Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Dysregulation of lipid and amino acid metabolism precedes islet autoimmunity in children who later progress to type 1 diabetes

Dysregulation of lipid and amino acid metabolism precedes islet autoimmunity in children who... ARTICLE Dysregulation of lipid and amino acid metabolism precedes islet autoimmunity in children who later progress to type 1 diabetes 1 2 1 2 Matej Ore š i c ˇ , Satu Simell , Marko Sysi-Aho, Kirsti N ä nt ö -Salonen , 1 2 1 Tuulikki Sepp ä nen-Laakso , Vilhelmiina Parikka , Mikko Katajamaa , 4 1 5 1 Anne Hekkala , Ismo Mattila , P ä ivi Keskinen , Laxman Yetukuri , 6 5 1 2 Arja Reinikainen , Jyrki L ä hde , Tapani Suortti , Jari Hakalax , 2 7,8 4 3,9 Tuula Simell , Heikki Hy ö ty , Riitta Veijola , Jorma Ilonen , 6 5,10 2 Riitta Lahesmaa , Mikael Knip , and Olli Simell VTT Technical Research Centre of Finland, Espoo FI-02044, Finland 2 3 Department of Pediatrics and Immunogenetics Laboratory, University of Turku, Turku FI-20520, Finland Department of Pediatrics, University of Oulu, Oulu FI-90014, Finland Department of Pediatrics, Tampere University Hospital, Tampere FI-33521, Finland Turku Centre for Biotechnology, Turku FI-20521, Finland Department of Virology, University of Tampere, Tampere FI-33520, Finland Centre for Laboratory Medicine, University Hospital of Tampere, Tampere FI-33520, Finland Department of Clinical Microbiology, University of Kuopio, FI-70211 Kuopio, Finland Hospital for Children and Adolescents, University of Helsinki, Helsinki FI-00014, Finland The risk determinants of type 1 diabetes, initiators of autoimmune response, mechanisms regulating progress toward  cell failure, and factors determining time of presentation of clinical diabetes are poorly understood. We investigated changes in the serum metabolome prospectively in children who later progressed to type 1 diabetes. Serum metabolite profi les were compared between sample series drawn from 56 children who progressed to type 1 diabetes and 73 controls who remained nondiabetic and permanently autoantibody nega- tive. Individuals who developed diabetes had reduced serum levels of succinic acid and phosphatidylcholine (PC) at birth, reduced levels of triglycerides and antioxidant ether phospholipids throughout the follow up, and increased levels of proinfl ammatory lysoPCs several months before seroconversion to autoantibody positivity. The lipid changes were not attributable to HLA-associated genetic risk. The appearance of insulin and glutamic acid CORRESPONDENCE decarboxylase autoantibodies was preceded by diminished ketoleucine and elevated glu- Matej Ore š ic ˇ : tamic acid. The metabolic profi le was partially normalized after the seroconversion. Auto- [email protected] OR immunity may thus be a relatively late response to the early metabolic disturbances. Olli Simell: Recognition of these preautoimmune alterations may aid in studies of disease pathogenesis [email protected] and may open a time window for novel type 1 diabetes prevention strategies. Abbreviations used: BCAA, branched chain amino acids; DIPP, Type 1 Diabetes predic- The incidence of type 1 diabetes among chil- a linear steady increase from 1965 to 1996 ( 3 ), tion and prevention study; dren and adolescents has increased markedly in was recently found to be increasing even faster GABA,  -aminobutyric acid; the Western countries during the recent de- than before, with the number of cases diag- GADA, glutamic acid decarbox- ylase antibody; GCxGC-TOF/ cades ( 1 ). The incidence has reached record nosed at or before 14 yr of age expected to MS, two-dimensional gas chro- levels in Finland, where, currently, 1/120 chil- double in the next 15 yr ( 4 ). Although  70% matography coupled to time-of- dren develops type 1 diabetes before the age of of subjects with type 1 diabetes carry defi ned fl ight mass spectrometry; IAA, insulin autoantibody; ICA, islet 15 yr ( 2 ). The annual incidence, which showed cell autoantibody; PC, phospha- © 2008 Orešicˇ et al. This article is distributed under the terms of an Attribu- tion–Noncommercial–Share Alike–No Mirror Sites license for the fi rst six months tidylcholine; STRIP, Special after the publication date (see http://www.jem.org/misc/terms.shtml). After six Turku Coronary Risk Factor S. Simell, M. Sysi-Aho, K. N ä nt ö -Salonen, T. Sepp ä nen- months it is available under a Creative Commons License (Attribution–Noncom- Intervention Project for Laakso, V. Parikka, and M. Katajamaa contributed equally to mercial–Share Alike 3.0 Unported license, as described at http://creativecommons Children. this paper. .org/licenses/by-nc-sa/3.0/). The Rockefeller University Press $30.00 J. Exp. Med. Vol. 205 No. 13 2975-2984 2975 www.jem.org/cgi/doi/10.1084/jem.20081800 The Journal of Experimental Medicine risk-associated genotypes at the HLA locus, only 3 – 7% of the To compare the fi ndings in the genetically defi ned DIPP carriers of such genetic risk markers develop the disease ( 5 ). cohort to those in another cohort of children that was not Seroconversion to islet autoantibody positivity has been genetically defi ned, the metabolomes of another six children the fi rst detectable signal for the onset of autoimmunity and in the Special Turku Coronary Risk Factor Intervention progression toward diabetes. Although seroconversion usually Project for Children (STRIP) ( 12 ) who had developed type 1 precedes the clinical disease by months to years, its occur- diabetes were investigated before disease presentation and rence may already be too late for therapeutic approaches compared with their six age- and sex-matched nondiabetic aimed at preventing progression to overt diabetes. As long as controls from the same study. The STRIP series also com- the initiators of the autoimmune response remain unknown prised 87 samples, which had been collected annually be- and the mechanisms supporting progression toward  cell tween the ages of 7 mo and 5 – 15 yr. Three of the children failure are poorly understood, the estimation of the absolute who developed diabetes turned out to carry the defi ned disease risk and time of disease presentation in genetically HLA-DQB1 risk alleles, but all had had multiple diabetes- susceptible individuals and the discovery of eff ective preven- associated autoanti bodies for 2 – 13 yr before the presentation tion remain a challenge. In 1994, an ongoing birth cohort study (Type 1 Diabetes prediction and prevention study [DIPP]) was launched in Finland ( 6 ). After informed parental consent, HLA alleles associ- ated with type 1 diabetes risk or protection were analyzed from cord blood, and subjects carrying risk-associated genotypes were followed over time to establish the time point of sero- conversion and diabetes development. Over a period of 11.5 yr until 2006, > 100,000 newborn infants have been screened for genetic risk and > 8,000 at-risk children are being followed. Metabolomics systematically studies the chemical fi nger- prints in cells, tissues, and biofl uids in a given physiological and environmental context. New analytical methods combined with bioinformatics provide a way to measure the extended metabolome ( 7 ). The metabolic phenotype is sensitive to subtle but pathogenically relevant factors such as age, lifestyle, nutrition, and the microbe environment of the gut ( 8 – 10 ). Changes in the concentrations of metabolites during early development may thus refl ect both genetic and environmental factors infl uencing later susceptibility to chronic diseases ( 11 ). In this paper, we hypothesize that alterations in the meta- bolic phenotypes characterize the early pathogenesis of type 1 diabetes. We apply the metabolomics strategy to the unique preautoimmune and prediabetic sample series collected in DIPP and provide evidence that metabolic disturbances pre- cede the autoimmunity that is characteristically observed be- fore development of type 1 diabetes. RESULTS Subjects Serum metabolite profi les were compared between 50 DIPP children who progressed to type 1 diabetes (the progressors) and 67 nonprogressors, who remained healthy and autoanti- body negative and were matched with the progressors for time and site of birth, gender, and HLA risk group ( Fig. 1 ). The 117 DIPP children contributed 1,109 samples (9.5 sam- ples per child on average) covering the time from 3 mo of age to clinical diabetes in the progressors. Additionally, the cord blood samples of 39 DIPP children from Turku, 15 of which Figure 1. Subject selection fl owchart for the metabolomics study. progressed to diabetes before the age of 12 yr, were analyzed. Nonprogressors were matched with progressors for time and site of birth, Two progressors included in the cord blood series were also gender, and HLA risk group. Status of DIPP ( 6 ) and STRIP ( 12 ) studies as of included in the longitudinal sample series. June 6, 2006. 2976 METABOLOME EN ROUTE TO TYPE 1 DIABETES | Ore š ic ˇ et al. ARTICLE of diabetes. Selected characteristics of the study subjects are phosphatidylcholine (PC), as well as high lysoPC and leucine shown in T able I . ( Fig. 2 B ). This period was also accompanied by abnormally high glutamic acid and  -aminobutyric acid (GABA), as well Analytical platforms as low glutamine and  -ketoglutarate ( Fig. 2 C ). During glu- Current lipidomics platforms constitute an effi cient way to tamic acid decarboxylase antibody (GADA) positivity, these separate and quantify hundreds of lipid molecules ( 13, 14 ). metabolite concentrations almost normalized, whereas ala- For this study, ultra performance liquid chromatography nine diminished and 2-hydroxybutyric acid increased tran- coupled to mass spectrometry was applied as previously des- siently (unpublished data). cribed ( 8, 15 ). The fi nal lipidomics dataset, not including the cord blood data, comprised 53 identifi ed lipids mea- Metabolome in prospective sample series and cord blood sured in each of the 1,196 samples. A total of 515 samples The analysis of longitudinal series of serum metabolomes were from progressors to type 1 diabetes, among which during the asymptomatic period from children who later de- 112 were collected before seroconversion to autoantibody veloped diabetes revealed characteristic metabolite patterns positivity. In each of the 39 cord blood samples, 249 lipids that preceded or accompanied the initial seroconversion to were identifi ed. Once the complete dataset was constructed, autoantibody positivity ( Fig. 3 ). Compared with children we fi rst confi rmed that the samples were not aff ected by who remained autoantibody negative, the serum lipidome of the time of storage (Fig. S1, available at http://www.jem progressors showed a consistent decrease in triglycerides (P = .org/cgi/content/full/jem.20081800/DC1). 0.005) and multiple phospholipids, including ether PCs (P < Serum samples were also analyzed by two-dimensional 0.001), PCs (P = 0.04), and sphingomyelins (P = 0.09), in gas chromatography coupled to time of fl ight mass spectrom- samples obtained from infancy to early school years ( Fig. 3, etry (GCxGC-TOF/MS). This analytical technology pro- A and B ). The diff erence in lipidome was observed before the vides a broad profi le of small molecules ( 16 ). The analysis was autoantibodies appeared and it persisted throughout the en- performed on a random subset of DIPP study samples from tire age range covered. Low phospholipid levels were also Turku, comprising sample series from 13 progressors and 26 observed in cord blood, although only one of the lipids de- matched nonprogressors. The fi nal metabolomics dataset tected in the longitudinal sample series diff ered signifi cantly obtained by GCxGC-TOF/MS consisted of 75 identifi ed between the progressors and nonprogressors. However, the metabolites measured in each of the 419 samples. total amount of cord serum PC was 1.22-fold decreased in the progressors (P = 0.004), as shown in Fig. S2 (available at Case report http://www.jem.org/cgi/content/full/jem.20081800/DC1). The interdependence of metabolic and immune system fac- We also compared the lipid concentration diff erences be- tors is fi rst demonstrated in Fig. 2 for a girl with an extended tween the children at high and moderate HLA-associated gene- sample series available. She developed overt type 1 diabetes tic risk of developing type 1 diabetes. No signifi cant diff erences close to 9 yr of age ( Fig. 2 A ). The time before the appear- between the genetic risk groups were identifi ed among the ance of the fi rst autoantibody was characterized by abnor- progressors or the nonprogressors (Fig. S3, available at http:// mally low succinic acid (not depicted), ketoleucine, and ether www.jem.org/cgi/content/full/jem.20081800/DC1), thus Table I. Selected characteristics of the subjects