Tuesday, August 24, 2010

Agenda for NuGOweek 2010

The following is the agenda for the nutrigenomics conference NuGOweek 2010. For more information on NuGO, see this link. I will try to provide updates and notes from the conference as long as wireless is functional...

Tuesday 31st of August 2010
Welcome and opening lectures
Welcome: Dr Baukje de Roos, University of Aberdeen, UK
In short introductory remarks, she noted that this is the first NuGOweek conference without FP6 funding. Thus, overall number of registrants is down from about 250 to about 130. The conference is funded in part by NuGO and Unilever.

Professor Naveed Sattar, Glasgow University, UK
Nutrigenomics - A perspective from the world of metabolic disease

NS was invited to kick things off and to provide the perspective of the physician who is seeing and treating patients with metabolic-based diseases such as type 2 diabetes (T2DM) and cardiovascular disease (CVD).

Amid rising rates worldwide for obesity, T2DM and CVD, the challenge becomes to slow obesity. Not only does obesity lead to increased risk of CVD and T2DM, but also to fatty liver, sleep apnea, some cancers and fertility and pregnancy complications. This said, CVD death rates are falling. Thus, the challenge is to slow the age-relted weight gain trajectory.

Using a biochemical marker (i.e., phenotype), we can screen for CVD risk quite well, but T2DM is complicated by an oft-changing marker. Perhaps that will be HbAc (acetylated hemoglobin). He uses a scoring system incorporating sex, age, BMI, blood pressure, family history of T2Dm and coronary heart disease, ethnicity, smoking status. Something similar can be found at www.qdscore.org.

If nutrigenomics research is to identify a new predictor of disease risk, that marker (or panel of markers) must be cost-effective because the above test is 80-85% accurate. One way of putting this is weight gain pulls the T2DM trigger. Perhaps slowly... So, which other marker might he wish to add to a T2DM test? One could be GGT.

CVD risk. Only when the increase in CVD risk exceeds 20% is the patient treated - with statins. However, most events occur when the risk is elevated by just 10-20% and these people are not treated. (LP - Is this where lifestyle intervention could help?) Thus, we need new phenotypes for risk factors. CRP associates with CVD but even after long, expensive studies it remains unclear if elevated CRP in the plasma enhances prediction. So, use genomics/proteomics - e.g., peptide patterns in urine of coronary artery stenosis.

T2DM risk. T2Dm trials are coming: 12 trials are now ongoing with from 5000 to 20000 subjects, 5-7 years duration. This is tough, costly, long.

Confounders. In Glasgow, vitamin D levels show a seasonal fluctuation, but also are linked to wealth - poorest subjects have lower levels than most affluent subjects.

A big chalenge is to link omics results to disease outcome / risk.

Conclusions. The real goal is to prevent obesity as this leads to other complications. Omics research, while in its infancy, shows promise. We need lots more data keeping in mind both the clinical questions and the translational potential of the results.

Wednesday 1st September 2010
Plenary Session 1: Nutrigenomics and novel biomarkers of health
Chairs: Professor Christian A. Drevon and Dr Lorraine Brennan

Professor Barbara Cannon, Stockholm University, Sweden
The adipostat hypothesis for body-weight control

Certain chemicals are mitochondrial uncouplers. One is DNP, dinitrophenol acting by proton leakage and she wonders if this could counteract obesity. Work done in chicken and quail showed that RQ is down and fat is burned when DNP is added to diet. In 1933, Cutting et al. showed that in humans metabolism increases and body weight decreases after DNP in the diet. Tainter, et al. (1933) showed weight loss of 0.5 - 1.0 kg/week, mostly around the hips/waist. Side effects were catarats/blindness, skin rash, loss of taste. This work, however, is proof that in humans thermogenesis works against obesity.

So, adipostat set point must be flexible.

Since 2007 it is clear that humans have brown adipose tissue (BAT). Questions: How many people have it? How much do they have? Does it matter? (She cannot answer this last question yet...)

