwide, it has been reported that mtDNA polymorphisms that increase heat generation are seen in groups living in cold climates. Balloux et al. (2009) compared mtDNA, Y chromosomes, and single terminal repeats (STR), and reported that mtDNA is subjected to selective pressure from climate, since only mtDNA has lower diversity when minimum temperatures are low. Thus, people in cold climates may have mitochondria that generate more heat even while consuming the same amount of oxygen, by increasing the generation of heat from ATP. This suggests that, by adapting to climate, mtDNA affects the energy metabolism function in humans.

From the above, it is thought that mtDNA polymorphism is related to cold tolerance. There have been many earlier reports of a relation between haplogroup and physiological function. Examples include haplogroup J in Europe, which is associated with small maximum oxygen consumption (Marcuello et al., 2009), and haplogroup D in Japan, which tends to be associated with longevity (Tanaka et al., 1998; Bilal et al., 2008). Another example is that there are differences between groups in terms of the risk of acquiring diabetes mellitus or lifestyle-related diseases (Fuku et al., 2007; Tanaka et al., 2007). However, very few studies have attempted to explain physiological characteristics such as individual differences in human cold tolerance by mitochondrial polymorphism.

Japanese people have a large number of haplogroups. Both the M and N haplogroups exist in Japan, and many haplogroups are interwoven. This is thought to reflect the arrival in Japan of ancestral groups through various routes (Shinoda, 2009). Forty percent of all Japanese are haplogroup D, showing a frequency nearly two times that in other Asian regions. Haplogroup D groups have also settled in Arctic regions, and, according to Wallace (2005), they are groups that have adapted to cold climates. If haplogroup D Japanese, who live in a temperate zone, are found to have superior cold tolerance to other groups, we may gain some understanding of the genetic background for cold tolerance. In the present study, therefore, to elucidate individual differences and genetic influences in cold tolerance in humans, we focused on haplogroup D (D4) subjects, the largest group in Japan, and investigated whether they have superior cold tolerance to other groups (not predominantly northern) in Japan.

Methods

DNA analysis

DNA was collected from a hair shaft and analyzed. Total DNA was extracted from the hair shaft by digestion in extraction buffer using the ISOHAIR (NIPPON GENE Code No. 319–03401) in accordance with the manufacturer’s instructions. The nuclear mtDNA gene spacer D-loop was amplified by polymerase chain reaction (PCR) using primers M13RV-L15996 and M13(-21)-H408. The analyzed sequences of the D-loop primers were as follows: mtDNA L15996 (5′-CTCCACCATTAGCACCCAAAGC-3′) and mtDNA H408 (5′-CTGTTA AAAGTGCATACCGCCA-3′). The thermocycling profile consisted of an initial denaturation step at 94°C for 1 min, followed by 32 cycles of 30 s at 94°C, 30 s at 56°C and 75 s at 72°C. Purified DNA was sequenced in both directions using the ABI PRISM 310 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA) with a BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems).

The nucleotide sequences were aligned using Clustal W (Thompson et al., 1994). mtDNA were classified into haplogroups based on D-loop motifs of the specific sequences described in recent studies (Tanaka et al., 2004; Lee et al., 2006; Uchiyama et al., 2007). The mtDNA sequenced by M13 without setting the inner primers. The haplogoroups were determined based on the sequence length as follows: the sequence including stretch of C at the 16189 position were 16000–16189 and 16200–16310, the others were 16000–16310.

Study population

After obtaining consent from 50 male university students, their morphological characteristics and mtDNA were investigated. Participants were given an explanation of the experimental procedure, including its strain and danger to their body; agreements were obtained from 21 participants. To focus on genetic affects, variation in the morphological characteristics within subjects were minimized since cold adaptability depends to a large degree on morphological characteristics (Steegmann, 2007). As numbers were limited, participants were divided into group D (D4) and non-D (not predominantly northern). A total of 16 subjects therefore participated in the cold exposure experiment, including eight haplotype D (D4) students and eight students of other groups. The subjects were selected so that there were no significant differences in morphological characteristics (height, weight, body mass index, body surface area, body fat), as shown in Table 1. Body surface area (BSA) was calculated by Kurazumi’s formula (Kurazumi et al., 2003) and body fat was calculated by Brozek’s formula (Brozek et al., 1963). The subjects were born in Fukuoka Prefecture, or neighboring prefectures, and did not include any individuals who participated regularly in vigorous sports. The haplogroups of non-D subjects were M7a (four subjects), M7c (one subject), F2a (one subject), and B4 (two subjects).