included in the study. Progressors Nonprogressors Gender (girls, boys) 31, 25 40, 33 City of birth (Turku, Oulu) 29, 27 45, 28 Season at time of birth 11, 19, 14, 12 15, 25, 19, 14 (winter, spring, summer, autumn) Season at time of diagnosis 11, 21, 10, 13 NA (winter, spring, summer, autumn) Age at time of diagnosis 44, 6 – 162 NA (median, range in months) Age at time of fi rst seroconversion 12, 6 – 96 NA (median, range in months) HLA risk haplotypes Moderate risk DR4-DQ8/x 33 46 (x = any haplotype except DR2-DQ6, DR5-DQ7 or DR3-DQ2) High risk (DR3-DQ2/DR4-DQ8) 17 20 Seasons are defi ned as winter for December to February, spring for March to May, summer for June to August, and autumn for September to November. HLA genotype information is shown for the DIPP children only. One nonprogressor, not included in the table, had the moderate risk DR3-DQ2/x haplotype. JEM VOL. 205, December 22, 2008 2977 confi rming that the observed lipid changes are not attribut- able to HLA-associated genetic risk. One subclass of phospholipids that was repeatedly low in progressors and showed low within-person variability were ether PCs ( Fig. 3, A and B ). Ethanolamine plasmalogen levels Figure 3. Serum lipidome in cord blood and prospective sample series. (A) Differences in serum lipidome between the progressors to type 1 diabetes and nonprogressors who remained autoantibody negative throughout the follow up. The age groups are divided into birth (cord Figure 2. Selected autoantibody and metabolite changes during blood) and then into groups covering 1-yr cohorts. Only one sample per the prediabetic period in a girl who progressed to type 1 diabetes at subject, closest to the mean age within the time window, is used in each close to 9 yr of age (HLA-haplotype DR7-DQ2/DR4-DQ8, moderate comparison. The number of subjects included in each age cohort is shown genetic risk). (A) GADA and IAA profi les. The relative autoantibody level at the bottom. The last sample of each progressor was the sample drawn values are calculated as the ratio of their measured levels and corre- last before diagnosis of diabetes. (B) Total triglyceride and ether PC con- sponding cutoff limits for autoantibody positivity. Scales are linear and centrations for the age cohorts shown in A calculated as the sum of lipid adjusted separately for each autoantibody for clarity. Time of seroconver- concentrations within each class. Both lipid classes were found consis- sion to positivity for each autoantibody is marked, as is the time of diag- tently down-regulated in the progressors, as tested by the linear mixed nosis of type 1 diabetes. The child also seroconverted to ICA and IA-2A effects model for the overall trend throughout the follow up. The error autoantibody positivity within 6 mo before the appearance of IAA. bars show SEM. (C) Levels of the lysoPC PC(18:0/0:0) and the ether PC (B) Changes of ketoleucine, leucine, lysoPC (the profi le shown is for the most PC(O-18:1/20:4) at the age of 18 mo for DIPP children. The two lipids abundant lysoPC species PC(18:0/0:0)), and ether PC (the profi le shown is were up-regulated (P = 0.0009) and down-regulated (P = 0.04) in pro- for the most abundant ether PC species PC(O-18:1/20:4)) with age. Fold gressors, respectively. Only one measurement, closest to the age of 18 mo changes are calculated as ratios of metabolite level at a given time point within the 12 – 24-mo age interval is shown for each subject. Subjects and the normal level, i.e., the mean value across all samples in the non- who already seroconverted to autoantibody positivity against one or mul- progressors. Scales are linear and adjusted separately for each metabolite tiple islet autoantigens are specifi cally marked with different colors, for clarity. (C) The changes of glutamic acid, GABA,  -ketoglutarate whereas the subjects in moderate and high HLA-conferred risk groups, as (  -KG), and glutamine with age. defi ned in Table I , are marked with circles and squares, respectively. 2978 METABOLOME EN ROUTE TO TYPE 1 DIABETES | Ore š ic ˇ et al. ARTICLE were also low in progressors (Fig. S4, available at http:// (P = 0.009) and 2.5-fold (P = 0.02) lower in the progressors www.jem.org/cgi/content/full/jem.20081800/DC1). Con- than in the nonprogressors, respectively, whereas the BCAAs centration of the most abundant lysoPC, PC(18:0/0:0), was showed a rising trend before IAA emerged. elevated in the progressors during the fi rst 2 yr of life ( Fig. 3 A ). High glutamic acid values were associated with subse- During this period, concentrations of PC(18:0/0:0) and quent appearance of GADA and IAA. Glutamic acid was in- PC(O-18:1/20:4), together with most other ether phospho- creased between 4.1-fold (0 – 9 mo; P = 0.06) and 4.7-fold lipids, were inversely related ( Fig. 3 C ). (9 – 18 mo; P = 0.03) before seroconversion to GADA positivity Although no clear persistent changes were observed in small molecule profi les using GCxGC-TOF/MS ( Fig. 4 ), succinic acid, a key metabolite of citric acid cycle, was diminis- hed fourfold in progressors at birth and during the fi rst year of life (P = 0.04), whereas the citric acid was 1.7-fold de- creased at birth (P = 0.02). Metabolic changes before emergence of islet autoantibodies and diagnosis of type 1 diabetes To determine whether the metabolome abnormalities showed association with the emergence of islet autoantibodies, the metabolomes of the progressors and of the nonprogressors were compared using samples obtained from ± 18 mo from the emergence of the fi rst autoantibody ( Fig. 5 A ). The lyso PC PC(18:0/0:0) was increased 1.5-fold within 9 – 18 mo before seroconversion (P = 0.03) and 1.3-fold within the succeeding 9 mo (P = 0.005; Fig. 5 B ). Metabolomes in the progressors and nonprogressors showed several unexpected diff erences around the time when auto- antibodies appeared. Glutamic acid was 5.2-fold and 32-fold increased 9 – 18 mo (P = 0.02) and 0 – 9 mo before serocon- version (P = 0.02), respectively. Branched chain amino acids (BCAAs) such as leucine and isoleucine were also increased before seroconversion, whereas ketoleucine concentration was diminished. We also studied whether serum metabolome profi les at the presentation of clinical diabetes resembled those seen be- fore the fi rst autoantibody species emerged. The profi les in the progressors during the last visits before diagnosis of type 1 diabetes revealed no clear diff erences as compared with the profi les of matched nonprogressors (Fig. S5, available at http://www.jem.org/cgi/content/full/jem.20081800/DC1), except for specifi c phospholipids which were similarly diminis- hed as they had been at an early age and around the time of seroconversion to autoantibody positivity. When we tested whether the concentrations of the key lipids found altered in progressors before autoimmunity correlated with the age at diagnosis or the time interval between the fi rst seroconver- sion to islet autoimmunity and diagnosis, no signifi cant trends were found (Figs. S6 and S7). Figure 4. Comparison of serum metabolomes between progressors to type 1 diabetes and nonprogressors who remained autoantibody Associations of metabolome with the appearance negative throughout the follow up. The age groups are divided into of specifi c autoantibodies birth (cord blood) and then into groups covering 1-yr periods of followup. The appearance of each autoantibody (insulin autoantibody Only one sample per subject, closest to the mean age within the time [IAA], GADA, IA-2A, or islet cell autoantibody [ICA]) was window (0.5, 1.5, 2.5 yr, etc.), is used in each comparison. Clustering was preceded by a relative increase in the lysoPC PC(18:0/0:0) performed for the profi les across all available samples ( n = 419) using and followed by a transient two to fourfold elevation of Ward linkage and the nonparametric Spearman rank correlation – based 2-hydroxybutyric acid ( Fig. 5 C ). Within 9 – 18 mo and 0 – 9 distance metric. The number of subjects included in each age cohort is mo before the appearance of IAA, ketoleucine was 20-fold shown at the bottom. JEM VOL. 205, December 22, 2008 2979 and 8.4-fold (0 – 9 mo; P = 0.02) before seroconversion to We found that the children who developed type 1 diabe- IAA positivity. tes have reduced serum levels of succinic acid and PC at birth and reduced levels of multiple phospholipids and triglycerides DISCUSSION throughout the follow up. As PC is a major source of choline In this longitudinal study, we compared serum metabolome in in the body, our data suggest that the progressors are choline- children who, during follow up, progressed to type 1 diabetes- defi cient since birth. Besides being an important epigenetic associated autoimmunity and, further, to clinical diabetes regulator ( 17 ), choline also controls the secretion of triglyc- (progressors) or who remained permanently healthy and au- eride-rich very low-density lipoprotein particles. Choline toantibody negative. Our study strongly suggests that metabolic defi ciency leads to lower serum triglyceride levels and their dysregulation precedes overt autoimmunity in type 1 diabetes. increased accumulation in liver ( 18 ). This is consistent with our observation that during infancy and early childhood the progressors have low triglyceride levels. Although diminished choline values were observed in the progressors already at birth, nongenetic regulation cannot be excluded because, for example, choline metabolism is dependent on the composi- tion of the intestinal microbiota ( 19 ). Consistent with this, succinic acid, which is also diminished in the progressors at birth, is as well produced by the human intestinal microbiota ( 20, 21 ). Interaction of gut microbiota with the innate im- mune system is an important factor aff ecting T1D predisposi- tion in the nonobese diabetic mouse ( 22 ). Accordingly, it is conceivable that maternal diet or intestinal fl ora during preg- nancy may aff ect the choline and succinic acid levels in new- born infants and thereby modify systemic energy metabolism and the immune system in the off spring. Our data show that ether PCs were consistently low in the progressors. The plasmalogens, which are the most abundant ether phospholipids in normal serum, protect cells from oxi- dative damage ( 23 ). Plasmalogen defi ciency may also indicate that subjects who are genetically susceptible to type 1 diabetes are particularly prone to specifi c metabolic stress, which not only leads to plasmalogen defi ciency but also triggers progres- sion toward diabetes. One potential pathway linking low Figure 5. Selected metabolite differences between progressors and nonprogressors within 18-mo periods before and after seroconver- sion to islet autoantibody positivity, divided into four 9-mo periods. (A) Illustrative matching of progressors and nonprogressors for studies of seroconversion-related changes in metabolome. For each progressor ’ s samples near the selected period of seroconversion, the age-matched control samples were selected from the matched nonprogressor. (B) Changes before (Ab ) and after (Ab ) fi rst seroconversion to islet auto- antibody positivity. Only one sample from each subject, closest to the time of seroconversion, is included within each time period. Clustering was performed as described in Fig. 4 . (C) Changes before and after seroconver- sion to positivity for each autoantibody. Only one sample from each sub- Figure 6. Biochemical model of metabolic changes in progressors ject, closest to the time of seroconversion, is included within each time to type 1 diabetes prior and after seroconversion to islet autoim- period. Lipids were measured in 56 progressors and 73 nonprogressors, munity. Detected changes of specifi c metabolites of citric acid cycle and whereas the metabolites were measured in 13 progressors and 26 non- glutamic acid metabolism before (left boxes) and shortly after (right progressors as shown in Table S1. boxes) the appearance of GADA are shown. 