Christiensen et al (2006) showed a temperature dependence to the ability to detect BAT in humans. See Zingaretti et al (2009) where BAT is densely assocaited with nerves. Human BAT contains UCP1, as in rodents. But does it matter? Look in rodents and Ucp1 -/- mouse. There is no thermogenesis in BAT when mice housed at thermoneutrality (30 oC). Similar phenotypes were observed in UCP1-/- mice on obesity-prone (C57) and obesity-resistant (129SV) backgrounds.

After norepinepherine treatment, respiration increases in normal vs Ucp1-/- mouse. There is no increase or difference in VO2 max in Ucp1-/- mice on chow vs high-fat diets. Basal metabolism is unchanged in these animals. Thus, there is an adaptive thermogenesis dependent on Ucp1.

Humans always go around with clothing - more or less in a thermoneutral state. OK, the above animals are obesity-prone

So, is more BAT good? Only if it is activated. Lower human BMI and age correlate with presence of BAT (Zingaretti et al (2009)). Hence, there is a diet-induced thermogenesis (DIT) and decreased DIT may cause obesity. Adaptive thermogenesis counteracts obesity.

Professor Helga Refsum, University of Oslo, Norway
Cysteine in relation to body composition

Cysteine gives rise to taurine and glutathione (an anti-oxidant). Cysteine is converted to glutathione by GGT (GGT1 and GGT2). But why look at cysteine? Change in BMI predicted change in total cysteine levels in plasma over time in a Swedish study. This change associated with fat mass and not lean mass. But are other sulfur amino acids involved (e.g., taurine, glutathione, methionine)? She showed that it is not the case, only cysteine.

Does high cysteine lead to obesity? Or does obesity lead to high cysteine? Or are there confounding factors? No: dietary factors and energy intake, physical activity, lipid related factors, serum glucose, GGT levels all show no confounding effects. Baritric surgery with rapid weight loss suggests that high cysteine levels lead to obesity.

CBS (gene) deficient humans are thin and CBS in excess in humans leads to overweight. Numerous genes are implicated: SCD-1, PLTP, ABCA1, et al.

Adding cysteine to rodents fed a methionine-restricted diet reverses the phenotypes. Fatty acid synthesis increased in diet supplemented with cysteine, as suggested by gene expression analysis.

Hannelore Daniel asked about cysteine oxidation - it is not impaired in the mouse experiments.

What about dieting (e.g., Atkins and high-protein). Diets fail. Soy is low in sulfur amino acids but associates with satiety. Need weight maintenance and not weight loss.

Dr Lorraine Brennan, University College Dublin, Ireland
Nutrityping and phenotyping people using metabolomics

She wishes to understand the interactions between lifestyle factors and nutrition-based phenotypes. She uses cluster analysis to find three dietary patterns in her group of about 160 Irish. She uses NMR to find differences in biomarkers of intake: fatty acids, O-acetylcarnitine in the urine and phenyl... in plasma. One of the latter two is a marker of red meat intake and the other of vegetable intake.

Phenotyping - an intervention study was conducted for 4 weeks with vitamin D. They found 5 clusters by k-means based on 14 biomarkers. One is 25(OH)D (vitamin D). So, which biomarkers respond? Cluster 5 responded by healthier profiles in adiponectin, HOMA, insulin. Metabolites altered in cluster 5 are VLDL/LDL (decreased), glucose (d), lactate (d) and glutamine (increased).

Using one biomarker is not sufficient and dividing a population based on a number (n>1) is necessary.

Professor Hannelore Daniel, Technical University Munich, Germany
The human metabolic accordion

The normal human metabolome is boring, right? Not really because of the time-component. [LP: she did not go into detail, but assumed one such t-c. I see several that I believe she would acknowledge: after a meal, post-exercise, throughout aging, etc.)