In view of respect for human rights and privacy rights, mtDNA analysis was executed with approval from the Ethics Committee for Genome–Gene analysis of the Graduate School of Medicine, Kyushu University. In addition, mtDNA information obtained in our study was treated anonymously and managed by the Gene Therapeutic Information Center of this institution.

Measurements

The experiments were conducted in summer (August–September). Various measurement sensors were attached to subjects in an environment with a temperature of 27°C in preparation for the experiment. The subjects rested quietly for 15 min in an artificial climate chamber, and then the cold exposure commenced. The artificial climate chamber was programmed so that the ambient temperature dropped to 10°C in approximately 30 min, and this temperature was then maintained for for a further 60 min.

The parameters recorded were rectal temperature, skin temperature (seven locations), oxygen consumption, blood pressure, electrocardiogram, and a subjective evaluation. The rectal temperature probe was inserted to a depth of 13 cm beyond the anal sphincter. The skin temperature sensors were attached with surgical tape to measurement sites on the forehead, shoulder, chest, forearm, back of the hand, thigh, and dorsal side of the foot. Measurements were made continuously at intervals of 2 s using a data logger (LT-8A, Gram Corporation, Saitama, Japan). Mean skin temperature was calculated from the seven-point method of Hardy–DuBois. Oxygen consumption was measured with a respiratory gas analyzer (AE-300S, Minato Medical Science, Osaka, Japan) through a breathing tube using a mask to measure expired gas (Rudolph mask, Nihon Kohden, Tokyo, Japan). The measurements were made in expiratory mode. Inspired air was taken to be atmospheric concentration, and the oxygen concentration of expired air only was measured at 10 min intervals. Blood pressure was measured every 10 min using a digital automated sphygmomanometer (HEM-737 IntelliSense, Omron, Kyoto, Japan). A subjective evaluation (thermal comfort sensation) was made according to the following scores: very uncomfortable (−3), uncomfortable (−2), slightly uncomfortable (−1), normal (0), slightly comfortable (+1), comfortable (+2), very comfortable (+3).

Statistical analysis

Morphological data were compared by a unpaired t-test. Physiological data were compared using two-way (Haplogroup and Time) repeated-measures analysis of variance (ANOVA). The result of the ANOVA was as follows: F_(df1,df2) = variance ratio, df1 = degrees of freedom 1, df2 = degrees of freedom 2. Microsoft Excel 2010 was used for t-test and ANOVA4 (Kiriki, 2002) was used for two-way repeated-measures ANOVA. All data were expressed as the mean ± SD and P < 0.05 was considered to be statistically significant.

Results

ANOVA revealed significant differences in main effect of group (F_(1,14) = 5.717, P < 0.05) and main effect of time (F_(9,126) = 21.13, P < 0.001) for rectal temperature (Figure 1). Interaction was also significant (F_(9,126) = 3.81, P < 0.001). In a post-hoc test, the decrease in rectal temperature from 40 min after the start was significantly smaller in haplogroup D subjects than in the other groups. There were no significant main effects between groups in mean skin temperature (Figure 2) or oxygen consumption (Figure 3). There were also no main effects between groups for blood pressure or electrocardiograms.

View Details

Figure 1. Rectal temperature during cold exposure (closed squares: haplogroup D; open squares: haplogroup non-D with SD bars, *P < 0.05, **P < 0.01,***P < 0.005).

View Details

Figure 2. VO2 during cold exposure (closed squares: haplogroup D and open squares: haplogroup non-D with SD bars).

View Details

Figure 3. Mean skin temperature during cold exposure (closed squares: haplogroup D and open squares: haplogroup non-D with SD bars).

