There has been considerable interest in composition of gut microbiota in recent years, leading to a better understanding of the role the gut microbiota plays in health and disease. Most studies have been limited in their geographical and socioeconomic diversity to high-income settings, and have been conducted using small sample sizes. To date, few analyses have been conducted in low-income settings, where a better understanding of the gut microbiome could lead to the greatest return in terms of health benefits. Here, we have used quantitative real-time polymerase chain reaction targeting dominant and sub-dominant groups of microorganisms associated with human gut microbiome in 115 people living a subsistence lifestyle in rural areas of Papua New Guinea. Quantification of Clostridium coccoides group, C. leptum subgroup, C. perfringens, Bacteroides fragilis group, Bifidobacterium, Atopobium cluster, Prevotella, Enterobacteriaceae, Enterococcus, Staphylococcus, and Lactobacillus spp. was conducted. Principle coordinates analysis (PCoA) revealed two dimensions with Prevotella, clostridia, Atopobium, Enterobacteriaceae, Enterococcus and Staphylococcus grouping in one dimension, while B. fragilis, Bifidobacterium and Lactobacillus grouping in the second dimension. Highland people had higher numbers of most groups of bacteria detected, and this is likely a key factor for the differences revealed by PCoA between highland and lowland study participants. Age and sex were not major determinants in microbial population composition. The study demonstrates a gut microbial composition with some similarities to those observed in other low-income settings where traditional diets are consumed, which have previously been suggested to favor energy extraction from a carbohydrate rich diet.
Citation: Greenhill AR, Tsuji H, Ogata K, Natsuhara K, Morita A, Soli K, et al. (2015) Characterization of the Gut Microbiota of Papua New Guineans Using Reverse Transcription Quantitative PCR. PLoS ONE 10(2): e0117427. https://doi.org/10.1371/journal.pone.0117427
Academic Editor: Jose Luis Balcazar, Catalan Institute for Water Research (ICRA), SPAIN
Received: November 9, 2013; Accepted: December 23, 2014; Published: February 6, 2015
Copyright: © 2015 Greenhill et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Funding: Funding was provided by Funding Program for Next Generation World-Leading Researchers (NEXT program): LS024 (Masahiro UMEZAKI, the University of Tokyo), http://www.jsps.go.jp/j-jisedai/. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have the following interests: Hirokazu Tsuji, Kiyohito Ogata and Koji Nomoto are employed by Yakult Central Institute. There are no patents, products in development or marketed products to declare. This does not alter the authors’ adherence to PLOS ONE policies on sharing data and materials.
The human digestive system comprises a huge number of bacteria (∼1014) which interact closely with the host to impact on our health. The gut microbiome aids digestion, thus delivering nutrients and vitamins; and helps prevent infection through protection against colonization and modulation of the immune system [1,2]. In addition to these fundamental roles, we are now recognizing the broader implications for human health that the gut microbiome may have. Over the past 10–12 years the potential interactions between the gut microbiome and non-communicable diseases such as obesity and autism spectrum disorder have been of great interest [3–6]. These and other studies have highlighted the need for a better understanding of the core composition of the gut microbiota, and the impact of diet and other factors on gut composition. As a consequence, and as a result of rapidly advancing techniques, there has been a rapid increase in studies investigating the microbiome of the human digestive system in recent years.
Researchers have sought descriptions of the composition of the gut microbiota, resulting in the notion of enterotypes, i.e. clusters of microbes in which the composition of the population is ‘driven’ by a genus or group of specific bacteria . Initially three principle human enterotypes were proposed; and were considered ‘cross-national’, in that they were not geographically confined. However, the conclusions were drawn from only 39 individuals, all from high-income countries. One year later, an independent study  based on 98 individuals looked at the impact of diet on gut microbial composition, and sought to determine the validity of the previously proposed enterotypes . Two major findings of the study were that the gut composition was better described by two enterotypes; and these enterotypes correlated with long-term diet. The two predominant enterotypes were a Bacteroides-dominated type and a Prevotella-dominated type. The Bacteroides-dominated enterotype was associated with high protein and animal fat consumption, whereas individuals with a high carbohydrate intake more commonly had a Prevotella dominated enterotype .