2980 METABOLOME EN ROUTE TO TYPE 1 DIABETES | Ore š ic ˇ et al. ARTICLE phospholipids with the increased risk of type 1 diabetes is the spectively collected sample series, which covered the time enhanced susceptibility of pancreatic  cells to increased oxi- from birth to overt type 1 diabetes. However, it is essential that dative damage ( 24 ). Decreased antioxidant capacity of  cells these results are validated using other well characterized popu- caused by low ether phospholipids might so enhance their lation cohorts. The German BabyDiab ( 36 ), the US-based susceptibility to the eff ects of reactive oxygen species. DAISY ( 37 ) and PANDA studies ( 38 ), and the multinational The appearance of each islet autoantibody was preceded TEDDY study ( 39 ) will hopefully, in time, be able to confi rm by elevated serum concentration of lysoPCs, which are reac- or reject our fi ndings. To address the hypotheses derived from tive lipid byproducts generated in the hydrolysis of PCs by our fi ndings and to understand the tissue-specifi c mechanisms phospholipase A2. These are lipid mediators able to activate behind the early metabolic disturbances and associated autoim- a range of proinfl ammatory molecules ( 25 ) that function as mune changes preceding the disease onset, experimental mod- natural adjuvants for the immune system ( 26 ). The increased els refl ecting both the metabolic and autoimmune components levels of lysoPCs observed just a few months before serocon- of early type 1 diabetes will have to be established. version to autoantibody positivity may be consistent with the In conclusion, dysregulation of metabolism precedes eff ects of increased oxidative stress associated with, for exam- cell autoimmunity and overt type 1 diabetes. The factors ple, a viral infection ( 2 ) or some other proinfl ammatory event. leading to metabolic stress and autoimmune responses clearly The appearances of autoantibodies against insulin and need to be investigated in further studies in the context of GADA, an enzyme that decarboxylates glutamic acid to autoimmune diseases in general. Our fi ndings also imply that GABA, were preceded by diminished ketoleucine, elevated metabolic or immunomodulatory interventions during the BCAAs, and elevated glutamic acid. Augmentation of insulin preautoimmune period may be used as a potential strategy for secretion by elevated serum BCAAs is well documented ( 27 ) prevention of type 1 diabetes. and high blood levels of BCAAs were previously reported in overt type 1 diabetes ( 28 ). BCAA and glutamic acid levels are MATERIALS AND METHODS restored to normal after the appearance of GADA or IAA DIPP study. The DIPP project has been performed in three Finnish cities ( Fig. 6 ). This raises an intriguing possibility that the initial with a combined annual birth rate of 11,000, representing almost 20% of all births in Finland. The project was launched in the city of Turku in Novem- autoimmune response is physiological and aimed at restoring ber 1994; Oulu joined the study 1 yr later and Tampere 2 yr after that. HLA- the metabolic homeostasis and that the disease may be caused DQ – based genetic typing was used to select children positive for DR3-DQ2 by or be at least partially infl uenced by a defective response and/or DR4-DQ8 risk haplotypes without protective haplotypes ( 6 ). By toward the  cell autoantigens ( 29, 30 ). A protective role of June 6, 2006, 104,111 consecutive newborn infants had been screened, and the diabetes-associated autoantibodies is indirectly supported 8,026 children with genetic risk continued in the follow up. The DIPP study by the observation that the autoantibody titers frequently de- was approved by the local Ethical Committees at the University Hospitals in Turku, Oulu, and Tampere. crease before the presentation of clinical diabetes, which can In Turku the children were monitored at 3-mo intervals until 2 yr of be interpreted as a sign of failing protection ( 31 ). age and then were monitored twice a year, and in Oulu and Tampere they Our fi ndings favor early immunomodulation rather than were monitored at 3, 6, 12, 18, and 24 mo and then annually ( 40 ). At each immunosuppression as a preventive therapy for type 1 diabetes, visit, a venous blood sample was collected from the children without fasting. with the aim of boosting the benefi cial components of auto- After 30 – 60 min at room temperature, serum was separated and transferred immunity. However, despite apparent effi cacy in animal mod- to  70 ° C in cryovials within 3 h from the draw. ICA were measured in all samples. If a child seroconverted to ICA positivity, IAA, GADA, and els of several immunomodulatory agents, success has been poor protein tyrosine phosphatase-like antigen IA-2 (IA-2A) were measured in all in human type 1 diabetes prevention trials, including our re- samples obtained from that child. After seroconversion to autoantibody cent double-blind placebo-controlled nasal insulin study ( 32 ). positivity the visits took place at 3-mo intervals in all clinical centers. As of Notably, in that study the intervention was conducted in January 1, 2003, all four autoantibody types, ICA, GADA, IAA, and IA-2A, children who recently had seroconverted to positivity for two were measured from all samples drawn as previously described ( 32 ). Auto- or more types of diabetes-associated autoantibodies. Our antibodies that were present in cord blood and disappeared before 18 mo of age were regarded as maternal in origin and such values were excluded from fi ndings suggest that this period may already be late for such the analysis. a therapeutic intervention and that the preautoimmune period Between 1997 and 2007, study children with two or more types of au- could be a more suitable window for disease prevention. toantibodies detected in at least two consecutive serum samples have been Early therapeutic strategies aimed at improving or restor- recruited into a randomized double-blind placebo-controlled trial testing the ing the metabolic phenotype, e.g., by reestablishing ether effi cacy of nasal insulin for preventing type 1 diabetes (http://clinicaltrials phospholipid availability or by modulating gut microbial .gov, identifi er: NCT00223613). The study showed that daily nasal insulin administration in an  1 U/kg dose failed to delay diabetes development in composition, might also be of value by preventing  cell de- children with genetic risk and two or more types of autoantibodies ( 32 ). struction and delaying progression to overt type 1 diabetes. In There were no meaningful diff erences in serum metabolome components in support of this, administration of docosahexaenoic acid, the enrolled children, who comprised roughly half of those eligible, com- which is known to be selectively targeted to plasmalogens pared with those who did not participate in the prevention trial. ( 33 ), has been associated with decreased risk of  cell autoim- munity and type 1 diabetes ( 34, 35 ). Lipidomic analysis. 10- μ l serum samples diluted with 10 μ l of 0.15 M One of the strengths of our study is that we used state of NaCl and spiked with a standard mixture containing 10 lipid species the art metabolomics technology to examine our large pro- were extracted with 100 μ l of a 2:1 mixture of chlorofor m and methanol. JEM VOL. 205, December 22, 2008 2981 The extraction time was 0.5 h and the lower organic phase was separated by m/z peaks having more weight than the lower. A similarity value is assigned centrifuging at 10,000 rpm for 3 min. Another standard mixture containing between 0 and 999, with 999 being a perfect match and 750 generally con- three labeled lipid species was added to the extracts and the lipids were ana- sidered as a reasonable match. We used the conservative cut-off criterion of lyzed on a mass spectrometer (Q-Tof Premier; Waters) combined with an 850 for identifi cation. Response curves are shown for selected metabolites Acquity UPLC (Waters). The column, kept at 50 ° C, was an Acquity UPLC of relevance to the study in Figs. S8 and S9 (available at http://www.jem BEH C18, 50 mm, with 1.7- μ m particles (Waters). The solvent system in- .org/cgi/content/full/jem.20081800/DC1). cluded water (1% 1 M ammonium acetate and 0.1% HCOOH) and a mix- ture of acetonitrile and 2-propanol (5:2; 1% 1 M NH Ac and 0.1% HCOOH). 4 Statistical methods. R statistical software (http://www.r-project.org/) The fl ow rate was 0.2 ml/min and the total run time, including column re- was used for data analyses and visualization. The concentrations were equilibration, was 18 min. compared using the Wilcoxon rank-sum test, with p-values < 0.05 con- Raw centroid data from the LC/MS instrument was converted to sidered signifi cant. To account for multiple comparisons, false discovery netCDF fi les, which were processed with MZmine software version 0.60 ( 41, rate was estimated as the maximum q -value ( 43 ) in the set of signifi cant 42 ). Calibration was performed as follows: all monoacyl lipids except choles- diff erences for each dataset (i.e., lipidomics or metabolomics) and the terol esters, such as monoacylglycerols and monoacyl glycerophospholipids, family of hypotheses tested (e.g., diff erences between progressors and were calibrated with the lysoPC PC(17:0/0:0) as internal standard. All nonprogressors in a specifi c age cohort). False discovery rates were com- diacyl lipids except phosphatidylethanolamines were calibrated with the puted using the R package qvalue. Lipid class concentration diff erences PC PC(17:0/17:0), the phosphatidylethanolamines with PE(17:0/17:0), between the mean longitudinal profi les of progressors and nonprogressors and the triacylglycerols and cholesterol esters with the triacylglycerol were tested using linear mixed eff ects models implemented with the lme TG(17:0/17:0/17:0). Calibration-based concentrations are expressed in μ mol/ function in the R package nlme. The fi xed eff ects were modeled as con- liter unless otherwise noted. stant levels for each group, i.e., for progressors and nonprogressors, and The samples were analyzed in four separate runs within 12 mo and the the random eff ects were modeled as constant deviations from these con- data were processed for each analytical run separately (Table S1, available at stant group level trajectories. The signifi cance of the group diff erences http://www.jem.org/cgi/content/full/jem.20081800/DC1). For the fi nal was evaluated by the p-value for the fi xed eff ect parameter estimate of combined dataset consisting of data from all four analytical runs, we matched group diff erences. the lipids across diff erent runs based on the following criteria: lipid identity The fold diff erence was calculated by dividing the median concentra- is the same across all runs; lipid is detected in all four analytical runs (this tion in progressors by the median concentration in nonprogressors. If this stringent criterion is the main reason for relatively low number of lipids in number was less than one, the negative is listed (e.g., 0.75, or a drop of 25% vs. the fi nal dataset and, thus, only the most abundant lipids within each class are nonprogressors, is reported as a  1.3-fold diff erence). included); starting from the fi rst analytical run as reference (Turku DIPP se- ries), retention time deviation is < 2%; and starting from the fi rst analytical Online supplemental material. Fig. S1 shows lysoPC levels across all run as reference (Turku DIPP series), deviation of median intensity (calcu- 1,196 samples measured as a function of sample storage time. Fig. S2 shows lated within the analytical run) between analytical runs is < 50%. the total PC concentration in cord serum of progressors and nonprogressors. Fig. S3 compares the lipidomes between children with the high-risk HLA genotype and those with moderate-risk HLA genotypes. Fig. S4 shows the Metabolomic analysis using the GCxGC-TOF/MS platform. 10 μ l of concentrations of ethanolamine plasmalogen detected in a subset of progres- an internal standard-labeled palmitic acid (16:0-16,16,16d ; 500 mg/l) and sors and nonprogressors. Diff erences in selected metabolite concentrations 400 μ l of methanol solvent were added to 20 μ l of the sample. After vortex- between the progressors to type 1 diabetes and nonprogressors within the ing for 1 min and incubating for 30 min at room temperature, the superna- four 9-mo intervals before disease diagnosis in the progressors are shown in tant was separated by centrifugation at 10,000 rpm for 5 min at room Fig. S5. Associations of lysoPC and ether PC with the rate of progression to temperature. The sample was dried under constant fl ow of nitrogen gas. 25 μ l type 1 diabetes are shown in Figs. S6 and S7, respectively. Figs. S8 and S9 MOX (2% methoxyamine HCl in pyridine) was added to the dried sample. show the selected characteristics of the GCxGC-TOF/MS analytical plat- The mixture was then incubated at 45 ° C for 1 h and derivatized with 25 μ l form. Table S1 provides an overview of the four analytical runs used to gen- N -methyl- N -(trimethylsilyl)trifl uoroacetamide by incubating at 45 ° C for 1 h. erate the data for this study. Online supplemental material is available at 10 μ l of retention index standard mixture with fi ve alkanes at 800 ppm was http://www.jem.org/cgi/content/full/jem.20081800/DC1. added to the metabolite mixture. Sample order for analysis was established by randomization. We thank the dedicated personnel of the DIPP study in Turku, Oulu, and Tampere The instrument used was a GCxGC-TOF mass spectrometer (Pegasus and the DIPP families for their support. We acknowledge the technical support of 4D; Leco) with an autosampler (6890N GC and Combi PAL;Agilent Tech- Leena Ö hrnberg, Anna-Liisa Ruskeep ä ä , Jarkko Miettinen, Airi Hyrk ä s, and Jing Tang. nologies). The instrument parameters were as follows: 1- μ l split injection 1:20; We thank Antonio Vidal-Puig, Fredrik B ä ckhed, and Kari Alitalo for discussions fi rst column, RTX-5, 10 m × 180 μ m × 0.20 μ m; second column, BPX-50, and comments. 