Up-front questions: Urine and plasma represent what? What is normal in the face of physical constitution, genetic heterogeneity, etc.? Is the static metabolome a good measure of health vs. disease?

Experiment: Young men, all within BMI of 23.7 +/- 1.7 (or so), were put through a battery of tests, beginning with a ~36-hr fast, glucose tolerance, exercise test, etc. etc, over the course of 4 days. During this time, blood was taken at many time points, urine, too. many metabolites were measured and many observations were made. For example, several amino acids change in remarkable ways during this treatment.


- Metabolic plasticity is important to evolution in order to rapidly respond in time/space (=organs, cells) to catabolic vs. anabolic states.

- Don't know what is "normal" when taking one snapshot after a overnight fast. Is this the best reference?

- Based on enormous plasticity of metabolic responses, it seems more advised to "titrate" the capacity of adaptation in time and space by defined and standardized changes for identifying deviations from normal.

Plenary Session 2: Modeling human metabolism
Chair: Professor Hannelore Daniel and Dr Grietje Holtrop

Dr Kevin Hall, National Institutes of Health, Bethesda, USA
Modeling Metabolism of Mice and Men

Modeling can be thought of as mathematics or of using a surrogate organism to learn about the human condition. He uses math. We can take longitudinal (i.e., long-term) measures of body weight, fat mass, lean mass, even food intake. However, getting long-term measures of energy expenditure is tricky. So, use short-term, then ask if mathematical modeling helps to get long-term values in numbers that easily, directly relate to values of food intake (kJ/kg body wt/day).

d(pBW)/dt = I - E,

where the change of body weight (with some factor rho) over time equals Intake - Expenditure. This is the energy balance equation.

Food intake and physical activity both allow mathematical modeling of human metabolism. This in turn allows calculation or determination of fluxes and changes of various sorts, e.g., metabolism of carbohydrates and lipids, energy expenditure, et al.

He used such to assess the USDA/ERS calculation that placing a tax on soda would lead to a linear weight loss over 5 years of about 10 kg for a 100 kg person. He found that this weight loss reaches a plateau and amounts to just 2 kg because the model, which uses more complicated mathematics than shown here, has 1) an exponent and thus reaches saturation; 2) a long time constant of 410 days.

Human weight change is dynamic and occurs over a long time scale. See their paper.

Professor Claudio Cobelli, University of Padova
Glucose Metabolism in Health and Diabetes: Necessity of Models

He takes the engineering approach - a simple experiment using complex mathematics to model it - as opposed to a biologist's complex experiment with a simple model. He uses the IVGTT - to measure glucose, insulin, C-peptide. A meal or OGTT is too complex because one needs to consider gut influences in order to model the observations.

He has moved to cellular models of insulin secretion. See the paper from 2008. Insulin sensitivity x beta-cell function = a constant. So, some people have low insulin sensitivity and need a boost with therapy, while others have reduced beta-cell response requiring a different therapeutic approach.
Dr Gerald Lobley & Dr Grietje Holtrop, University of Aberdeen, UK
Theoretical and practical considerations for measurement of glucose and protein kinetics

Moderated poster session 1

Scientific Session 1: Inflammation, metabolic health and obesity
Chairs: Professor Aldona Dembinska-Kiec and Professor John Mathers

Nadja Schulz, German Institute of Human Nutrition, Potsdam, Germany
Adp3, a protein involved in beta-oxidation is a putative regulator of insulin secretion

[LP: This is continuation of work I have seen from some 3 years ago, with reference to a gene that is not defined in literature nor in EntrezGene. Perhaps it is in patent applications.]

They began to work on this protein after a screen of C. elegans genes. Adp3-/- knock out mice show reduced body weight gain, but no differences in food intake. Some differences were noted in the light phase in locomotor activity. Increased body temperature in Adp3-/- in both phases was a key to the metabolism issue. These mice have impaired oral glucose tolerance tests but the insulin response and fat tolerance are like wildtype.