Prior to the proposal of enterotypes, a difference between the gut microbiota of European children to that of children consuming a traditional high fiber diet in Burkina Faso was demonstrated , suggesting that gut microbial composition is not consistently cross-national. In the Burkina Faso children the predominant species included Prevotella spp. and Xylanibacter spp., bacteria able to hydrolyse cellulose and xylan. It was hypothesized that this Prevotella-rich gut flora may enable children to maximize energy intake from their carbohydrate and fiber rich diet.
Despite the role the gut microbiome plays in human health, to date most studies investigating the microbial composition have focused on people living in socio-economically developed countries. Relatively few studies have investigated the composition of the gut microbiota in low socio-economic settings where the greatest improvement in health outcomes is needed; and specifically, where the burden of infectious diseases remains high. Based on the limited data to date, it appears that there are differences in the composition of gut microbiota, both within high-income European countries ; and particularly when comparing people from socio-economically developed settings to those that live a non-Western or traditional subsistence lifestyle [9,11]. Moreover, when detailed analyses have been conducted in low socio-economic settings, typically studies have focused on children and sample sizes have been small [9,12,13]. Two recent studies in Malawi have used appreciable sample sizes, although with the focus on children only [14,15]; with only one sizable study looking at adults and children from high- and low-income settings . Thus, there remain large gaps in our knowledge in this field, particularly from the Asia-Pacific region. To address these issues, we have used reverse transcriptase real-time PCR to determine the prevalence of the recognized important groups of bacteria in the gut of adults and children in the highlands and lowlands of Papua New Guineans living a predominantly traditional lifestyle.
Participation in the study was voluntary, and all participants or parent/guardian (for participants under the age of 18 years) provided written informed consent. Ethics approval was granted by the Papua New Guinea (PNG) Institute of Medical Research (IMR) institutional review board (Ethics 10.25), and the PNG Medical Research Advisory Council (Ethics # 11.25).
Samples were collected from two PNG highland regions (Asaro Valley, approximately 20 km from the provincial capital of Goroka in Eastern Highlands Province and the Tari Basin, Hela Province) and one lowland region (East Maprik, East Sepik Province). Participants from the highland regions consume sweet potatoes as their staple food, supplemented by edible leafy vegetables, while those from the lowland consume sago starch and edible leafy vegetables, and various starchy crops (i.e., banana, yam, taro). The participants from both regions occasionally consume pig meat, but the contribution to total protein intake is limited [17,18].
Samples were collected in highland regions in February to March, 2012; and in the lowland region in September, 2012. Demographic information (i.e. sex, age) and history of antibiotic use was collected by face-to-face interview. Age was estimated by referring to the demographic database completed by members of the research team during earlier fieldwork [17,19]. Height was measured to the nearest 1 mm using a field anthropometer (GPM, Switzerland). Weight measurements were taken to the nearest 0.1 kg using a digital scale (Tanita Japan), without shoes and in light clothing. Body Mass Index (BMI) was calculated as a function of body weight (kg) divided by squared height (m) and rounded to 1 decimal place.
Participants were given a clean plastic bag for faecal collection at the participant’s earliest convenience. Once a stool specimen had been passed it was given to a member of the study team, who preserved it in RNAlater (Life Technologies) to stabilise nucleic acids. Initially a 10-fold dilution (w:v) of stool was prepared in RNAlater.