1.50 m × 100 μ m × 0.10 μ m; helium 39.6 psig constant pressure; temperature The study was funded by Juvenile Diabetes Research Foundation International programs, primary oven, Initial 50 ° C, 1 min → 295 ° C, 7 ° C/min, 3 min, and (4-1999-731 and 4-2001-435 to O. Simell), Tekes FinnWell Program (M. Ore š ic ˇ , R. secondary oven, +20 ° C above primary oven temperature; second dimension Lahesmaa, and O. Simell), DIAPREPP project of the Seventh Framework Program of the separation time 5 s; MS measurement 45 – 700 amu, 100 spectra/s. Raw data European Community (HEALTH-F2-2008-202013 to M. Ore š ic ˇ and O. Simell), Academy were processed using ChromaTOF software (Leco) followed by alignment of Finland, Sigrid Juselius Foundation, P ä ivikki and Sakari Sohlberg Foundation, Signe and normalization using the in house – developed software. Unwanted back- and Ane Gyllenberg Foundation, Diabetes Research Foundation, Finland, and Special ground peaks were eliminated using the classifi cations feature of ChromaTOF Federal Funds to the Turku, Oulu, and Tampere University Hospitals. software. In house – developed software was used to perform additional fi ltering The authors have no confl icting fi nancial interests. using compound identifi cations by ChromaTOF. The metabolites were identifi ed using an in house reference compound Submitted: 12 August 2008 library as well as by searching the reference mass spectral library. Mass spectra Accepted: 18 November 2008 from the GCxGC-TOF/MS analysis were searched against The Palisade Complete Mass Spectral Library, 600K Edition (Palisade Mass Spectrome- try), which includes all spectra available from the NIST 2002 and Wiley reg- REFERENCES istry collections as well as 150,000 other spectra. The matches to reference 1 . Gale , E.A.M. 2002 . The rise of childhood type 1 diabetes in the 20th spectra are based on a weighted dot product of the two spectra, with higher century. Diabetes . 51 : 3353 – 3361 . 2982 METABOLOME EN ROUTE TO TYPE 1 DIABETES | Ore š ic ˇ et al. ARTICLE 2 . Knip , M. , R. Veijola , S.M. Virtanen , H. Hy ö ty , O. Vaarala , and H.K. 21 . Gill , S.R. , M. Pop , R.T. DeBoy , P.B. Eckburg , P.J. Turnbaugh , B.S. Å kerblom . 2005 . Environmental triggers and determinants of type 1 Samuel , J.I. Gordon , D.A. Relman , C.M. Fraser-Liggett , and K.E. diabetes. Diabetes . 54 : S125 – S136 . Nelson . 2006 . Metagenomic analysis of the human distal gut microbi- 3 . Tuomilehto , J. , M. Karvonen , J. Pitk ä niemi , E. Virtala , K. Kohtam ä ki , ome. Science . 312 : 1355 – 1359 . L. Toivanen , and E. Tuomilehto-Wolf . 1999 . Record-high incidence 22 . Wen , L. , R.E. Ley , P.Y. Volchkov , P.B. Stranges , L. Avanesyan , A.C. of type 1 (insulin-dependent) diabetes mellitus in Finnish children. Stonebraker , C. Hu , F.S. Wong , G.L. Szot , J.A. Bluestone , et al . 2008 . Diabetologia . 42 : 655 – 660 . Innate immunity and intestinal microbiota in the development of Type 4 . Harjutsalo , V. , L. Sj ö berg , and J. Tuomilehto . 2008 . Time trends in the 1 diabetes. Nature . 455 : 1109 – 1113 . incidence of type 1 diabetes in Finnish children: a cohort study. Lancet . 23 . Zoeller , R.A. , T.J. Grazia , P. LaCamera , J. Park , D.P. Gaposchkin , 371 : 1777 – 1782 . and H.W. Farber . 2002 . Increasing plasmalogen levels protects human 5 . Achenbach , P. , E. Bonifacio , K. Koczwara , and A.-G. Ziegler . 2005 . endothelial cells during hypoxia. Am. J. Physiol. Heart Circ. Physiol. Natural history of type 1 diabetes. Diabetes . 54 : S25 – S31 . 283 : H671 – H679 . 6 . Kupila , A. , P. Muona , T. Simell , P. Arvilommi , H. Savolainen , A.-M. 24 . Lenzen , S. , J. Drinkgern , and M. Tiedge . 1996 . Low antioxidant enzyme H ä m ä l ä inen , S. Korhonen , T. Kimpim ä ki , M. Sj å roos , J. Ilonen , et al . gene expression in pancreatic islets compared with various other mouse 2001 . Feasibility of genetic and immunological prediction of type 1 dia- tissues. Free Radic. Biol. Med. 20 : 463 – 466 . betes in a population-based birth cohort. Diabetologia . 44 : 290 – 297 . 25 . Mehta , D. 2005 . Lysophosphatidylcholine: an enigmatic lysolipid. Am. 7 . Ore š i c ˇ , M. , A. Vidal-Puig , and V. H ä nninen . 2006 . Metabolomic ap- J. Physiol. Lung Cell. Mol. Physiol. 289 : L174 – L175 . proaches to phenotype characterization and applications to complex dis- 26 . Kabarowski , J.H.S. , Y. Xu , and O.N. Witte . 2002 . Lysophospha- eases. Expert Rev. Mol. Diagn. 6 : 575 – 585 . tidylcholine as a ligand for immunoregulation. Biochem. Pharmacol. 8 . Nikkil ä , J. , M. Sysi-Aho , A. Ermolov , T. Sepp ä nen-Laakso , O. Simell , 64 : 161 – 167 . S. Kaski , and M. Ore š ic ˇ . 2008 . Gender dependent progression of sys- 27 . Floyd , J.C.J. , S.S. Fajans , R.F. Knopf , and J.W. Conn . 1963 . Evidence temic metabolic states in early childhood. Mol. Syst. Biol. 4 : 197 . that insulin release is the mechanism for experimentally induced leucine 9 . Lenz , E.M. , J. Bright , I.D. Wilson , A. Hughes , J. Morrisson , H. hypoglycemia in man. J. Clin. Invest. 42 : 1714 – 1719 . 28 . Marliss , E.B. , A.F. Nakhooda , P. Poussier , and A.A. Sima . 1982 . The Lindberg , and A. Lockton . 2004 . Metabonomics, dietary infl uences and cultural diff erences: a 1H NMR-based study of urine samples obtained diabetic syndrome of the ‘ BB ’ Wistar rat: possible relevance to type 1 (insulin-dependent) diabetes in man. Diabetologia . 22 : 225 – 232 . from healthy British and Swedish subjects. J. Pharm. Biomed. Anal. 36 : 841 – 849 . 29 . Nevo , U. , I. Golding , A.U. Neumann , M. Schwartz , and S. Akselrod . 2004 . Autoimmunity as an immune defense against degenerative pro- 10 . B ä ckhed , F. , H. Ding , T. Wang , L.V. Hooper , G.Y. Koh , A. Nagy , C.F. Semenkovich , and J.I. Gordon . 2004 . The gut microbiota as an en- cesses: a primary mathematical model illustrating the bright side of auto- immunity. J. Theor. Biol. 227 : 583 – 592 . vironmental factor that regulates fat storage. Proc. Natl. Acad. Sci. USA . 101 : 15718 – 15723 . 30 . Hauben , E. , M.G. Roncarolo , U. Nevo , and M. Schwartz . 2005 . Benefi cial autoimmunity in type 1 diabetes mellitus. Trends Immunol. 11 . Gluckman , P.D. , M.A. Hanson , C. Cooper , and K.L. Thornburg . 2008 . Eff ect of in utero and early-life conditions on adult health and disease. 26 : 248 – 253 . 31 . Kimpimaki , T. , P. Kulmala , K. Savola , A. Kupila , S. Korhonen , T. N. Engl. J. Med. 359 : 61 – 73 . 12 . Simell , O. , H. Niinikoski , T. R ö nnemaa , H. Lapinleimu , T. Routi , Simell , J. Ilonen , O. Simell , and M. Knip . 2002 . Natural history of { beta } -cell autoimmunity in young children with increased genetic H. Lagstr ö m , P. Salo , E. Jokinen , and J. Viikari . 2000 . Special Turku Coronary Risk Factor Intervention Project for Babies (STRIP). Am. susceptibility to type 1 diabetes recruited from the general population. J. Clin. Endocrinol. Metab. 87 : 4572 – 4579 . J. Clin. Nutr. 72 : 1316S – 1331S . 13 . Watson , A.D. 2006 . Thematic review series: systems biology ap- 32 . N ä nt ö -Salonen , K. , A. Kupila , S. Simell , H. Siljander , T. Salonsaari , A. Hekkala , S. Korhonen , R. Erkkola , J.I. Sipil ä , L. Haavisto , et al . 2008 . proaches to metabolic and cardiovascular disorders. Lipidomics: a global approach to lipid analysis in biological systems. J. Lipid Res. Nasal insulin to prevent type 1 diabetes in children with HLA genotypes and autoantibodies conferring increased risk of disease: a double-blind, 47 : 2101 – 2111 . 14 . Ore š ic ˇ , M. , V.A. H ä nninen , and A. Vidal-Puig . 2008 . Lipidomics: a randomised controlled trial. Lancet . 372 : 1746 – 1755 . 33 . Gaposchkin , D.P. , and R.A. Zoeller . 1999 . Plasmalogen status infl u- new window to biomedical frontiers. Trends Biotechnol. 26 : 647 – 652 . 15 . Laaksonen , R. , M. Katajamaa , H. P ä iv ä , M. Sysi-Aho , L. Saarinen , P. ences docosahexaenoic acid levels in a macrophage cell line: insights using ether lipid-defi cient variants. J. Lipid Res. 40 : 495 – 503 . Junni , D. L ü tjohann , J. Smet , R.V. Coster , T. Sepp ä nen-Laakso , et al . 2006 . A systems biology strategy reveals biological pathways and plasma 34 . Stene , L.C. , and G. Joner , and the Norwegian Childhood Diabetes Study Group. 2003 . Use of cod liver oil during the fi rst year of biomarker candidates for potentially toxic statin induced changes in muscle. PLoS ONE . 1 : e97 . life is associated with lower risk of childhood-onset type 1 diabe- tes: a large, population-based, case-control study. Am. J. Clin. Nutr. 16 . Welthagen , W. , R.A. Shellie , J. Spranger , M. Ristow , R. Zimmermann , and O. Fiehn . 2005 . Comprehensive two-dimensional gas chroma- 78 : 1128 – 1134 . 35 . Norris , J.M. , X. Yin , M.M. Lamb , K. Barriga , J. Seifert , M. tography time-of-fl ight mass spectrometry (GC — GC-TOF) for high resolution metabolomics: biomarker discovery on spleen tissue ex- Hoff man , H.D. Orton , A.E. Baron , M. Clare-Salzler , H.P. Chase , et al . 2007 . Omega-3 polyunsaturated fatty acid intake and islet au- tracts of obese NZO compared to lean C57BL/6 mice. Metabolomics . toimmunity in children at increased risk for type 1 diabetes. JAMA . 1 : 65 – 73 . 298 : 1420 – 1428 . 17 . Niculescu , M.D. , C.N. Craciunescu , and S.H. Zeisel . 2006 . Dietary cho- 36 . Achenbach , P. , K. Koczwara , A. Knopff , H. Naserke , A.-G. Ziegler , line defi ciency alters global and gene-specifi c DNA methylation in the and E. Bonifacio . 2004 . Mature high-affi nity immune responses to developing hippocampus of mouse fetal brains. FASEB J. 20 : 43 – 49 . (pro)insulin anticipate the autoimmune cascade that leads to type 1 dia- 18 . Yao , Z.M. , and D. Vance . 1988 . The active synthesis of phosphatidyl- betes. J. Clin. Invest. 114 : 589 – 597 . choline is required for very low density lipoprotein secretion from rat 37 . Baschal , E.E. , T.A. Aly , S.R. Babu , M.S. Fernando , L. Yu , D. Miao , hepatocytes. J. Biol. Chem. 263 : 2998 – 3004 . 19 . Dumas , M.-E. , R.H. Barton , A. Toye , O. Cloarec , C. Blancher, A. K.J. Barriga , J.M. Norris , J.A. Noble , H.A. Erlich , et al . 2007 . HLA- DPB1*0402 protects against type 1A diabetic autoimmunity in the Rothwell , J. Fearnside , R. Tatoud , V. Blanc , J.C. Lindon , et al . 2006 . Metabolic profi ling reveals a contribution of gut microbiota to fatty highest risk DR3-DQB1*0201/DR4-DQB1*0302 DAISY population. Diabetes . 56 : 2405 – 2409 . liver phenotype in insulin-resistant mice. Proc. Natl. Acad. Sci. USA . 103 : 12511 – 12516 . 38 . Carmichael , S.K. , S.B. Johnson , A. Baughcum , K. North , D. Hopkins , 20 . Macy , J.M. , L.G. Ljungdahl , and G. Gottschalk . 1978 . Pathway of suc- M.G. Dukes , J.X. She , and D.A. Schatz . 2003 . Prospective assessment cinate and propionate formation in Bacteroides fragilis . J. Bacteriol. in newborns of diabetes autoimmunity (PANDA): maternal understand- 134 : 84 – 91 . ing of infant diabetes risk. Genet. Med. 5 : 77 – 83 . JEM VOL. 205, December 22, 2008 2983 39 . Hagopian , W.A. , A. Lernmark , M.J. Rewers , O.G. Simell , J.-X. She , 41 . Katajamaa , M. , and M. Ore š ic ˇ . 2005 . Processing methods for dif- A.G. Ziegler , J.P. Krischer , and B. Akolkar . 2006 . TEDDY - the en- ferential analysis of LC/MS profi le data. BMC Bioinformatics . vironmental determinants of diabetes in the young: an observational 6 : 179 – 190 . clinical trial. Ann. N. Y. Acad. Sci. 1079 : 320 – 326 . 42 . Katajamaa , M. , J. Miettinen , and M. Ore š ic ˇ . 2006 . MZmine: toolbox 40 . Kupila , A. , P. Keskinen , T. Simell , S. Erkkil ä , P. Arvilommi , S. for processing and visualization of mass spectrometry based molecular Korhonen , T. Kimpim ä ki , M. Sj ö roos , M. Ronkainen , J. Ilonen , profi le data. Bioinformatics . 22 : 634 – 636 . et al . 2002 . Genetic risk determines the emergence of diabetes-associated 43 . Storey , J.D. 2002 . A direct approach to false discovery rates. Journal of the autoantibodies in young children. Diabetes . 51 : 646 – 651 . Royal Statistical Society Series B . 64 : 479 – 498 . 2984 METABOLOME EN ROUTE TO TYPE 1 DIABETES | Ore š ic ˇ et al. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Journal of Experimental Medicine Pubmed Central

Dysregulation of lipid and amino acid metabolism precedes islet autoimmunity in children who later progress to type 1 diabetes

Download PDF
10 pages

Loading next page...