- Decreased insulin secretion in response to glucose in the KOs

- Increased insulin secretion in response to fat in the KOs.

Hannah R. Elliott, Newcastle University, UK
Novel epigenetic biomarkers of T2D susceptibility

Three questions:

- Do DNA methylation patterns associate with T2DM traits?

- Do such methylation patterns alter with age?

- Does #2 above (altered patterns) associate with T2DM severity?

She looks at the first question using the RISC cohort and CpG islands in the promoter and exon-1 regions of FTO and ADCY5. Specifically, she is most interested in CpG islands in the promoter and transcription factor binding sites. They use a MALDI-TOF approach to get a percentage of differential mass, which is an indicator of methylation.

- BMI and age correlated positively with ACY5 methylation.

- No correlation was observed between FTO methylation and age.

Thomas Skurk, Technische Universität München, Germany
Cell size of human adipocytes affects endocrine and metabolic functions

Fat cell size in adipose tissue. Adipocytes increase in size as BMI increases. He size-fractionated adipocytes. There is a shift to pro-inflammation mode in larger fat cells, assessed by measures of cytokines. It looks like ER-stress is not the only relevant factor but he is looking at more genes. Small adipocytes are insulin sensitive; large cells appear insulin resistant, but this is really true only when the person is a type 2 diabetic.

James C. McConnell, Newcastle University, UK
Genome wide DNA methylation is associated with lipid profiles at age 50

They used the Newcastle Thousand Families Study, a longitudinal birth cohort from 1947. Global DNA methylation was assessed by pyrosequencing in 231 individuals at 3 CpG islands in LINE-1 retrotransposon elements. Significant positive correlations were observed between methyl-DNA and levels of fasting glucose and C-peptide. Also, blood lipids of total cholesterol, LDL-cholesterol (increased), APOB, triglycerides (increased) and HDL-cholesterol (decreased). Thus, a perturbed pattern of DNA methylation is suggested in pathogenesis of common complex diseases.

Miguel A. Lucena, IMABIS Foundation, Malaga, Spain
Metabolic alterations in the abdominal muscle of obese rats - a proteomic approach

In obese rats, muscle saw decreased levels of glycolysis-related enzymes: glucose-6-phosphate isomerase, alpha-enolase and lactate dehydrogenase. Increased levels of FABP3 and FABP4 were noted as well as B-crystallin and HP (haptoglobin). It looks like glucose and fatty acid metabolism are affected by obesity in skeletal muscle.

Andreas Kolb, University of Aberdeen, UK
B-vitamin deficiency and phenotypic variation in vascular cells

They used A7r5 cells. Treatment was high folate, 100 ng/ml. This induced expression of many cholesterol and lipid metabolism genes - more so than any other pathways or funcitonal group. However, some genes were up-regulated and some were down-regulated. (It wasn't entirely clear to me, but I believe that these genes function in both synthesis and metabolism.) B-vitamin deficiency increased expression of pro-inflammation cytokines and decreased NO production.

Scientific Session 2: Novel food models to investigate metabolic health
Chairs: Dr Suzan Woperies and Professor Edwin Mariman

Suzan Wopereis, TNO Quality of Life, Zeist, the Netherlands
Postprandial challenge test to demonstrate subtle dietary effects on human health

MPO and MDC show less increase after the high-fat challenge (these are AIDM genes). VCAM1 showed greater reduction. ACE was reduced compared to the placebo at baseline.

Claire Merrifield, Imperial College, London, United Kingdom
NMR-based urinary metabolic profiling of the pig reveals a sustainable metabolic reprogramming event related to weaning diet

Laurence D. Parnell, Tufts University, Boston, MA, United States
Network analysis defines the impact of gene-physical activity interactions

Mark Boekschoten, Netherlands Nutrigenomics Centre, Wageningen, Netherlands
Effect of dietary fat on the transcriptome in white adipose tissue of C57BL/6J mice

Thursday 2nd September 2010
Plenary Session 3: Inflammation, metabolic health and obesity
Chairs: Professor Michael Muller and Professor Harry McArdle

Professor Michael Muller, Wageningen University, the Netherands
Metabolism and Inflammation

NAFLD = non-alcoholic fatty liver diseases, is a component/manifestation of metabolic syndrome where PPARA plays a role, especially in Kupffer cells.