A further 5-fold dilution of the stool sample in RNAlater (final ratio 1:50) was conducted. A 200 μl aliquot was washed in phosphate buffered saline, centrifuged, and the pellet stored at IMR laboratories at-80°C. The pellet was then sent to Japan for analysis. RNA was extracted using a modified method described by Matsuda et al. . Briefly, the thawed sample was resuspended in a solution containing 346.5 μl of RLT buffer (Qiagen GmbH, Hilden, Germany), 3.5 μl of β-mercaptoethanol, and 100 μl of Tris-EDTA buffer. Glass beads (300 mg; diameter, 0.1 mm) (Tomy Seiko, Tokyo, Japan) were added to the suspension, and the mixture was disrupted vigorously for 5min using a ShakeMaster Auto (Biomedical Science, Tokyo, Japan). 500 μl of acid phenol was added and mixed, and the mixture was incubated for 10 min at 60°C. After incubation, the mixture was added to 100 μl of chloroform-isoamilalcohol (24:1) and mixed by vortex. Following the centrifugation at 12,000 × g for 10 min at 4°C, 450 μl of the supernatant was collected and added to an equal volume of chloroform-isoamilalcohol. After mixing by vortex, the mixture was centrifuged at 12,000 × g for 5 min, 400 μl of supernatant was collected and subjected to isopropanol precipitation. Finally, the nucleic acid fraction was suspended in 200 μl of nuclease-free water (Ambion, Inc., Austin, TX, USA) and frozen at-80°C until use.
Bacterial microbiota composition
We conducted Yakult Intestinal Flora-SCAN (YIF-SCAN) analysis based on reverse transcription-quantitative PCR (RT-qPCR) analysis using methods previously described [20,21]. In brief, primers targeting either the 16S rRNA or 23S rRNA region of the genome were used to detect dominant and sub-dominant gut microbes. Target organisms and their lower limit of detection are listed in Table 1.RT-qPCR was conducted by using an OneStep RT-PCR kit (Qiagen GmbH, Hilden, Germany). RT-PCR amplification and detection were performed in 384-well optical plates on an ABI PRISM 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA, USA). The total bacterial count obtained by RT-qPCR is shown as the sum of the counts of 10 bacterial groups and one species. The count of Lactobacillus obtained by RT-qPCR is expressed as the sum of the counts of six subgroups and two species.
Quantification of bacterial populations was conducted by correlating RT-qPCR outputs to previously obtained qPCR, FISH and/or culture results .
Data were entered into a spreadsheet (Microsoft Excel) as log bacterial counts, and basic statistical analysis conducted. All statistical analysis was conducted using SPSS v20. Non Linear Principle Components Analysis, also known as Principle coordinates analysis (PCoA), was conducted using the original data and the CATPCA program in SPSS. All variables were transformed as spline ordinal (degree = 2, the number of internal knots = 2) and two components were extracted. Variables were removed from analysis if the total variance explained for the variable in the analysis was less than 25%. Relationships between gut microbial composition and age, sex, regions were sought. For age, study participants were grouped as <5 years old (children), 5–17 years old (older children and adolescents, but referred to as ‘adolescents’ to avoid confusion with the ‘children’ group), and ≥18 years (adults). Participants were categorised as either living in the highlands or the lowlands for regional analysis. Comparisons were conducted using the Mann-Whitney U test and Kruskal-Wallis test.
A total of 126 samples were collected: 29 from Asaro Valley, 60 from Tari Basin (collectively the highland samples) and 37 from Maprik (lowland samples). Of the 126 samples tested using reverse transcriptase qPCR, 11 samples failed to amplify and were excluded from analysis. An overview of the basic demographic data for the remaining 115 participants is provided in Table 2. The age range of participants was 2 to 66 years, with an average of 30 years (median 28 years). Twenty-six participants were under 18 years of age, but only three of the juveniles were under 5 years of age. The majority of participants (71/115) were male. Participants < 18 years old were excluded from BMI analysis.
Table 1 provides the average number of each bacterial group detected in the 115 study participants, and the frequency of detection (prevalence). The most abundant bacteria were the Clostridium leptum subgroup (log10 9.3±0.7), followed by Prevotella spp (log10 9.0±0.9), C. coccoides group (log10 8.9±0.7) and Atopobium cluster (log10 8.1±0.8). Each of these four groups of bacteria was present in ≥92% of samples. S1 Table provides the raw data.