 
/lp/pubmed-central/dysregulation-of-lipid-and-amino-acid-metabolism-precedes-islet-Vws34qe5Qj

References (50)

Publisher
Pubmed Central
Copyright
© 2008 Orešič et al.
ISSN
0022-1007
eISSN
1540-9538
DOI
10.1084/jem.20081800
Publisher site
See Article on Publisher Site

Abstract

ARTICLE Dysregulation of lipid and amino acid metabolism precedes islet autoimmunity in children who later progress to type 1 diabetes 1 2 1 2 Matej Ore š i c ˇ , Satu Simell , Marko Sysi-Aho, Kirsti N ä nt ö -Salonen , 1 2 1 Tuulikki Sepp ä nen-Laakso , Vilhelmiina Parikka , Mikko Katajamaa , 4 1 5 1 Anne Hekkala , Ismo Mattila , P ä ivi Keskinen , Laxman Yetukuri , 6 5 1 2 Arja Reinikainen , Jyrki L ä hde , Tapani Suortti , Jari Hakalax , 2 7,8 4 3,9 Tuula Simell , Heikki Hy ö ty , Riitta Veijola , Jorma Ilonen , 6 5,10 2 Riitta Lahesmaa , Mikael Knip , and Olli Simell VTT Technical Research Centre of Finland, Espoo FI-02044, Finland 2 3 Department of Pediatrics and Immunogenetics Laboratory, University of Turku, Turku FI-20520, Finland Department of Pediatrics, University of Oulu, Oulu FI-90014, Finland Department of Pediatrics, Tampere University Hospital, Tampere FI-33521, Finland Turku Centre for Biotechnology, Turku FI-20521, Finland Department of Virology, University of Tampere, Tampere FI-33520, Finland Centre for Laboratory Medicine, University Hospital of Tampere, Tampere FI-33520, Finland Department of Clinical Microbiology, University of Kuopio, FI-70211 Kuopio, Finland Hospital for Children and Adolescents, University of Helsinki, Helsinki FI-00014, Finland The risk determinants of type 1 diabetes, initiators of autoimmune response, mechanisms regulating progress toward  cell failure, and factors determining time of presentation of clinical diabetes are poorly understood. We investigated changes in the serum metabolome prospectively in children who later progressed to type 1 diabetes. Serum metabolite profi les were compared between sample series drawn from 56 children who progressed to type 1 diabetes and 73 controls who remained nondiabetic and permanently autoantibody nega- tive. Individuals who developed diabetes had reduced serum levels of succinic acid and phosphatidylcholine (PC) at birth, reduced levels of triglycerides and antioxidant ether phospholipids throughout the follow up, and increased levels of proinfl ammatory lysoPCs several months before seroconversion to autoantibody positivity. The lipid changes were not attributable to HLA-associated genetic risk. The appearance of insulin and glutamic acid CORRESPONDENCE decarboxylase autoantibodies was preceded by diminished ketoleucine and elevated glu- Matej Ore š ic ˇ : tamic acid. The metabolic profi le was partially normalized after the seroconversion. Auto- [email protected] OR immunity may thus be a relatively late response to the early metabolic disturbances. Olli Simell: Recognition of these preautoimmune alterations may aid in studies of disease pathogenesis [email protected] and may open a time window for novel type 1 diabetes prevention strategies. Abbreviations used: BCAA, branched chain amino acids; DIPP, Type 1 Diabetes predic- The incidence of type 1 diabetes among chil- a linear steady increase from 1965 to 1996 ( 3 ), tion and prevention study; dren and adolescents has increased markedly in was recently found to be increasing even faster GABA,  -aminobutyric acid; the Western countries during the recent de- than before, with the number of cases diag- GADA, glutamic acid decarbox- ylase antibody; GCxGC-TOF/ cades ( 1 ). The incidence has reached record nosed at or before 14 yr of age expected to MS, two-dimensional gas chro- levels in Finland, where, currently, 1/120 chil- double in the next 15 yr ( 4 ). Although  70% matography coupled to time-of- dren develops type 1 diabetes before the age of of subjects with type 1 diabetes carry defi ned fl ight mass spectrometry; IAA, insulin autoantibody; ICA, islet 15 yr ( 2 ). The annual incidence, which showed cell autoantibody; PC, phospha- © 2008 Orešicˇ et al. This article is distributed under the terms of an Attribu- tion–Noncommercial–Share Alike–No Mirror Sites license for the fi rst six months tidylcholine; STRIP, Special after the publication date (see http://www.jem.org/misc/terms.shtml). After six Turku Coronary Risk Factor S. Simell, M. Sysi-Aho, K. N ä nt ö -Salonen, T. Sepp ä nen- months it is available under a Creative Commons License (Attribution–Noncom- Intervention Project for Laakso, V. Parikka, and M. Katajamaa contributed equally to mercial–Share Alike 3.0 Unported license, as described at http://creativecommons Children. this paper. .org/licenses/by-nc-sa/3.0/). The Rockefeller University Press $30.00 J. Exp. Med. Vol. 205 No. 13 2975-2984 2975 www.jem.org/cgi/doi/10.1084/jem.20081800 The Journal of Experimental Medicine risk-associated genotypes at the HLA locus, only 3 – 7% of the To compare the fi ndings in the genetically defi ned DIPP carriers of such genetic risk markers develop the disease ( 5 ). cohort to those in another cohort of children that was not Seroconversion to islet autoantibody positivity has been genetically defi ned, the metabolomes of another six children the fi rst detectable signal for the onset of autoimmunity and in the Special Turku Coronary Risk Factor Intervention progression toward diabetes. Although seroconversion usually Project for Children (STRIP) ( 12 ) who had developed type 1 precedes the clinical disease by months to years, its occur- diabetes were investigated before disease presentation and rence may already be too late for therapeutic approaches compared with their six age- and sex-matched nondiabetic aimed at preventing progression to overt diabetes. As long as controls from the same study. The STRIP series also com- the initiators of the autoimmune response remain unknown prised 87 samples, which had been collected annually be- and the mechanisms supporting progression toward  cell tween the ages of 7 mo and 5 – 15 yr. Three of the children failure are poorly understood, the estimation of the absolute who developed diabetes turned out to carry the defi ned disease risk and time of disease presentation in genetically HLA-DQB1 risk alleles, but all had had multiple diabetes- susceptible individuals and the discovery of eff ective preven- associated autoanti bodies for 2 – 13 yr before the presentation tion remain a challenge. In 1994, an ongoing birth cohort study (Type 1 Diabetes prediction and prevention study [DIPP]) was launched in Finland ( 6 ). After informed parental consent, HLA alleles associ- ated with type 1 diabetes risk or protection were analyzed from cord blood, and subjects carrying risk-associated genotypes were followed over time to establish the time point of sero- conversion and diabetes development. Over a period of 11.5 yr until 2006, > 100,000 newborn infants have been screened for genetic risk and > 8,000 at-risk children are being followed. Metabolomics systematically studies the chemical fi nger- prints in cells, tissues, and biofl uids in a given physiological and environmental context. New analytical methods combined with bioinformatics provide a way to measure the extended metabolome ( 7 ). The metabolic phenotype is sensitive to subtle but pathogenically relevant factors such as age, lifestyle, nutrition, and the microbe environment of the gut ( 8 – 10 ). Changes in the concentrations of metabolites during early development may thus refl ect both genetic and environmental factors infl uencing later susceptibility to chronic diseases ( 11 ). In this paper, we hypothesize that alterations in the meta- bolic phenotypes characterize the early pathogenesis of type 1 diabetes. We apply the metabolomics strategy to the unique preautoimmune and prediabetic sample series collected in DIPP and provide evidence that metabolic disturbances pre- cede the autoimmunity that is characteristically observed be- fore development of type 1 diabetes. RESULTS Subjects Serum metabolite profi les were compared between 50 DIPP children who progressed to type 1 diabetes (the progressors) and 67 nonprogressors, who remained healthy and autoanti- body negative and were matched with the progressors for time and site of birth, gender, and HLA risk group ( Fig. 1 ). The 117 DIPP children contributed 1,109 samples (9.5 sam- ples per child on average) covering the time from 3 mo of age to clinical diabetes in the progressors. Additionally, the cord blood samples of 39 DIPP children from Turku, 15 of which Figure 1. Subject selection fl owchart for the metabolomics study. progressed to diabetes before the age of 12 yr, were analyzed. Nonprogressors were matched with progressors for time and site of birth, Two progressors included in the cord blood series were also gender, and HLA risk group. Status of DIPP ( 6 ) and STRIP ( 12 ) studies as of included in the longitudinal sample series. June 6, 2006. 2976 METABOLOME EN ROUTE TO TYPE 1 DIABETES | Ore š ic ˇ et al. ARTICLE of diabetes. Selected characteristics of the study subjects are phosphatidylcholine (PC), as well as high lysoPC and leucine shown in T able I . ( Fig. 2 B ). This period was also accompanied by abnormally high glutamic acid and  -aminobutyric acid (GABA), as well Analytical platforms as low glutamine and  -ketoglutarate ( Fig. 2 C ). During glu- Current lipidomics platforms constitute an effi cient way to tamic acid decarboxylase antibody (GADA) positivity, these separate and quantify hundreds of lipid molecules ( 13, 14 ). metabolite concentrations almost normalized, whereas ala- For this study, ultra performance liquid chromatography nine diminished and 2-hydroxybutyric acid increased tran- coupled to mass spectrometry was applied as previously des- siently (unpublished data). cribed ( 8, 15 ). The fi nal lipidomics dataset, not including the cord blood data, comprised 53 identifi ed lipids mea- Metabolome in prospective sample series and cord blood sured in each of the 1,196 samples. A total of 515 samples The analysis of longitudinal series of serum metabolomes were from progressors to type 1 diabetes, among which during the asymptomatic period from children who later de- 112 were collected before seroconversion to autoantibody veloped diabetes revealed characteristic metabolite patterns positivity. In each of the 39 cord blood samples, 249 lipids that preceded or accompanied the initial seroconversion to were identifi ed. Once the complete dataset was constructed, autoantibody positivity ( Fig. 3 ). Compared with children we fi rst confi rmed that the samples were not aff ected by who remained autoantibody negative, the serum lipidome of the time of storage (Fig. S1, available at http://www.jem progressors showed a consistent decrease in triglycerides (P = .org/cgi/content/full/jem.20081800/DC1). 0.005) and multiple phospholipids, including ether PCs (P < Serum samples were also analyzed by two-dimensional 0.001), PCs (P = 0.04), and sphingomyelins (P = 0.09), in gas chromatography coupled to time of fl ight mass spectrom- samples obtained from infancy to early school years ( Fig. 3, etry (GCxGC-TOF/MS). This analytical technology pro- A and B ). The diff erence in lipidome was observed before the vides a broad profi le of small molecules ( 16 ). The analysis was autoantibodies appeared and it persisted throughout the en- performed on a random subset of DIPP study samples from tire age range covered. Low phospholipid levels were also Turku, comprising sample series from 13 progressors and 26 observed in cord blood, although only one of the lipids de- matched nonprogressors. The fi nal metabolomics dataset tected in the longitudinal sample series diff ered signifi cantly obtained by GCxGC-TOF/MS consisted of 75 identifi ed between the progressors and nonprogressors. However, the metabolites measured in each of the 419 samples. total amount of cord serum PC was 1.22-fold decreased in the progressors (P = 0.004), as shown in Fig. S2 (available at Case report http://www.jem.org/cgi/content/full/jem.20081800/DC1). The interdependence of metabolic and immune system fac- We also compared the lipid concentration diff erences be- tors is fi rst demonstrated in Fig. 2 for a girl with an extended tween the children at high and moderate HLA-associated gene- sample series available. She developed overt type 1 diabetes tic risk of developing type 1 diabetes. No signifi cant diff erences close to 9 yr of age ( Fig. 2 A ). The time before the appear- between the genetic risk groups were identifi ed among the ance of the fi rst autoantibody was characterized by abnor- progressors or the nonprogressors (Fig. S3, available at http:// mally low succinic acid (not depicted), ketoleucine, and ether www.jem.org/cgi/content/full/jem.20081800/DC1), thus Table