NASH = non-alcoholic steatosis hepatitis.

Their goal is to isolate biomarkers of NASH. BLACK 6 mice develop NASH on a high-fat diet (45% fat vs 10% fat for control). Many genes show altered expression in the high-fat/high-responder group. This is about twice the number of genes as in the high-fat/low-responder and low-fat/high-responder groups. Many genes fall into three categories: fibrosis, inflammation, lipid metabolism.

Furthermore, changes in gene expression indicate adipose dysfunction. This is emphasized by macrophage infiltration.

The search for a plasma biomarker: CRP, haptoglobin, IL1B, MIP-1alpha - early markers of NASH development.

The Angptl4-/- mouse on a high-fat diet is very ill. But adipose tissue and liver are small. They detect systemic inflammation. Saa2, haptoglobin and this is independent of microbiota. This is observed only when the fat source is lard or palm oil, not with safflower oil.

Angptl4, under control of PPARD, represses LPL (lipoprotein lipase). In the Angptl4-/- KO, triglycerides in the chylomicrons go to fatty acids. This is in press in Cell Metabolism.

Dr Matthijs Hesselink, University of Maastricht, the Netherlands
Muscle physiology in insulin resistance and type 2 diabetes

Fatty acid derivatives disrupt IRS-PI3K-SLC2A4 signaling but evidence for such is lacking in T2DM subjects. Skeletal muscle is responsible for about 40% of postprandial glucose uptake. The focus is on storage of fat in muscle (ectopic fat). The balance between fat storage and fat metabolism in muscle is indicative of cell function. The literature shows that more fat there is in muscle, the more insulin resistance there is. Muscle triglyceride (TG) storage is augmented by increases in free fatty acids and the TG levels decrease after exercise, but there is a differential effect on insulin sensitivity.

Lipid droplet (LD) proteins (also known as perilipins or PAT proteins) are indeed important in muscle: PLIN 5 (OXPAT), PLIN2 (adipophilin, ADRP), PLIN3 (Tip47). PLIN4 (S3-12) is also expressed in muscle. Expression of PAT genes in muscle of T2DM subjects vs those without T2DM: control for age, BMI: (see Meex 2010 Diabetes). PLIN2 and PLIN5 showed no differences in expression levels, but PLIN3 is down-regulated in T2DM skeletal muscle. In this case, new LDs are not made. Gene PNPLA2 (ATGL) shows no difference. Now add exercise training. Of those genes reported, only PLIN2 and PLIN5 are up-regulated in both muscle types (T2DM and non-T2DM) post-exercise. PNPLA2 is up-regulated only in T2DM post-exercise.

The adaptive response of PLIN5 and PLIN2 may improve fuel selection or use in hyperinsulinemic subjects.

Dr Lydia Afman, Wageningen University, the Netherlands
The challenge of nutritional phenotyping in human nutrigenomics

Her goal is to identify early biomarkers of disease at a time when nutrition can be used to treat the pre-disease state.

1) Gene expression in PBMCs comparing MUFA-based diet vs. diet high in EPA/DHA. This was done for long-term (20-26 weeks). The main finding: a diet high in EPA/DHA elicited an anti-inflammation anti-atherogenic gene expression profile.

2) Adipose gene expression after 8 weeks on one of three diets: Mediterranean, high MUFA (20%), high SFA (20%). She presented only on the MUFA:SFA comparison. There was no difference in insulin sensitivity; no effect on HOMA was observed. Both the MUFA and saturated fat diets were about 40% in fat, with 20% of energy coming from the respective fat type. SFA increased expression of many inflammation pathways, notably T- and B-cell receptor signaling, leukocyte extravasation and complement. The SFA diet induced a pro-inflammatory, obesity-linked gene profile. MUFA showed a reduced inflammatory profile.