The average number of Prevotella spp. was significantly greater than average number of Bacteroides fragilis group (log10 6.8±0.9) (S1 Fig.). Individually, 101 of 115 participants had a higher number of Prevotella spp. than B. fragilis group. In nine of the 14 participants for which B. fragilis group was more abundant than Prevotella spp, the later was not detected (detection limit of log105.0); and in one sample numbers of Prevotella spp. and B. fragilis group were comparable (log108.0). Due to nonlinear relationships present between variables in the data set, PCoA was used to extract two components. Prevotella grouped with six other groups of bacteria tested for in this study, whereas B. fragilis grouped with two other groups (Table 3). The two components explain 62.09% of the variance in the data set. The outcome of PCoA is illustrated in Fig. 1. The results of the PCoA support an inverse relationship between Prevotella spp. and B. fragilis group with the two factor loading for these species having opposite signs.
Prevotella groups into dimension 1 (red circle) with other commonly detected bacterial groups such as the clostridia, Enterobacteriaceae and Atopobium cluster. Bacteroides fragilis group is a component of dimesion 2 (blue eclipse) and has an inverse relationship with Lactobacillus and Bifidobacterium.
Statistical analysis was conducted to determine whether socio-demographic factors (sex, region, and age) impacted on the composition of the gut microbiota. Using PCoA outcomes, clustering occurred according to region (Fig. 2). No clear trends were observable by sex or age (Figs. 3 and 4). Specific comparisons were conducted for key microbial groupings primarily based of phyla (Bacteroidetes, Firmicutes, Proteobacteria and Actinobacteria), but also using total bacterial numbers and total Lactobacillus. Kolmogovov-Smirnov analysis revealed non-normal distribution of data. Appropriate descriptive statistic analyses were conducted, revealing differences in multiple bacterial populations in highland and lowland populations (Table 4), and relatively fewer differences due to age (Table 5) and sex (Table 6). The differences observed in bacterial populations of highland compared to lowland individuals were also observed when only adults were included in analysis (S2 Table).
There is a lower number of dimension 1 organisms in lowlanders than highlanders.
To date, human gut composition studies have lacked global representation; and emphasis has been placed on very detailed analysis of often undersized sample sizes. We have applied RT-qPCR to detect and quantify dominant and sub-dominant groups of microbes known to inhabit the human gut of Papua New Guinean people living a mostly traditional subsistence-based lifestyle. Our study revealed differences in the gut composition of highland and lowland people in PNG, and differences in children and adults. Also of interest is that Prevotella was detected in higher numbers than Bacteroides in the majority (88%) of study participants, and there was an inverse relationship between the two genera.
Differences in gut microbial composition based on location and age of study participants have been observed in other studies, though geographical variation has more commonly been investigated at the inter-country level rather than the intra-country level [9,16]. In this study we compared populations with similar socio-economic conditions and subsistence lifestyle. Of the seven phylogenetic groups that cluster in dimension 1 (Table 3), five were present in the vast majority of samples analysed (≥92%), and the enterococci were present in 84% of samples (Table 1). This component represents the organisms that are commonly present, and often present in high numbers relative to many of the other organisms detected in this study. The staphylococci were an exception to this observation; they clustered in dimension 1 despite a relatively low rate of detection and mean bacterial numbers. Organisms in dimension 1 (excluding staphylococci) contributed to the core gut microbiota of people in PNG.