In the subjective evaluation (Figure 4), significant differences in main effect of group (F_(1,14) = 5.978, P < 0.05) and main effect of time (F_(9,126) = 105.344, P < 0.001), and in the results of a post-hoc test, haplogroup D subjects were judged to be significantly uncomfortable at 20 and 30 min after the start of cold exposure compared with the non-D groups (Figure 4).

View Details

Figure 4. Thermal comfort sensation during cold exposure (closed squares: haplogroup D and open squares: haplogroup non-D with SD bars, *P < 0.05).

Rectal temperature and oxygen consumption during cold exposure were plotted, and in haplogroup D subjects, oxygen consumption was seen to increase at an earlier time after rectal temperature decreased, compared with the other group (Figure 5).

View Details

Figure 5. Rectal temperature vs. VO2 during cold exposure (closed squares: haplogroup D and open squares: haplogroup non-D).

Discussion

The clearest result of this study is that the decrease in rectal temperature was significantly smaller in haplogroup D subjects than in the other group during cold exposure (P < 0.001, Figure 1). There was no significant difference between the haplogroups in oxygen consumption itself (Figure 2); that is, there was no difference between the haplogroups in energy consumed. Moreover, the fact that there was no significant difference in peripheral skin temperature, including mean skin temperature, at the four measurement sites on the limbs shows that there were no differences in thermal insulation from vasoconstriction. Therefore, the higher rectal temperature in haplotype D subjects than in the other subjects suggests that other factors that are not attributable to vascular constriction or oxygen consumption are involved.

One possible factor is that non-shivering thermogenesis is high in the body core in haplogroup D subjects. Brown adipose tissue is thought to contribute as a heat source. It is well known that brown adipose cells contain an abundance of mitochondria. In small animals, e.g. rats, body temperature is maintained by non-shivering thermogenesis rather than shivering. In humans, brown adipose tissue was previously thought to disappear with growth, but in recent years, brown adipose tissue has come to be considered a source of non-shivering heat generation in adults as well, and the level of its contribution to body temperature regulation has been investigated. The results have suggested that subjects with high brown adipose tissue activity maintain body temperature with heat generated in the body core (Van Marken et al., 2009). As a further genetic factor, from reports of high non-shivering thermogenesis in the greater white-toothed shrew, a high-frequency haplogroup that lives at high altitudes, a relation between haplogroup and non-shivering thermogenesis has been suggested (Fontanillas et al., 2005). Wallace claimed that the mtDNA of humans living in cold regions generates heat that is used in maintaining body temperature via ATP, using the same amount of oxygen and lipids/carbohydrates in oxidative phosphorylation. Therefore, in haplogroup D subjects, mitochondria generate more heat, and the larger thermogenesis in the core is thought to be one factor that limits the decrease in rectal temperature. The data in Figure 5 show that, when VO₂ is 7 ml/kg/min, haplotype D subjects maintain a significantly higher rectal temperature with the same oxygen consumption. Therefore, thermogenesis originating in brown adipose tissue is thought to increase the amount of thermogenesis of the entire core. The second factor, as claimed by Wallace (2005) is that since ATP generation and thermogenesis in mitochondria are affected by mitochondrial haplogroup, haplogroup D subjects have mitochondria that more easily generate heat in all cells, which may contribute to body temperature maintenance. Our finding of no significant difference between groups in body surface temperatures supports the claim of high thermogenesis in the core.

Incidentally, the ambient temperature at which increased thermogenesis starts in order to maintain the body temperature of animals in the cold is called the lower critical temperature. It is thought to be lower in species living in colder regions. This is thought to be because, as mentioned above, the limiting of heat loss in response to cold occurs first, and the lower the temperature at which increased thermogenesis is reached, the better the cold tolerance. This is also thought to be one indicator of cold tolerance in humans. According to Yoshimura and Yoshimura (1969), a lower critical temperature is closely related to accelerated metabolism, and the consumption of energy during cold exposure is greatly suppressed if the lower critical temperature is low. However, a metabolic adaptation that maintains body temperature by actively accelerating metabolism exists in very cold regions, such as in the Inuit people living in the Arctic zone (Hart et al., 1962; Andersen et al., 1963). In sub-Arctic zones, group physiological polymorphisms are seen in cold tolerance, such as in isolative adaptation that maintains the body temperature in the core in Australian Aborigines and other groups (Scholander et al., 1958; Hammel et al., 1959), and hypothermic adaptation that lowers the core and skin temperature in bushmen and others (Carlson et al., 1953). Behind such differences in adaptive type, typically, is thought to be a balance between thermogenesis and heat dissipation resulting from the living environment, including eating habits. Thus, since Inuit people can obtain high-energy animal protein and fat, they depend on energy metabolism to maintain body temperature. Conversely, since food is more difficult to obtain for aborigines and bushmen, it is conjectured that they need to conserve energy to maintain body temperature, and they acquired isolative adaptation to reduce heat loss. This means that it may not always be appropriate to evaluate cold tolerance by lower critical temperature only.