Professor Christian A. Drevon, University of Oslo, Norway
New myokines and potential actions

He began with a list reviewing literature of positive, beneficial effects of exercise on a number of diseases. He also mentioned the review of BK Pedersen (Physiol Rev 2008) describing contraction-induced release of IL6 leading to increased glucose uptake (via PI3-K) and increased fat oxiation (via STAT3).

IL7 is secreted from skeletal muscle cells. IL7 mRNA increases linearly with myogenic differentiation. LPS increased IL7 mRNA but not protein levels.

IL7 is localized to myotubes expressing myosin heavy chain. Like myostatin, IL7 decreases expression by about 35% of myosin heavy chain (MYH2) and MYOG. IL7 enhanced myotube migration.

In human subjects undergoing strength training for 2 and 11 weeks, increased expression of IL7 was noted in skeletal muscle. Also increased were IL8, TLR1, TLR2, TLR3, TLR4, TL5, TLR6, TLR7; not TLR9. See their paper.

Plenary Session 4: Gut metabolism and chronic disease development
Chairs: Professor Harry Flint and Dr Elizabeth Lund

Professor Michael Blaut, DIFE, Potsdam, Germany
Impact of food ingredients on intestinal microbiota-associated obesity development in mice

Dr Patrice Cani, University Catholique de Louvain, Belgium
The contribution of gut micro-organisms in promoting and preventing insulin resistance

Dr Petra Louis, University of Aberdeen, UK
Impact of diet upon the human gut microbiota and gut metabolism in obese subjects

Moderated poster session 2

Scientific Session 3: Food, nutrigenomics, biomarkers and health
Chairs: Dr Jill McKay and Professor Sean Strain

Jill McKay, Newcastle University, Newcastle upon Tyne, United Kingdom
Folate depletion during development and high fat intake from weaning: consequences for DNA methylation and gene regulation

Maryam Rakhshandehroo, Nutrigenomics Consortium, Wageningen, the Netherlands
Mannose binding lectin is a circulating mediator of hepatic PPARα activity in human

The aim was to screen for novel circulating mediators of PPARA activity in human. They identified MBL2 as a circulating mediator of PPARA likely affecting innate immunity.

Jildau Bouwman, TNO Quality of Life, Zeist, the Netherlands
Let's visualize personalized health

Luisa M. Ostertag, University of Aberdeen, UK
Dark secrets of chocolate, platelet function and cardiovascular health

Emilie Martinez, INRA, Clermont-Ferrand, Auvergne, France
Changes in the myocardium proteome of rat pups after maternal deficiency of methyl donors

Siv Kjølsrud Bøhn, University of Oslo, Oslo, Norway
Bilberry and grape juice decreases plasma biomarkers of inflammation in aged men with subjective memory impairment

Results: Compared to placebo, plasma biomarkers of inflammation (EGF, VEGF, IL6, MIP1b, IL10, IL9 and TNF) and a biomarker of tissue damage (LDH) significantly decreased after bilberry/grape consumption while several plasma polyphenols increased.

Debate: The future of personalised nutrition
Moderator: Dr Ben van Ommen

Argument 1: Personalized nutrition is alive and kicking
Personal health monitoring will be daily practice. Everyone has smartphone and internet access to his health status, based on electronic health records, a series of frequent bioassays in the home setting, genomics information, coupled to life style and dietary advice. Industry has skipped the concept of functional foods, and provides tailored foods in the context of life style coaching, integrated with personal health monitoring. Nutrition science has finally understood how to deal with genetic variation, that is, of course not by further refining epidemiology but by exploiting systems biology modeling. Also, major breakthroughs in mechanistic nutrition research embedded in the biology revolution provided a wealth of knowledge on food bioactives. Healthy ageing is a reality!
Speakers: George Lietz, Barbara Stewart-Knox & Christian Drevon