In Papua New Guineans Prevotella predominated over Bacteroides: this is likely a reflection of the diet and subsistence lifestyle of the study participants. De Filippo et al  found Prevotella (along with Xylanibacter and Treponema) to be present in children from Burkina Faso but not in European children. The authors hypothesized that the presence of these genera were on account of the high fiber diet, and that the bacteria maximize energy extraction from plant polysaccharides. Wu and colleagues  recently confirmed a link between Bacteroides-rich enterotypes and high protein, high animal-fat diets, and Prevotella-rich enterotypes with high carbohydrate diets. Although “western foods” such as rice and tinned meat are available and commonly consumed in PNG, these foods contribute more to the diet of urbanized Papua New Guineans than people living a rural subsistence lifestyle. For the latter, the focus of this study, garden foods continue to primarily contribute to daily energy intake [21,22]. As such, much of the diet consists of complex carbohydrates rather than refined carbohydrates; and protein intake is lower than in western diets . A food frequency questionnaire recently conducted by our research team in the highland study sites confirmed that people living it these communities do not meet the biologically required protein intake .
Bäckhed et al  observed that Prevotella and Bacteroides coexist if the community is predominant in Firmicutes, whereas in communities dominated by Bacteroidetes the two Gram negative genera are mutually exclusive. In our study participants Prevotella and Bacteroides coexisted (both organisms present in 80% of samples), and Firmicutes were detected in higher numbers than Bacteroidetes. However, our methods did not comprehensively detect all bacteria within the Firmicutes and Bacteroidetes, thus it is difficult to make a direct comparison to other settings.
Although always represented in PNG samples and typically in high numbers, there are lower numbers of bacteria clustering in dimension 1 in lowland participants than in highland participants (Fig. 2 and Table 4). The causes of the difference between highland and lowland populations are difficult to ascertain in this study. People living in the PNG highlands (1,500–2,300 meters altitude) have a diet that differs to some degree from the lowland diet, though both are subsistence based and are broadly similar (high in carbohydrates, the primary source of carbohydrate varies; and low in protein). The homology/heterogeneity of the core gut microbiome across the genetic spectrum has not been fully elucidated [7,9]. If genetic factors influence human microbiome composition this may be a contributing factor; as PNG and nearby island communities are renowned for the genetic diversity of their people. In general lowland people have recent Austronesian influences on their genetic composition (including non-Papuan Melanesian, Polynesian and Micronesian influences), relative to the more remote highland people in whom genetic composition is more homogenous [26–28]. The true extent of microbial composition differences between highland and lowland populations, and the mechanisms of variation, warrants further investigation, ideally in conjunction with detailed analysis of dietary intake.
We observed differences in Actinobacteria (Bifidobacterium and Atopobium) and Lactobacillus in children, adolescents (children over 5 to 17 years old) and adults (Table 5). These findings correlate with recent findings, in which changes in the composition of the gut flora are shown to occur with age. Perhaps the greatest change occurs at the time of weaning, prior to which Bifidobacterium (or other genera of Actinobacteria) constitute a major proportion of the gut flora . However, the change is not as abrupt as once thought, with Bifidobacterium being more abundant in adolescents than adults in a study conducted in the USA . Differences in gut composition have also been noted in later life, though in participants of a greater age than our most senior study participants .
Our lactobacilli primer set detects a broad range of Lactobacillus spp. , included those identified as major genera in the human gut in independent studies [31,32]. Thus we were able to gain an approximation of the overall number of lactobacilli while also determining the predominant subgroups. In doing so we have shown L. ruminis subgroup to be the most prevalent and abundant group of lactobacilli in our study participants relative to other groups.
The association with increased Firmicutes in obese humans and animal models [4,6] has led to speculation about a role for these organisms in obesity. The majority of studies have been conducted in high income settings; however, Xu and colleagues  used quantitative PCR to demonstrate reduced Bacteriodetes and an increased Firmicutes:Bacteroidetes ratio in obese school children in the minority Kazakh people from under-developed farming communities in rural China. Our analyses did not use universal primers to target all Firmicutes and Bacteroidetes, thus it is difficult to interpret how our findings pertain to obesity. Moreover, the relationship between Firmicutes:Bacteroidetes ratio and obesity is more complex than initially thought [34,35]. Nonetheless, it is possible that high numbers of Firmicutes may be an evolutionary advantage in people living a truly subsistence lifestyle (such as the people of PNG) due to an increased energy harvesting efficiency.Metagenomic approaches are required to further elucidate relationships between gut microbiota and obesity in this and other Pacific populations, where obesity and type 2 diabetes are becoming problematic.