I. Selected characteristics of the subjects included in the study. Progressors Nonprogressors Gender (girls, boys) 31, 25 40, 33 City of birth (Turku, Oulu) 29, 27 45, 28 Season at time of birth 11, 19, 14, 12 15, 25, 19, 14 (winter, spring, summer, autumn) Season at time of diagnosis 11, 21, 10, 13 NA (winter, spring, summer, autumn) Age at time of diagnosis 44, 6 – 162 NA (median, range in months) Age at time of fi rst seroconversion 12, 6 – 96 NA (median, range in months) HLA risk haplotypes Moderate risk DR4-DQ8/x 33 46 (x = any haplotype except DR2-DQ6, DR5-DQ7 or DR3-DQ2) High risk (DR3-DQ2/DR4-DQ8) 17 20 Seasons are defi ned as winter for December to February, spring for March to May, summer for June to August, and autumn for September to November. HLA genotype information is shown for the DIPP children only. One nonprogressor, not included in the table, had the moderate risk DR3-DQ2/x haplotype. JEM VOL. 205, December 22, 2008 2977 confi rming that the observed lipid changes are not attribut- able to HLA-associated genetic risk. One subclass of phospholipids that was repeatedly low in progressors and showed low within-person variability were ether PCs ( Fig. 3, A and B ). Ethanolamine plasmalogen levels Figure 3. Serum lipidome in cord blood and prospective sample series. (A) Differences in serum lipidome between the progressors to type 1 diabetes and nonprogressors who remained autoantibody negative throughout the follow up. The age groups are divided into birth (cord Figure 2. Selected autoantibody and metabolite changes during blood) and then into groups covering 1-yr cohorts. Only one sample per the prediabetic period in a girl who progressed to type 1 diabetes at subject, closest to the mean age within the time window, is used in each close to 9 yr of age (HLA-haplotype DR7-DQ2/DR4-DQ8, moderate comparison. The number of subjects included in each age cohort is shown genetic risk). (A) GADA and IAA profi les. The relative autoantibody level at the bottom. The last sample of each progressor was the sample drawn values are calculated as the ratio of their measured levels and corre- last before diagnosis of diabetes. (B) Total triglyceride and ether PC con- sponding cutoff limits for autoantibody positivity. Scales are linear and centrations for the age cohorts shown in A calculated as the sum of lipid adjusted separately for each autoantibody for clarity. Time of seroconver- concentrations within each class. Both lipid classes were found consis- sion to positivity for each autoantibody is marked, as is the time of diag- tently down-regulated in the progressors, as tested by the linear mixed nosis of type 1 diabetes. The child also seroconverted to ICA and IA-2A effects model for the overall trend throughout the follow up. The error autoantibody positivity within 6 mo before the appearance of IAA. bars show SEM. (C) Levels of the lysoPC PC(18:0/0:0) and the ether PC (B) Changes of ketoleucine, leucine, lysoPC (the profi le shown is for the most PC(O-18:1/20:4) at the age of 18 mo for DIPP children. The two lipids abundant lysoPC species PC(18:0/0:0)), and ether PC (the profi le shown is were up-regulated (P = 0.0009) and down-regulated (P = 0.04) in pro- for the most abundant ether PC species PC(O-18:1/20:4)) with age. Fold gressors, respectively. Only one measurement, closest to the age of 18 mo changes are calculated as ratios of metabolite level at a given time point within the 12 – 24-mo age interval is shown for each subject. Subjects and the normal level, i.e., the mean value across all samples in the non- who already seroconverted to autoantibody positivity against one or mul- progressors. Scales are linear and adjusted separately for each metabolite tiple islet autoantigens are specifi cally marked with different colors, for clarity. (C) The changes of glutamic acid, GABA,  -ketoglutarate whereas the subjects in moderate and high HLA-conferred risk groups, as (  -KG), and glutamine with age. defi ned in Table I , are marked with circles and squares, respectively. 2978 METABOLOME EN ROUTE TO TYPE 1 DIABETES | Ore š ic ˇ et al. ARTICLE were also low in progressors (Fig. S4, available at http:// (P = 0.009) and 2.5-fold (P = 0.02) lower in the progressors www.jem.org/cgi/content/full/jem.20081800/DC1). Con- than in the nonprogressors, respectively, whereas the BCAAs centration of the most abundant lysoPC, PC(18:0/0:0), was showed a rising trend before IAA emerged. elevated in the progressors during the fi rst 2 yr of life ( Fig. 3 A ). High glutamic acid values were associated with subse- During this period, concentrations of PC(18:0/0:0) and quent appearance of GADA and IAA. Glutamic acid was in- PC(O-18:1/20:4), together with most other ether phospho- creased between 4.1-fold (0 – 9 mo; P = 0.06) and 4.7-fold lipids, were inversely related ( Fig. 3 C ). (9 – 18 mo; P = 0.03) before seroconversion to GADA positivity Although no clear persistent changes were observed in small molecule profi les using GCxGC-TOF/MS ( Fig. 4 ), succinic acid, a key metabolite of citric acid cycle, was diminis- hed fourfold in progressors at birth and during the fi rst year of life (P = 0.04), whereas the citric acid was 1.7-fold de- creased at birth (P = 0.02). Metabolic changes before emergence of islet autoantibodies and diagnosis of type 1 diabetes To determine whether the metabolome abnormalities showed association with the emergence of islet autoantibodies, the metabolomes of the progressors and of the nonprogressors were compared using samples obtained from ± 18 mo from the emergence of the fi rst autoantibody ( Fig. 5 A ). The lyso PC PC(18:0/0:0) was increased 1.5-fold within 9 – 18 mo before seroconversion (P = 0.03) and 1.3-fold within the succeeding 9 mo (P = 0.005; Fig. 5 B ). Metabolomes in the progressors and nonprogressors showed several unexpected diff erences around the time when auto- antibodies appeared. Glutamic acid was 5.2-fold and 32-fold increased 9 – 18 mo (P = 0.02) and 0 – 9 mo before serocon- version (P = 0.02), respectively. Branched chain amino acids (BCAAs) such as leucine and isoleucine were also increased before seroconversion, whereas ketoleucine concentration was diminished. We also studied whether serum metabolome profi les at the presentation of clinical diabetes resembled those seen be- fore the fi rst autoantibody species emerged. The profi les in the progressors during the last visits before diagnosis of type 1 diabetes revealed no clear diff erences as compared with the profi les of matched nonprogressors (Fig. S5, available at http://www.jem.org/cgi/content/full/jem.20081800/DC1), except for specifi c phospholipids which were similarly diminis- hed as they had been at an early age and around the time of seroconversion to autoantibody positivity. When we tested whether the concentrations of the key lipids found altered in progressors before autoimmunity correlated with the age at diagnosis or the time interval between the fi rst seroconver- sion to islet autoimmunity and diagnosis, no signifi cant trends were found (Figs. S6 and S7). Figure 4. Comparison of serum metabolomes between progressors to type 1 diabetes and nonprogressors who remained autoantibody Associations of metabolome with the appearance negative throughout the follow up. The age groups are divided into of specifi c autoantibodies birth (cord blood) and then into groups covering 1-yr periods of followup. The appearance of each autoantibody (insulin autoantibody Only one sample per subject, closest to the mean age within the time [IAA], GADA, IA-2A, or islet cell autoantibody [ICA]) was window (0.5, 1.5, 2.5 yr, etc.), is used in each comparison. Clustering was preceded by a relative increase in the lysoPC PC(18:0/0:0) performed for the profi les across all available samples ( n = 419) using and followed by a transient two to fourfold elevation of Ward linkage and the nonparametric Spearman rank correlation – based 2-hydroxybutyric acid ( Fig. 5 C ). Within 9 – 18 mo and 0 – 9 distance metric. The number of subjects included in each age cohort is mo before the appearance of IAA, ketoleucine was 20-fold shown at the bottom. JEM VOL. 205, December 22, 2008 2979 and 8.4-fold (0 – 9 mo; P = 0.02) before seroconversion to We found that the children who developed type 1 diabe- IAA positivity. tes have reduced serum levels of succinic acid and PC at birth and reduced levels of multiple phospholipids and triglycerides DISCUSSION throughout the follow up. As PC is a major source of choline In this longitudinal study, we compared serum metabolome in in the body, our data suggest that the progressors are choline- children who, during follow up, progressed to type 1 diabetes- defi cient since birth. Besides being an important epigenetic associated autoimmunity and, further, to clinical diabetes regulator ( 17 ), choline also controls the secretion of triglyc- (progressors) or who remained permanently healthy and au- eride-rich very low-density lipoprotein particles. Choline toantibody negative. Our study strongly suggests that metabolic defi ciency leads to lower serum triglyceride levels and their dysregulation precedes overt autoimmunity in type 1 diabetes. increased accumulation in liver ( 18 ). This is consistent with our observation that during infancy and early childhood the progressors have low triglyceride levels. Although diminished choline values were observed in the progressors already at birth, nongenetic regulation cannot be excluded because, for example, choline metabolism is dependent on the composi- tion of the intestinal microbiota ( 19 ). Consistent with this, succinic acid, which is also diminished in the progressors at birth, is as well produced by the human intestinal microbiota ( 20, 21 ). Interaction of gut microbiota with the innate im- mune system is an important factor aff ecting T1D predisposi- tion in the nonobese diabetic mouse ( 22 ). Accordingly, it is conceivable that maternal diet or intestinal fl ora during preg- nancy may aff ect the choline and succinic acid levels in new- born infants and thereby modify systemic energy metabolism and the immune system in the off spring. Our data show that ether PCs were consistently low in the progressors. The plasmalogens, which are the most abundant ether phospholipids in normal serum, protect cells from oxi- dative damage ( 23 ). Plasmalogen defi ciency may also indicate that subjects who are genetically susceptible to type 1 diabetes are particularly prone to specifi c metabolic stress, which not only leads to plasmalogen defi ciency but also triggers progres- sion toward diabetes. One potential pathway linking low Figure 5. Selected metabolite differences between progressors and nonprogressors within 18-mo periods before and after seroconver- sion to islet autoantibody positivity, divided into four 9-mo periods. (A) Illustrative matching of progressors and nonprogressors for studies of seroconversion-related changes in metabolome. For each progressor ’ s samples near the selected period of seroconversion, the age-matched control samples were selected from the matched nonprogressor. (B) Changes before (Ab ) and after (Ab ) fi rst seroconversion to islet auto- antibody positivity. Only one sample from each subject, closest to the time of seroconversion, is included within each time period. Clustering was performed as described in Fig. 4 . (C) Changes before and after seroconver- sion to positivity for each autoantibody. Only one sample from each sub- Figure 6. Biochemical model of metabolic changes in progressors ject, closest to the time of seroconversion, is included within each time to type 1 diabetes prior and after seroconversion to islet autoim- period. Lipids were measured in 56 progressors and 73 nonprogressors, munity. Detected changes of specifi c metabolites of citric acid cycle and whereas the metabolites were measured in 13 progressors and 26 non- glutamic acid metabolism before (left boxes) and shortly after (right progressors as shown in Table S1. boxes) the appearance of GADA are shown. 