Argument 2: Personalized nutrition is dead, long live nutrition
Although mechanistic nutrition has provided a lot of new views on modes of actions, this appeared to have no real impact whatsoever on actual health, except for some fine-tuning. The obesity outbreak made diet the 'silent killer,' which made nutrition research split into two mainstream lines, driven by health care costs. One side merged with biomedical research focusing on prevention of pathologies. The other side merged with social science to focus on 'social engineering' of food intake control. Food intake quantification has improved and epidemiology readily incorporated this, to finally optimize public health dietary recommendations.
Speakers: Piero Dolara & Anne-Marie Minihane

Argument 3 – Nutrigenomics is a waste of money
While a lot of money has been burned on high-tech nutrition research, the marginal advances in health optimization did not justify further spending. In fact, this money could have been more wisely spent on international nutrition, as more than half of the global population still receives an inadequate diet. Anyhow, a series of events caused the decay of nutrition research. EFSA regulations in the end depressed food industry, which stopped submitting claims but rather returned to consumer persuasion via commercials. This was encouraged by the media coverage of many conflicting messages from the nutrition research community. In the end, funding for nutrition research diminished and mainstream biological research absorbed diet as one of the 'environmental factors.'
Speakers: Hannelore Daniel, Helen Roche & Duccio Cavalieri

Friday 3rd September 2010
Plenary Session 5: Insulin resistance and the brain
Chairs: Dr Ben van Ommen and Dr Lynda Williams

Dr Kenneth Kornman, Interleukin Genetics
Genetic patterns predict weight loss success at 12 months: The right diet does matter

KK: You have to reduce calories to lose weight, but how much you lose is genetically determined.

They looked at three gene variants: rs1799883 in FABP2, rs 1801282 in PPARG, rs1042714 in ADRB2 because these had substantial data from the literature and are functional (i.e., amino acid change. See Gardner, et al 2007).

Individuals (all females and overweight to obese, n=~140) were randomly assigned to one of four diets for a period of 12 months. For the first two months, they came into the clinic once a week. After that, they were contacted by phone to assess eating behaviors and status with respect to the diet. Diet types were either low-fat, low-carbohydrate or neither. Diets included Atkins, the Zone and Ornish. Genotyping of individuals was done after completion of the study!

The hypothesis is, of course, there is a diet to match the genotype of the individual. Diets for the appropriate genotype lost 2- to 3-fold more weight (closer really to 2-fold) at 6 and 12 months after initiation of the study than those on the inappropriate diet. Weight loss at 6 months was about 5.5 kg on the appropriate diet and about 4.5 kg at 12 months. Waist, triglycerides also dropped; HDL-cholesterol went up. Weight loss was steeper for both groups (on appropriate and on inappropriate diets) at 2 months and this is likely due to the weekly clinic visits.

The data at two months shows something on satiety. Subjects on the appropriate diets took in ~100 cal less (but this was not explained further in response to my question).

Professor Oren Froy, The Hebrew University of Jerusalem, Israel
Metabolism and Circadian Rhythms--Implications for Obesity

See paper by Froy in Clin. Sci. (2010) on core clock components and metabolism factors. While that must be in press, one can view this paper. There is a master clock in the brain and peripheral clocks in many organs/tissues. Only the master clock appears to sensitive to feeding.

Mice were put on restricted feeding for four months. Restricted feeding is allowing the animals to eat as much as they want but only during the 3-5 hours that food is available during each 24-hr period. They found that this feeding regimen attenuates the peripheral clock and lowers inflammation markers. See this paper for details.

Restricted feeding stimulated the food entrained oscillator, leading to high amplitude circadian rhythms and reduced levels of inflammation markers.

A high-fat diet disrupts and flattens the circadian rhythms (Barnea, et al. 2009). [LP: I asked if it is intake of calories or even water that can trigger these observations. Response: It must be calories as water had no effect.]