The application of RT-qPCR does not enable the detection of all bacteria in the gut, only those targeted by the primers used (Table 1). The primers used in this study have been shown to detect a large range of bacteria, albeit culturable species. However, the primers have been shown to amplify various species within the target taxonomic group, while being specific for the target group [36,37]. As such, the primers used in our study are likely to amplify closely-related non-culturable species should they be present. Importantly, the higher taxonomic groups found to predominate in this study using RT-qPCR are consistent with the findings of other studies using sequence based technologies and microarrays [8,38,39]; and Tap et al  found cloning-based 16S rRNA sequence data to closely match qPCR data for the dominant species in their study. Thus, while this analysis does not allow for the detection of the full breadth of microbes present, we have been able to characterize the gut microbiota based on the predominant groups.
Our study documents important observations regarding the composition of the gut microbiota of people living a subsistence lifestyle in PNG, in particular the predominance of Prevotella over Bacteroides and an insight into the core gut microbiota. Future studies are warranted using PCR independent detection methods that accurately characterize the entire gut microbiome as opposed to targeting selected dominant and sub-dominant microbial taxa; and ideally correlate gut microbiome composition to diet and activity. Recent studies have begun exploring the interactions of gut microbiota with nutritional status and infectious diseases in Malawi [15,22]: further work should address such issues in other low-income settings such as PNG where the burden of infectious disease is high.
S1 Fig. Box plot comparing the average number (log10/g faeces) of Bacteroides compared to Prevotella in study participants (n = 115).
S1 Table. Quantification of bacterial populations using reverse transcriptase real time PCR for all individuals included in analysis.
Conceived and designed the experiments: AG PH PS MU. Performed the experiments: HT KO KN AM KS KT SO JB YN ET KN. Analyzed the data: AG JAL MU. Contributed reagents/materials/analysis tools: HT KO MU AG PH KS PS JAL. Wrote the paper: AG PH MU.
- 1. Festi D, Schiumerini R, Birtolo C, Marzi L, Montrone L, et al. (2011) Gut microbiota and its pathophysiology in disease paradigms. Dig Dis 29: 518–524. pmid:22179206
- 2. Robles Alonso V, Guarner F (2013) Linking the gut microbiota to human health. Br J Nutr 109 Suppl 2: S21–26. pmid:23360877
- 3. Finegold SM, Molitoris D, Song Y, Liu C, Vaisanen ML, et al. (2002) Gastrointestinal microflora studies in late-onset autism. Clin Infect Dis 35: S6–S16. pmid:12173102
- 4. Ley RE, Turnbaugh PJ, Klein S, Gordon JI (2006) Microbial ecology: human gut microbes associated with obesity. Nature 444: 1022–1023. pmid:17183309
- 5. Ming X, Stein TP, Barnes V, Rhodes N, Guo L (2012) Metabolic perturbance in autism spectrum disorders: a metabolomics study. J Proteome Res 11: 5856–5862. pmid:23106572
- 6. Turnbaugh PJ, Ley RE, Mahowald MA, Magrini V, Mardis ER, et al. (2006) An obesity-associated gut microbiome with increased capacity for energy harvest. Nature 444: 1027–1031. pmid:17183312
- 7. Arumugam M, Raes J, Pelletier E, Le Paslier D, Yamada T, et al. (2011) Enterotypes of the human gut microbiome. Nature 473: 174–180. pmid:21508958
- 8. Wu GD, Chen J, Hoffmann C, Bittinger K, Chen YY, et al. (2011) Linking long-term dietary patterns with gut microbial enterotypes. Science 334: 105–108. pmid:21885731