2980 METABOLOME EN ROUTE TO TYPE 1 DIABETES | Ore š ic ˇ et al. ARTICLE phospholipids with the increased risk of type 1 diabetes is the spectively collected sample series, which covered the time enhanced susceptibility of pancreatic  cells to increased oxi- from birth to overt type 1 diabetes. However, it is essential that dative damage ( 24 ). Decreased antioxidant capacity of  cells these results are validated using other well characterized popu- caused by low ether phospholipids might so enhance their lation cohorts. The German BabyDiab ( 36 ), the US-based susceptibility to the eff ects of reactive oxygen species. DAISY ( 37 ) and PANDA studies ( 38 ), and the multinational The appearance of each islet autoantibody was preceded TEDDY study ( 39 ) will hopefully, in time, be able to confi rm by elevated serum concentration of lysoPCs, which are reac- or reject our fi ndings. To address the hypotheses derived from tive lipid byproducts generated in the hydrolysis of PCs by our fi ndings and to understand the tissue-specifi c mechanisms phospholipase A2. These are lipid mediators able to activate behind the early metabolic disturbances and associated autoim- a range of proinfl ammatory molecules ( 25 ) that function as mune changes preceding the disease onset, experimental mod- natural adjuvants for the immune system ( 26 ). The increased els refl ecting both the metabolic and autoimmune components levels of lysoPCs observed just a few months before serocon- of early type 1 diabetes will have to be established. version to autoantibody positivity may be consistent with the In conclusion, dysregulation of metabolism precedes eff ects of increased oxidative stress associated with, for exam- cell autoimmunity and overt type 1 diabetes. The factors ple, a viral infection ( 2 ) or some other proinfl ammatory event. leading to metabolic stress and autoimmune responses clearly The appearances of autoantibodies against insulin and need to be investigated in further studies in the context of GADA, an enzyme that decarboxylates glutamic acid to autoimmune diseases in general. Our fi ndings also imply that GABA, were preceded by diminished ketoleucine, elevated metabolic or immunomodulatory interventions during the BCAAs, and elevated glutamic acid. Augmentation of insulin preautoimmune period may be used as a potential strategy for secretion by elevated serum BCAAs is well documented ( 27 ) prevention of type 1 diabetes. and high blood levels of BCAAs were previously reported in overt type 1 diabetes ( 28 ). BCAA and glutamic acid levels are MATERIALS AND METHODS restored to normal after the appearance of GADA or IAA DIPP study. The DIPP project has been performed in three Finnish cities ( Fig. 6 ). This raises an intriguing possibility that the initial with a combined annual birth rate of 11,000, representing almost 20% of all births in Finland. The project was launched in the city of Turku in Novem- autoimmune response is physiological and aimed at restoring ber 1994; Oulu joined the study 1 yr later and Tampere 2 yr after that. HLA- the metabolic homeostasis and that the disease may be caused DQ – based genetic typing was used to select children positive for DR3-DQ2 by or be at least partially infl uenced by a defective response and/or DR4-DQ8 risk haplotypes without protective haplotypes ( 6 ). By toward the  cell autoantigens ( 29, 30 ). A protective role of June 6, 2006, 104,111 consecutive newborn infants had been screened, and the diabetes-associated autoantibodies is indirectly supported 8,026 children with genetic risk continued in the follow up. The DIPP study by the observation that the autoantibody titers frequently de- was approved by the local Ethical Committees at the University Hospitals in Turku, Oulu, and Tampere. crease before the presentation of clinical diabetes, which can In Turku the children were monitored at 3-mo intervals until 2 yr of be interpreted as a sign of failing protection ( 31 ). age and then were monitored twice a year, and in Oulu and Tampere they Our fi ndings favor early immunomodulation rather than were monitored at 3, 6, 12, 18, and 24 mo and then annually ( 40 ). At each immunosuppression as a preventive therapy for type 1 diabetes, visit, a venous blood sample was collected from the children without fasting. with the aim of boosting the benefi cial components of auto- After 30 – 60 min at room temperature, serum was separated and transferred immunity. However, despite apparent effi cacy in animal mod- to  70 ° C in cryovials within 3 h from the draw. ICA were measured in all samples. If a child seroconverted to ICA positivity, IAA, GADA, and els of several immunomodulatory agents, success has been poor protein tyrosine phosphatase-like antigen IA-2 (IA-2A) were measured in all in human type 1 diabetes prevention trials, including our re- samples obtained from that child. After seroconversion to autoantibody cent double-blind placebo-controlled nasal insulin study ( 32 ). positivity the visits took place at 3-mo intervals in all clinical centers. As of Notably, in that study the intervention was conducted in January 1, 2003, all four autoantibody types, ICA, GADA, IAA, and IA-2A, children who recently had seroconverted to positivity for two were measured from all samples drawn as previously described ( 32 ). Auto- or more types of diabetes-associated autoantibodies. Our antibodies that were present in cord blood and disappeared before 18 mo of age were regarded as maternal in origin and such values were excluded from fi ndings suggest that this period may already be late for such the analysis. a therapeutic intervention and that the preautoimmune period Between 1997 and 2007, study children with two or more types of au- could be a more suitable window for disease prevention. toantibodies detected in at least two consecutive serum samples have been Early therapeutic strategies aimed at improving or restor- recruited into a randomized double-blind placebo-controlled trial testing the ing the metabolic phenotype, e.g., by reestablishing ether effi cacy of nasal insulin for preventing type 1 diabetes (http://clinicaltrials phospholipid availability or by modulating gut microbial .gov, identifi er: NCT00223613). The study showed that daily nasal insulin administration in an  1 U/kg dose failed to delay diabetes development in composition, might also be of value by preventing  cell de- children with genetic risk and two or more types of autoantibodies ( 32 ). struction and delaying progression to overt type 1 diabetes. In There were no meaningful diff erences in serum metabolome components in support of this, administration of docosahexaenoic acid, the enrolled children, who comprised roughly half of those eligible, com- which is known to be selectively targeted to plasmalogens pared with those who did not participate in the prevention trial. ( 33 ), has been associated with decreased risk of  cell autoim- munity and type 1 diabetes ( 34, 35 ). Lipidomic analysis. 10- μ l serum samples diluted with 10 μ l of 0.15 M One of the strengths of our study is that we used state of NaCl and spiked with a standard mixture containing 10 lipid species the art metabolomics technology to examine our large pro- were extracted with 100 μ l of a 2:1 mixture of chlorofor m and methanol. JEM VOL. 205, December 22, 2008 2981 The extraction time was 0.5 h and the lower organic phase was separated by m/z peaks having more weight than the lower. A similarity value is assigned centrifuging at 10,000 rpm for 3 min. Another standard mixture containing between 0 and 999, with 999 being a perfect match and 750 generally con- three labeled lipid species was added to the extracts and the lipids were ana- sidered as a reasonable match. We used the conservative cut-off criterion of lyzed on a mass spectrometer (Q-Tof Premier; Waters) combined with an 850 for identifi cation. Response curves are shown for selected metabolites Acquity UPLC (Waters). The column, kept at 50 ° C, was an Acquity UPLC of relevance to the study in Figs. S8 and S9 (available at http://www.jem BEH C18, 50 mm, with 1.7- μ m particles (Waters). The solvent system in- .org/cgi/content/full/jem.20081800/DC1). cluded water (1% 1 M ammonium acetate and 0.1% HCOOH) and a mix- ture of acetonitrile and 2-propanol (5:2; 1% 1 M NH Ac and 0.1% HCOOH). 4 Statistical methods. R statistical software (http://www.r-project.org/) The fl ow rate was 0.2 ml/min and the total run time, including column re- was used for data analyses and visualization. The concentrations were equilibration, was 18 min. compared using the Wilcoxon rank-sum test, with p-values < 0.05 con- Raw centroid data from the LC/MS instrument was converted to sidered signifi cant. To account for multiple comparisons, false discovery netCDF fi les, which were processed with MZmine software version 0.60 ( 41, rate was estimated as the maximum q -value ( 43 ) in the set of signifi cant 42 ). Calibration was performed as follows: all monoacyl lipids except choles- diff erences for each dataset (i.e., lipidomics or metabolomics) and the terol esters, such as monoacylglycerols and monoacyl glycerophospholipids, family of hypotheses tested (e.g., diff erences between progressors and were calibrated with the lysoPC PC(17:0/0:0) as internal standard. All nonprogressors in a specifi c age cohort). False discovery rates were com- diacyl lipids except phosphatidylethanolamines were calibrated with the puted using the R package qvalue. Lipid class concentration diff erences PC PC(17:0/17:0), the phosphatidylethanolamines with PE(17:0/17:0), between the mean longitudinal profi les of progressors and nonprogressors and the triacylglycerols and cholesterol esters with the triacylglycerol were tested using linear mixed eff ects models implemented with the lme TG(17:0/17:0/17:0). Calibration-based concentrations are expressed in μ mol/ function in the R package nlme. The fi xed eff ects were modeled as con- liter unless otherwise noted. stant levels for each group, i.e., for progressors and nonprogressors, and The samples were analyzed in four separate runs within 12 mo and the the random eff ects were modeled as constant deviations from these con- data were processed for each analytical run separately (Table S1, available at stant group level trajectories. The signifi cance of the group diff erences http://www.jem.org/cgi/content/full/jem.20081800/DC1). For the fi nal was evaluated by the p-value for the fi xed eff ect parameter estimate of combined dataset consisting of data from all four analytical runs, we matched group diff erences. the lipids across diff erent runs based on the following criteria: lipid identity The fold diff erence was calculated by dividing the median concentra- is the same across all runs; lipid is detected in all four analytical runs (this tion in progressors by the median concentration in nonprogressors. If this stringent criterion is the main reason for relatively low number of lipids in number was less than one, the negative is listed (e.g., 0.75, or a drop of 25% vs. the fi nal dataset and, thus, only the most abundant lipids within each class are nonprogressors, is reported as a  1.3-fold diff erence). included); starting from the fi rst analytical run as reference (Turku DIPP se- ries), retention time deviation is < 2%; and starting from the fi rst analytical Online supplemental material. Fig. S1 shows lysoPC levels across all run as reference (Turku DIPP series), deviation of median intensity (calcu- 1,196 samples measured as a function of sample storage time. Fig. S2 shows lated within the analytical run) between analytical runs is < 50%. the total PC concentration in cord serum of progressors and nonprogressors. Fig. S3 compares the lipidomes between children with the high-risk HLA genotype and those with moderate-risk HLA genotypes. Fig. S4 shows the Metabolomic analysis using the GCxGC-TOF/MS platform. 10 μ l of concentrations of ethanolamine plasmalogen detected in a subset of progres- an internal standard-labeled palmitic acid (16:0-16,16,16d ; 500 mg/l) and sors and nonprogressors. Diff erences in selected metabolite concentrations 400 μ l of methanol solvent were added to 20 μ l of the sample. After vortex- between the progressors to type 1 diabetes and nonprogressors within the ing for 1 min and incubating for 30 min at room temperature, the superna- four 9-mo intervals before disease diagnosis in the progressors are shown in tant was separated by centrifugation at 10,000 rpm for 5 min at room Fig. S5. Associations of lysoPC and ether PC with the rate of progression to temperature. The sample was dried under constant fl ow of nitrogen gas. 25 μ l type 1 diabetes are shown in Figs. S6 and S7, respectively. Figs. S8 and S9 MOX (2% methoxyamine HCl in pyridine) was added to the dried sample. show the selected characteristics of the GCxGC-TOF/MS analytical plat- The mixture was then incubated at 45 ° C for 1 h and derivatized with 25 μ l form. Table S1 provides an overview of the four analytical runs used to gen- N -methyl- N -(trimethylsilyl)trifl uoroacetamide by incubating at 45 ° C for 1 h. erate the data for this study. Online supplemental material is available at 10 μ l of retention index standard mixture with fi ve alkanes at 800 ppm was http://www.jem.org/cgi/content/full/jem.20081800/DC1. added to the metabolite mixture. Sample order for analysis was established by randomization. We thank the dedicated personnel of the DIPP study in Turku, Oulu, and Tampere The instrument used was a GCxGC-TOF mass spectrometer (Pegasus and the DIPP families for their support. We acknowledge the technical support of 4D; Leco) with an autosampler (6890N GC and Combi PAL;Agilent Tech- Leena Ö hrnberg, Anna-Liisa Ruskeep ä ä , Jarkko Miettinen, Airi Hyrk ä s, and Jing Tang. nologies). The instrument parameters were as follows: 1- μ l split injection 1:20; We thank Antonio Vidal-Puig, Fredrik B ä ckhed, and Kari Alitalo for discussions fi rst column, RTX-5, 10 m × 180 μ m × 0.20 μ m; second column, BPX-50, and comments. 