Dr Lynda Williams, University of Aberdeen, UK
Novel biomarkers of inflammation and leptin sensitivity

Does early onset sensitivity to leptin really matter? A high-fat diet compromises leptin action in the hypothalamus and not via the JAK/STAT signaling pathway.

Leptin is a potent insulin sensitizer acting on the hypothalamus and is necessary for the full response to glucose and glucose homeostasis. This is not due to caloric intake, but to high-fat diets.

Dr Ineke Klopping, TNO, the Netherlands
HPA linking metabolism, brain and psychological stress

Her two main points were nutrigenomics research needs to consider the stress level of the individual and timing of sampling (due to seasonal or circadian fluctuations).

Scientific Session 4: Gut metabolism and chronic disease development
Chairs: Professor Piero Dolara and Dr Robert Kleemann

Piero Dolara, University of Florence, Italy
Sodium butyrate enemas modify gene expression, atrophy and inflammation in mucosal enterostomy pouches

Lisa Gruber, Technische Universität München, Freising, Germany
The effect of high-fat feeding in a mouse model of inflammatory bowel disease

Didier Attaix, INRA/Clermont Université, Clermont-Ferrand, France
GLP-2 inhibits intestinal lysosomal proteolysis and improves skeletal muscle recovery in the starved/refed rat

Ben van Ommen, TNO Quality of Life, Zeist, the Netherlands
The nutritional phenotype database in practice

Wednesday, August 11, 2010

Tuberculosis susceptibility SNP under selective pressure

A recent publication by Thye, Vannberg, Horstmann, Hill, et al. reports on a genome-wide association study in Africans on association to the susceptibility of tuberculosis. The major finding is the identification of a susceptibility locus in a gene-poor region on chromosome 18q11.2 centered on SNP rs4331426.

My colleague Chao-Qiang Lai used Jonathan Pritchard's Haplotter data to discern that a comparison of YRI to CEU with Fst shows that this SNP, centered within a window of 151 SNPs, is within a group of variants (n=151) where 98% of the SNPs have Fst values greater than 95% of all SNPs. This is an indication that rs4331426 is under selection. With G as the minor allele and a minor allele frequency (MAF) in CEU of 0.025 and a MAF in YRI of 0.508, this difference in allele frequencies is noteworthy. In two Asian populations, MAFs are 0.03-0.04. Not surprisingly, we do not detect significant Fst between CEU and either of these two populations.

This is an important example, of which there is a growing body, in which genetic variation under selective pressure and disease phenotype(s) are linked. As diet constitutes a key aspect of the environment, we feel that there will be interesting findings at the intersection of selective pressure and metabolic-based disease.

Monday, August 9, 2010

More on the microRNAs that control SIRT1 expression

In the journal Aging, a report has just been published describing the control of expression of SIRT1 by microRNAs. See Lee and Kemper (2010).

Three microRNAs implicated in their study are MIRN34A, MIRN132 and MIRN199A. To add to the conversation, I provide some other details of these microRNAs:

MIRN34A - expression in subcutaneous fat is correlated positively with BMI (Ortega, et al. 2010 PLoS ONE 5: e9022); the pre-adipocyte from an obese individual has 1.23-fold higher expression than from a lean person (Ortega, et al.); furthermore, the gene is upregulated during adipocyte differentiation (Ortega, et al.); lastly, confluent fibroblasts show 4.1-fold higher expression than sub-confluent cells (Hwang, et al. 2009 PNAS 106: 7016-7021)

MIRN132 - no other data to offer

MIRN199A - similarly, expression in subcutaneous fat is correlated positively with BMI (Ortega, et al. 2010 PLoS ONE 5: e9022); there is 2.3 to 3-fold higher expression in confluent fibroblasts than in sub-confluent cells (Hwang, et al.)

The two genes above, those with data, are certainly interesting and strengthen the metabolic links involving SIRT1.