- 9. De Filippo C, Cavalieri D, Di Paola M, Ramazzotti M, Poullet JB, et al. (2010) Impact of diet in shaping gut microbiota revealed by a comparative study in children from Europe and rural Africa. Proc Natl Acad Sci U S A 107: 14691–14696. pmid:20679230
- 10. Mueller S, Saunier K, Hanisch C, Norin E, Alm L, et al. (2006) Differences in fecal microbiota in different European study populations in relation to age, gender, and country: a cross-sectional study. Appl Environ Microbiol 72: 1027–1033. pmid:16461645
- 11. Li M, Wang B, Zhang M, Rantalainen M, Wang S, et al. (2008) Symbiotic gut microbes modulate human metabolic phenotypes. Proc Natl Acad Sci U S A 105: 2117–2122. pmid:18252821
- 12. Gupta SS, Mohammed MH, Ghosh TS, Kanungo S, Nair GB, et al. (2011) Metagenome of the gut of a malnourished child. Gut Pathog 3: 7. pmid:21599906
- 13. Lin A, Bik EM, Costello EK, Dethlefsen L, Haque R, et al. (2013) Distinct distal gut microbiome diversity and composition in healthy children from Bangladesh and the United States. PLoS One 8: e53838. pmid:23349750
- 14. Grzeskowiak L, Collado MC, Mangani C, Maleta K, Laitinen K, et al. (2012) Distinct gut microbiota in southeastern African and northern European infants. J Pediatr Gastroenterol Nutr 54: 812–816. pmid:22228076
- 15. Smith MI, Yatsunenko T, Manary MJ, Trehan I, Mkakosya R, et al. (2013) Gut microbiomes of Malawian twin pairs discordant for kwashiorkor. Science 339: 548–554. pmid:23363771
- 16. Yatsunenko T, Rey FE, Manary MJ, Trehan I, Dominguez-Bello MG, et al. (2012) Human gut microbiome viewed across age and geography. Nature 486: 222–227. pmid:22699611
- 17. Natsuhara K, Ohtsuka R (1999) Nutritional ecology of a modernizing rural community in Papua New Guinea: an assessment from urinalysis. Man and Culture in Oceania 15: 91–111.
- 18. Umezaki M, Yamauchi T, Ohtsuka R (1999) Diet among the Huli in Papua New Guinea Highlands when they were influenced by the extended rainy period. Ecol Food Nutr 37: 409–427.
- 19. Umezaki M, Ohtsuka R (1998) Impact of rural-urban migration on fertility: a population ecology analysis in the Kombio, Papua New Guinea. J Biosoc Sci 30: 411–422. pmid:9746838
- 20. Matsuda K, Tsuji H, Asahara T, Matsumoto K, Takada T, et al. (2009) Establishment of an analytical system for the human fecal microbiota, based on reverse transcription-quantitative PCR targeting of multicopy rRNA molecules. Appl Environ Microbiol 75: 1961–1969. pmid:19201979
- 21. Yamauchi T, Umezaki M, Ohtsuka R (2001) Influence of urbanisation on physical activity and dietary changes in Huli-speaking population: a comparative study of village dwellers and migrants in urban settlements. Br J Nutr 85: 65–73. pmid:11227035
- 22. Umezaki M, Ohtsuka R (2003) Adaptive strategies of Highlands-origin migrant settlers in Port Moresby, Papua New Guinea. Hum Ecol 31: 3–25.
- 23. Temu PI, Saweri W (2001) Nutrition in transition. In: Bourke RM, Allen MG, Salisbury JG, editors. Food Security for Papua New Guinea Proceedings of the Papua New Guinea Food and Nutrition 2000 Conference PNG University of Technology, Lae, 26–30 June 2000 ACIAR Proceedings No 99. Canberra: Australian Centre for International Agricultural Research. pp. 395–406.
- 24. Morita A, Natsuhara K, Tomitsuka E, Odani S, Baba J, et al. (2014) Development, validation, and use of a semi-quantitative food frequency questionnaire for assessing protein intake in Papua New Guinean Highlanders. Am J Hum Biol. https://doi.org/10.1002/ajhb.22647.