The relationship between rectal temperature and oxygen consumption is shown in Figure 5. After cold exposure, haplogroup D subjects quickly increased oxygen consumption and maintained rectal temperature at a certain level for some time. From the finding that oxygen consumption increased earlier in haplogroup D subjects than in the other groups, it is conjectured that the lower critical temperature in these groups is high. It has been reported that even when shivering is not produced in humans, non-shivering thermogenesis occurs before shivering thermogenesis, since oxygen consumption increases (Dauncey, 1981). Therefore, from the present results, it is thought that core body temperature in haplogroup D subjects is maintained by the early start of thermogenesis. Moreover, it may be that, in haplogroup D subjects, the quick judgment of discomfort during cold exposure is related to the high lower critical temperature (early start of increased thermogenesis), although it would be difficult to explain this in terms of mtDNA. In other groups, the lower critical temperature is presumed to be low from the low rectal temperature and gradual increase in oxygen consumption. However, since rectal temperature decreased despite increased oxygen consumption, cold tolerance cannot be said to be high based on the idea of lower critical temperature. According to past reports, there is great variation (21–27°C) in lower critical temperature in modern humans. There is variation in lower critical temperature even among groups of Japanese, with mean values reported to be 21.7°C (Ishii, 1976), 24°C (Yoshimura and Yoshimura 1969), and 26.2°C (Sato et al., 1979). The size of this variation in lower critical temperature between groups, considered together with the differences in cold tolerance response between haplogroups, suggests that there is an influence from the differences in non-shivering thermogenesis that reflects genetic factors.

The variation between groups in the lower critical temperature partially reflects group characteristics, such as thermogenic and isolative types. There are also individual differences in cold tolerance within groups, showing a tendency for differences in thermogenesis that depend on non-shivering thermogenesis. The difference in cold tolerance response between haplogroups in this study therefore suggests that genetic background is involved even in individual differences in cold tolerance response within a group. From the above, it may be that, during the cold exposure in the present experiment, the physiological characteristics of haplogroup D were maintenance of rectal temperature as a result of an early start of increased thermogenesis and large thermogenesis in the core. From the perspective of maintaining core body temperature, the haplogroup D group can said to be a cold-adapted group. The present study thus explains part of the variation in cold tolerance among Japanese based on genetic background.

The neutrality of mtDNA is still being debated by population geneticists and anthropologists. The differences in cold tolerance between haplogroups in this study support the assertion of Wallace that haplogroups are subjected to selective pressure by temperature. While even though haplogroup did not have functional meaning, it would consider about hitchhiking of nuclear genome because the haplogoroup might have a primary adaptation based on population structure. Thus, many points remain to be clarified in this study. Viewing the present results from the perspective of population genetics, confirmation will be needed using a larger number of subjects, since there were few parameters, and other genes may have been involved. Moreover, other genes for cold tolerance and the effects of gene expression also need to be considered. From reports that mitochondria increase as a result of exercise or cold stimulation (Puigserver et al., 1998), it is also possible that PGC1-α affects cold tolerance between groups according to season. Since polymorphisms also exist in PGC1-α (Hara et al., 2002), it is thought these might contribute to individual differences in cold tolerance. With respect to variation in the lower critical temperature as well, it is possible that, for example, TRP gene polymorphism or expression is involved. Considering the above, a genome-wide study will be needed in the future.