1.50 m × 100 μ m × 0.10 μ m; helium 39.6 psig constant pressure; temperature The study was funded by Juvenile Diabetes Research Foundation International programs, primary oven, Initial 50 ° C, 1 min → 295 ° C, 7 ° C/min, 3 min, and (4-1999-731 and 4-2001-435 to O. Simell), Tekes FinnWell Program (M. Ore š ic ˇ , R. secondary oven, +20 ° C above primary oven temperature; second dimension Lahesmaa, and O. Simell), DIAPREPP project of the Seventh Framework Program of the separation time 5 s; MS measurement 45 – 700 amu, 100 spectra/s. Raw data European Community (HEALTH-F2-2008-202013 to M. Ore š ic ˇ and O. Simell), Academy were processed using ChromaTOF software (Leco) followed by alignment of Finland, Sigrid Juselius Foundation, P ä ivikki and Sakari Sohlberg Foundation, Signe and normalization using the in house – developed software. Unwanted back- and Ane Gyllenberg Foundation, Diabetes Research Foundation, Finland, and Special ground peaks were eliminated using the classifi cations feature of ChromaTOF Federal Funds to the Turku, Oulu, and Tampere University Hospitals. software. In house – developed software was used to perform additional fi ltering The authors have no confl icting fi nancial interests. using compound identifi cations by ChromaTOF. The metabolites were identifi ed using an in house reference compound Submitted: 12 August 2008 library as well as by searching the reference mass spectral library. Mass spectra Accepted: 18 November 2008 from the GCxGC-TOF/MS analysis were searched against The Palisade Complete Mass Spectral Library, 600K Edition (Palisade Mass Spectrome- try), which includes all spectra available from the NIST 2002 and Wiley reg- REFERENCES istry collections as well as 150,000 other spectra. The matches to reference 1 . Gale , E.A.M. 2002 . The rise of childhood type 1 diabetes in the 20th spectra are based on a weighted dot product of the two spectra, with higher century. Diabetes . 51 : 3353 – 3361 . 2982 METABOLOME EN ROUTE TO TYPE 1 DIABETES | Ore š ic ˇ et al. ARTICLE 2 . Knip , M. , R. Veijola , S.M. Virtanen , H. Hy ö ty , O. Vaarala , and H.K. 21 . Gill , S.R. , M. Pop , R.T. DeBoy , P.B. Eckburg , P.J. Turnbaugh , B.S. Å kerblom . 2005 . Environmental triggers and determinants of type 1 Samuel , J.I. Gordon , D.A. Relman , C.M. Fraser-Liggett , and K.E. diabetes. Diabetes . 54 : S125 – S136 . Nelson . 2006 . Metagenomic analysis of the human distal gut microbi- 3 . Tuomilehto , J. , M. Karvonen , J. Pitk ä niemi , E. Virtala , K. Kohtam ä ki , ome. Science . 312 : 1355 – 1359 . L. Toivanen , and E. Tuomilehto-Wolf . 1999 . Record-high incidence 22 . Wen , L. , R.E. Ley , P.Y. Volchkov , P.B. Stranges , L. Avanesyan , A.C. of type 1 (insulin-dependent) diabetes mellitus in Finnish children. Stonebraker , C. Hu , F.S. Wong , G.L. Szot , J.A. Bluestone , et al . 2008 . Diabetologia . 42 : 655 – 660 . Innate immunity and intestinal microbiota in the development of Type 4 . Harjutsalo , V. , L. Sj ö berg , and J. Tuomilehto . 2008 . Time trends in the 1 diabetes. Nature . 455 : 1109 – 1113 . incidence of type 1 diabetes in Finnish children: a cohort study. Lancet . 23 . Zoeller , R.A. , T.J. Grazia , P. LaCamera , J. Park , D.P. Gaposchkin , 371 : 1777 – 1782 . and H.W. Farber . 2002 . Increasing plasmalogen levels protects human 5 . Achenbach , P. , E. Bonifacio , K. Koczwara , and A.-G. Ziegler . 2005 . endothelial cells during hypoxia. Am. J. Physiol. Heart Circ. Physiol. Natural history of type 1 diabetes. Diabetes . 54 : S25 – S31 . 283 : H671 – H679 . 6 . Kupila , A. , P. Muona , T. Simell , P. Arvilommi , H. Savolainen , A.-M. 24 . Lenzen , S. , J. Drinkgern , and M. Tiedge . 1996 . Low antioxidant enzyme H ä m ä l ä inen , S. Korhonen , T. Kimpim ä ki , M. Sj å roos , J. Ilonen , et al . gene expression in pancreatic islets compared with various other mouse 2001 . Feasibility of genetic and immunological prediction of type 1 dia- tissues. Free Radic. Biol. Med. 20 : 463 – 466 . betes in a population-based birth cohort. Diabetologia . 44 : 290 – 297 . 25 . Mehta , D. 2005 . Lysophosphatidylcholine: an enigmatic lysolipid. Am. 7 . Ore š i c ˇ , M. , A. Vidal-Puig , and V. H ä nninen . 2006 . Metabolomic ap- J. Physiol. Lung Cell. Mol. Physiol. 289 : L174 – L175 . proaches to phenotype characterization and applications to complex dis- 26 . Kabarowski , J.H.S. , Y. Xu , and O.N. Witte . 2002 . Lysophospha- eases. Expert Rev. Mol. Diagn. 6 : 575 – 585 . tidylcholine as a ligand for immunoregulation. Biochem. Pharmacol. 8 . Nikkil ä , J. , M. Sysi-Aho , A. Ermolov , T. Sepp ä nen-Laakso , O. Simell , 64 : 161 – 167 . S. Kaski , and M. Ore š ic ˇ . 2008 . Gender dependent progression of sys- 27 . Floyd , J.C.J. , S.S. Fajans , R.F. Knopf , and J.W. Conn . 1963 . Evidence temic metabolic states in early childhood. Mol. Syst. Biol. 4 : 197 . that insulin release is the mechanism for experimentally induced leucine 9 . Lenz , E.M. , J. Bright , I.D. Wilson , A. Hughes , J. Morrisson , H. hypoglycemia in man. J. Clin. Invest. 42 : 1714 – 1719 . 28 . Marliss , E.B. , A.F. Nakhooda , P. Poussier , and A.A. Sima . 1982 . The Lindberg , and A. Lockton . 2004 . Metabonomics, dietary infl uences and cultural diff erences: a 1H NMR-based study of urine samples obtained diabetic syndrome of the ‘ BB ’ Wistar rat: possible relevance to type 1 (insulin-dependent) diabetes in man. Diabetologia . 22 : 225 – 232 . from healthy British and Swedish subjects. J. Pharm. Biomed. Anal. 36 : 841 – 849 . 29 . Nevo , U. , I. Golding , A.U. Neumann , M. Schwartz , and S. Akselrod . 2004 . Autoimmunity as an immune defense against degenerative pro- 10 . B ä ckhed , F. , H. Ding , T. Wang , L.V. Hooper , G.Y. Koh , A. Nagy , C.F. Semenkovich , and J.I. Gordon . 2004 . The gut microbiota as an en- cesses: a primary mathematical model illustrating the bright side of auto- immunity. J. Theor. Biol. 227 : 583 – 592 . vironmental factor that regulates fat storage. Proc. Natl. Acad. Sci. USA . 101 : 15718 – 15723 . 30 . Hauben , E. , M.G. Roncarolo , U. Nevo , and M. Schwartz . 2005 . Benefi cial autoimmunity in type 1 diabetes mellitus. Trends Immunol. 11 . Gluckman , P.D. , M.A. Hanson , C. Cooper , and K.L. Thornburg . 2008 . Eff ect of in utero and early-life conditions on adult health and disease. 26 : 248 – 253 . 31 . Kimpimaki , T. , P. Kulmala , K. Savola , A. Kupila , S. Korhonen , T. N. Engl. J. Med. 359 : 61 – 73 . 12 . Simell , O. , H. Niinikoski , T. R ö nnemaa , H. Lapinleimu , T. Routi , Simell , J. Ilonen , O. Simell , and M. Knip . 2002 . Natural history of { beta } -cell autoimmunity in young children with increased genetic H. Lagstr ö m , P. Salo , E. Jokinen , and J. Viikari . 2000 . Special Turku Coronary Risk Factor Intervention Project for Babies (STRIP). Am. susceptibility to type 1 diabetes recruited from the general population. J. Clin. Endocrinol. Metab. 87 : 4572 – 4579 . J. Clin. Nutr. 72 : 1316S – 1331S . 13 . Watson , A.D. 2006 . Thematic review series: systems biology ap- 32 . N ä nt ö -Salonen , K. , A. Kupila , S. Simell , H. Siljander , T. Salonsaari , A. Hekkala , S. Korhonen , R. Erkkola , J.I. Sipil ä , L. Haavisto , et al . 2008 . proaches to metabolic and cardiovascular disorders. Lipidomics: a global approach to lipid analysis in biological systems. J. Lipid Res. Nasal insulin to prevent type 1 diabetes in children with HLA genotypes and autoantibodies conferring increased risk of disease: a double-blind, 47 : 2101 – 2111 . 14 . Ore š ic ˇ , M. , V.A. H ä nninen , and A. Vidal-Puig . 2008 . Lipidomics: a randomised controlled trial. Lancet . 372 : 1746 – 1755 . 33 . Gaposchkin , D.P. , and R.A. Zoeller . 1999 . Plasmalogen status infl u- new window to biomedical frontiers. Trends Biotechnol. 26 : 647 – 652 . 15 . Laaksonen , R. , M. Katajamaa , H. P ä iv ä , M. Sysi-Aho , L. Saarinen , P. ences docosahexaenoic acid levels in a macrophage cell line: insights using ether lipid-defi cient variants. J. Lipid Res. 40 : 495 – 503 . Junni , D. L ü tjohann , J. Smet , R.V. Coster , T. Sepp ä nen-Laakso , et al . 2006 . A systems biology strategy reveals biological pathways and plasma 34 . Stene , L.C. , and G. Joner , and the Norwegian Childhood Diabetes Study Group. 2003 . Use of cod liver oil during the fi rst year of biomarker candidates for potentially toxic statin induced changes in muscle. PLoS ONE . 1 : e97 . life is associated with lower risk of childhood-onset type 1 diabe- tes: a large, population-based, case-control study. Am. J. Clin. Nutr. 16 . Welthagen , W. , R.A. Shellie , J. Spranger , M. Ristow , R. Zimmermann , and O. Fiehn . 2005 . Comprehensive two-dimensional gas chroma- 78 : 1128 – 1134 . 35 . Norris , J.M. , X. Yin , M.M. Lamb , K. Barriga , J. Seifert , M. tography time-of-fl ight mass spectrometry (GC — GC-TOF) for high resolution metabolomics: biomarker discovery on spleen tissue ex- Hoff man , H.D. Orton , A.E. Baron , M. Clare-Salzler , H.P. Chase , et al . 2007 . Omega-3 polyunsaturated fatty acid intake and islet au- tracts of obese NZO compared to lean C57BL/6 mice. Metabolomics . toimmunity in children at increased risk for type 1 diabetes. JAMA . 1 : 65 – 73 . 298 : 1420 – 1428 . 17 . Niculescu , M.D. , C.N. Craciunescu , and S.H. Zeisel . 2006 . Dietary cho- 36 . Achenbach , P. , K. Koczwara , A. Knopff , H. Naserke , A.-G. Ziegler , line defi ciency alters global and gene-specifi c DNA methylation in the and E. Bonifacio . 2004 . Mature high-affi nity immune responses to developing hippocampus of mouse fetal brains. FASEB J. 20 : 43 – 49 . (pro)insulin anticipate the autoimmune cascade that leads to type 1 dia- 18 . Yao , Z.M. , and D. Vance . 1988 . The active synthesis of phosphatidyl- betes. J. Clin. Invest. 114 : 589 – 597 . choline is required for very low density lipoprotein secretion from rat 37 . Baschal , E.E. , T.A. Aly , S.R. Babu , M.S. Fernando , L. Yu , D. Miao , hepatocytes. J. Biol. Chem. 263 : 2998 – 3004 . 19 . Dumas , M.-E. , R.H. Barton , A. Toye , O. Cloarec , C. Blancher, A. K.J. Barriga , J.M. Norris , J.A. Noble , H.A. Erlich , et al . 2007 . HLA- DPB1*0402 protects against type 1A diabetic autoimmunity in the Rothwell , J. Fearnside , R. Tatoud , V. Blanc , J.C. Lindon , et al . 2006 . Metabolic profi ling reveals a contribution of gut microbiota to fatty highest risk DR3-DQB1*0201/DR4-DQB1*0302 DAISY population. Diabetes . 56 : 2405 – 2409 . liver phenotype in insulin-resistant mice. Proc. Natl. Acad. Sci. USA . 103 : 12511 – 12516 . 38 . Carmichael , S.K. , S.B. Johnson , A. Baughcum , K. North , D. Hopkins , 20 . Macy , J.M. , L.G. Ljungdahl , and G. Gottschalk . 1978 . Pathway of suc- M.G. Dukes , J.X. She , and D.A. Schatz . 2003 . Prospective assessment cinate and propionate formation in Bacteroides fragilis . J. Bacteriol. in newborns of diabetes autoimmunity (PANDA): maternal understand- 134 : 84 – 91 . ing of infant diabetes risk. Genet. Med. 5 : 77 – 83 . JEM VOL. 205, December 22, 2008 2983 39 . Hagopian , W.A. , A. Lernmark , M.J. Rewers , O.G. Simell , J.-X. She , 41 . Katajamaa , M. , and M. Ore š ic ˇ . 2005 . Processing methods for dif- A.G. Ziegler , J.P. Krischer , and B. Akolkar . 2006 . TEDDY - the en- ferential analysis of LC/MS profi le data. BMC Bioinformatics . vironmental determinants of diabetes in the young: an observational 6 : 179 – 190 . clinical trial. Ann. N. Y. Acad. Sci. 1079 : 320 – 326 . 42 . Katajamaa , M. , J. Miettinen , and M. Ore š ic ˇ . 2006 . MZmine: toolbox 40 . Kupila , A. , P. Keskinen , T. Simell , S. Erkkil ä , P. Arvilommi , S. for processing and visualization of mass spectrometry based molecular Korhonen , T. Kimpim ä ki , M. Sj ö roos , M. Ronkainen , J. Ilonen , profi le data. Bioinformatics . 22 : 634 – 636 . et al . 2002 . Genetic risk determines the emergence of diabetes-associated 43 . Storey , J.D. 2002 . A direct approach to false discovery rates. Journal of the autoantibodies in young children. Diabetes . 51 : 646 – 651 . Royal Statistical Society Series B . 64 : 479 – 498 . 2984 METABOLOME EN ROUTE TO TYPE 1 DIABETES | Ore š ic ˇ et al.

Journal

The Journal of Experimental Medicine – Pubmed Central

Published: Dec 22, 2008

There are no references for this article.