- 25. Backhed F, Fraser CM, Ringel Y, Sanders ME, Sartor RB, et al. (2012) Defining a healthy human gut microbiome: current concepts, future directions, and clinical applications. Cell Host Microbe 12: 611–622. pmid:23159051
- 26. Friedlaender JS, Friedlaender FR, Reed FA, Kidd KK, Kidd JR, et al. (2008) The genetic structure of Pacific Islanders. PLoS Genet 4: e19. pmid:18208337
- 27. Mendez FL, Watkins JC, Hammer MF (2012) Global genetic variation at OAS1 provides evidence of archaic admixture in Melanesian populations. Mol Biol Evol 29: 1513–1520. pmid:22319157
- 28. Serjeanston SW, Board PG, Bhatia KK (1992) Population genetics in Papua New Guinea: a perspective on human evolution. In: Attenborough RD, Alpers MP, editors. Human Biology in Papua New Guinea, The Small Cosmos. Oxford: Clarendon Press.
- 29. Agans R, Rigsbee L, Kenche H, Michail S, Khamis HJ, et al. (2011) Distal gut microbiota of adolescent children is different from that of adults. FEMS Microbiol Ecol 77: 404–412. pmid:21539582
- 30. Biagi E, Nylund L, Candela M, Ostan R, Bucci L, et al. (2010) Through ageing, and beyond: gut microbiota and inflammatory status in seniors and centenarians. PLoS One 5: e10667. pmid:20498852
- 31. Reuter G (2001) The Lactobacillus and Bifidobacterium microflora of the human intestine: composition and succession. Curr Issues Intest Microbiol 2: 43–53. pmid:11721280
- 32. Tannock GW, Munro K, Harmsen HJ, Welling GW, Smart J, et al. (2000) Analysis of the fecal microflora of human subjects consuming a probiotic product containing Lactobacillus rhamnosus DR20. Appl Environ Microbiol 66: 2578–2588. pmid:10831441
- 33. Xu P, Li M, Zhang J, Zhang T (2012) Correlation of intestinal microbiota with overweight and obesity in Kazakh school children. BMC Microbiol 12: 283. pmid:23190705
- 34. Murphy EF, Cotter PD, Healy S, Marques TM, O'Sullivan O, et al. (2010) Composition and energy harvesting capacity of the gut microbiota: relationship to diet, obesity and time in mouse models. Gut 59: 1635–1642. pmid:20926643
- 35. Venema K (2010) Role of gut microbiota in the control of energy and carbohydrate metabolism. Curr Opin Clin Nutr Metab Care 13: 432–438. pmid:20531179
- 36. Matsuki T, Watanabe K, Fujimoto J, Miyamoto Y, Takada T, et al. (2002) Development of 16S rRNA-gene-targeted group-specific primers for the detection and identification of predominant bacteria in human feces. Appl Environ Microbiol 68: 5445–5451. pmid:12406736
- 37. Matsuki T, Watanabe K, Fujimoto J, Takada T, Tanaka R (2004) Use of 16S rRNA gene-targeted group-specific primers for real-time PCR analysis of predominant bacteria in human feces. Appl Environ Microbiol 70: 7220–7228. pmid:15574920
- 38. Rajilic-Stojanovic M, Heilig HG, Molenaar D, Kajander K, Surakka A, et al. (2009) Development and application of the human intestinal tract chip, a phylogenetic microarray: analysis of universally conserved phylotypes in the abundant microbiota of young and elderly adults. Environ Microbiol 11: 1736–1751. pmid:19508560
- 39. Turnbaugh PJ, Hamady M, Yatsunenko T, Cantarel BL, Duncan A, et al. (2009) A core gut microbiome in obese and lean twins. Nature 457: 480–484. pmid:19043404
- 40. Tap J, Mondot S, Levenez F, Pelletier E, Caron C, et al. (2009) Towards the human intestinal microbiota phylogenetic core. Environ Microbiol 11: 2574–2584. pmid:19601958